Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
Tilburg University
Texas Hold’em
van der Genugten, B.B.; Borm, P.E.M.
Document version:Early version, also known as pre-print
Publication date:2014
Link to publication
Citation for published version (APA):van der Genugten, B. B., & Borm, P. E. M. (2014). Texas Hold’em: A Game of Skill. Tilburg: Department ofEconometrics.
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
- Users may download and print one copy of any publication from the public portal for the purpose of private study or research - You may not further distribute the material or use it for any profit-making activity or commercial gain - You may freely distribute the URL identifying the publication in the public portal
Take down policyIf you believe that this document breaches copyright, please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Download date: 11. Apr. 2020
1
Texas Hold’em: a game of skill
Ben van der Genugten and Peter Borm
CentER and Department of Econometrics and OR, Tilburg University, Tilburg, The Netherlands
Abstract. On the basis of the detailed mathematical analysis of realistic approximations of Texas
Hold’em cash games and tournaments as provided in [1] and [2], respectively, this paper reconfirms
the findings of the court in The Hague [3] that on the basis of the methodology of relative skill it can
be concluded that Texas Hold’em in all its practical variants is a game of skill.
1. Introduction: on the methodology of relative skill.
The distinction between games of chance and games of skill is a prominent topic due to the rules set
out in the gaming acts of various countries. This paper will focus on this distinction within the
framework of the Dutch gaming act. Since different gaming acts typically have the same
characteristics our conclusions however do not only apply to the situation in the Netherlands but can
be applied in a rather broad and general sense. The game under consideration here is a specific
variant of poker: Texas Hold’em. Especially this variant of poker has gained worldwide popularity
over the past ten years, not only in casinos and on internet but also as the poker game played and
advertised in championships broadcasted on television throughout the world.
Roughly speaking the Dutch gaming act states that is not allowed to exploit, without license, games
with monetary payoffs in which the participants do not have a predominant influence on their game
results. These games are called games of chance. Since influencing game results requires skill
(studying, experience, counting cards, training memory, observational skills, etc.) other games are
called games of skill. Consequently, each game with monetary payoffs involving chance elements
(drawing cards, spinning of a wheel, throwing dice, etc.) in principle can be classified either as a game
of chance or as a game of skill. Of course the crucial word in the gaming act is predominant, which
means that the influence of skill should be weighed against the influence of chance.
The influence of skill is referred to as the learning effect LE in a game. Thus, the learning effect
measures the maximal difference in game result directly due to the influence or skill of players
themselves. It can be represented by the additional gains a player is able to obtain in evolving from a
serious but inexperienced beginner into a well experienced and established player who has mastered
all intricacies of the game: LE ≥ 0. In fact, empirical observations of real game play could provide valid
indications of the absolute size of the learning effect.
Regarding cash poker and tournament poker, a wide variety of books (cf. [4], [5], [6] and [7]), articles
(cf. [8] and [9]), and internet sites (cf. [10]) exclusively focus on the learning effect and, in particular,
on improving actual play by creating a player’s awareness by discussing general game aspects and
providing tips and tricks, by devising and analyzing theoretically strong poker strategies and/or by
analyzing expert play in practice. None of these studies however touch the issue of the influence of
2
an expert strategy to be predominant in a methodological way. For this reason these studies have
and will be ignored in the legal discussion of how to qualify Texas Hold’em according to a gaming act.
To adequately measure if players are able to execute predominant influence on their game result,
the learning effect should rather be considered in a relative sense. For this reason we compare the
learning effect with the so-called random effect RE that measures to what extent a skilful and
advanced player is restricted in further improving his gains by the presence of the chance elements.
It can be represented by the additional gains an experienced player would be able to obtain in the
fictive situation that he would know (but cannot influence) the outcomes of all possible chance
elements in advance before having to decide on his actions in the game: RE ≥ 0. It will be clear that
empirical observations of game play will not provide information about the absolute size of the
random effect. The random effect has to be calculated using an explicit and tailor-made analysis on
the basis of the rules and specifications of the game at hand. In a controlled experimental setting
however information on the random effect can be obtained as is shown and documented for the
specific case of Texas Hold’em in [11].
Having obtained the two numbers LE and RE in monetary units, the relative level of skill S is defined
as the quotient of LE and LE+RE: S = LE / (LE+RE). We want to note that games in which both the
learning effect and the random effect equal 0 are of no practical interest and are therefore left out of
consideration. It will be clear that the level of relative skill S is a number between 0 and 1.
A game with LE=0, i.e., if no player can improve upon the results of a beginner, has a level of relative
skill equal to 0. Such games are called pure games of chance. On the other hand, a game with RE=0,
i.e., if the chance elements put no restriction the results of a skilful player, has a level of relative skill
equal to 1. Such games are called pure games of skill. Although there are many different types of
games, all with different characteristics regarding game material, chance generators, number of
players, complexity etc., the methodology of relative skill can in principle be applied to any strategic
game where a player’s influence on his game results is through the quality of his strategy selection
only. A more detailed discussion on the definitions of learning effect, random effect, level of relative
skill and strategic games can be found in [12] and [13]. The way how to exactly quantify the level of
relative skill is elaborated upon in the next section.
The above described methodology of relative skill has shown its value as an operational tool in
practice and, moreover, turned out to be decisive in courts of law. The application of the
methodology of relative skill is flexible and mature in the sense that it requires a tailor-made
quantitative analysis for each game separately although the basic qualitative ideas and the
underlying conceptual framework have remained the same in all specific cases at hand throughout
the years. It has been not only been applied to one-person games as Roulette (S=0 for French
roulette), Caribbean stud poker (S= 0.003), and Blackjack (S =0.05 in Holland Casinos), and many,
many more, but also to more-person games with strong interaction between the players like cash
and tournament poker variants. For a partial overview on levels of relative skill for various types of
games we refer to [14]. Importantly the methodology of relative skill has played a decisive role in
court. The most recent example of this type can be found in [3] in which the court of The Hague
integrally accepts the methodology of relative skill and its underlying conceptual reasoning as
presented in [2] and concludes that the Texas Hold’em cash games and tournaments under
consideration could not be viewed as games of chance. At an earlier stage the court of Arnhem [15]
3
ruled on the basis of [16] and [17] that particular types of interactive more-person management
games like Grand Prix manager and Competition manager, with levels of relative skill approximately
equal to 0.3, should not be considered as games of chance.
As its main contribution the methodology of relative skill offers an objective and mathematical tool
to order and compare different types of strategic games on the basis of well-defined relevant
concepts and terminology and consistent reasoning. Moreover, it is legally constructive in the sense
that it offers the possibility to set a legal skill threshold level for the level of relative skill above which
a game should be considered a game of skill. Allowing for consistent legislation on the basis of the
ruling [15] on management games and the fact that no variant of the (legally classified) game of
chance Blackjack has a level of relative skill above 0.2, as is seen in [18] and [19], the skill threshold
level should be set between 0.2 and 0.3. In particular, if for a specific game the level of relative skill
well exceeds 0.3, then this game should be legally classified as a game of skill.
Due to the conceptual choices underlying the computation of the learning effect and the random
effect the methodology of relative skill has one additional advantage, especially in games with
several interacting players like Texas Hold’em. It requires the input of only one subjective parameter:
the strategic behaviour of a beginner. In this choice regarding the behaviour of a beginner the social
impact of the game and the general level of the participants in practice can be incorporated.
Moreover, the choice can also be based on statistical empirical evidence or professional knowledge.
This helps to better focus the discussion about the skill level of a game to its core features. From a
methodological perspective this input parameter allows for a sensitivity analysis of relative skill when
varying a particular choice of the strategy of a beginner possibly based on the development of the
game as played in practice over time.
In our mathematical analysis of realistic variants of Texas Hold’em the behaviour of a beginner is
selected carefully. Generally speaking, the format of our beginner’s strategy is generally applicable to
a wide range of complex Poker variants and therefore can already be considered as advanced.
Moreover, given the advanced format, our choice of the concrete parameters defining a beginner’s
strategy for real-life variants of Texas Hold’em is derived from extrapolating the choice made from
several options for a basic Texas Hold’em variant on the basis of having the lowest level of relative
skill. In this sense the approximated levels of relative skill of Texas Hold’em variants can be viewed as
underestimates for the real levels of relative skill.
2. Implementing relative skill
To formally implement the methodology of relative skill for a specific game via the general concepts
of learning effect and random effect described above, we consider three types of players: the
beginner, the optimal player and the fictive player.
The beginner with expected monetary result R(0) models a serious but inexperienced player who has
just mastered the rules of the game and who is endowed with a fixed particular (and typically naive)
strategy which is the methodology’s main choice parameter. Here (and elsewhere) the word
expected is used in the strict sense of probability theory, calculated on the basis of the chance
elements present in the game at hand. These chance elements can be induced both externally (e.g.
4
by the dealing of cards) and internally (by randomization imposed by the players themselves in taking
decisions).
The optimal player with expected monetary result R(m) models an expert player who completely
masters the rules and intricacies (in Texas Hold’em in particular the strategic aspects of bluffing and
sandbagging) of the game to maximize his monetary gains.
The fictive player with expected monetary result R(f) models an expert player who on top of his
knowledge and expertise will be informed about the realizations of all chance elements (of both
external and internal type) and who can adequately use this additional information to further
increase his monetary gains.
After having chosen a fixed beginner’s strategy the monetary results R(0), R(m) and R(f) are
unambiguously defined for one-player games with no strategic interaction between participants like
Roulette and Blackjack. For more-person games with strategic interaction between the participants
like Texas Hold’em there are several complications to deal with. Not only can the game involve
different player roles but typically the monetary result of each player type in a certain role will also
depend on the strategic behaviour of its opponents in the game. To allow for a consistent analysis,
for each player role we will calculate the monetary results of the three player types assuming that all
opponents are beginners and therefore endowed with the fixed strategy of a beginner (which may
depend on the specific role of such a player), while thereafter we will take the appropriate mean
over the player roles to get the monetary game results R(0), R(m) and R(f) of the three player types.
Subsequently we define the learning effect by LE= R(m)- R(0) and the random effect by RE=R(f)- R(m).
As a consequence the level of relative skill S = LE / (LE + RE) can be rewritten as
S = (R(m) – R(0)) / (R(f) – R(0))
3. On the level of relative skill of Texas Hold’em cash games
The more qualitative analysis of several poker variants provided in [20] concludes that Texas Hold’em
in all its cash variants should not be classified as a game of chance on the basis of the methodology of
relative skill. At this point in time we are able to offer further quantitative evidence for this
conclusion based on explicit computations of the level of relative skill for more realistic variants of
Texas Hold’em. For expositional purposes we will only consider two main variants here: fixed limit
and no limit Texas Hold’em. The document “Texas Hold’em cash games; a mathematical analysis”[1]
contains detailed descriptions of all variants under consideration, the computational set-up and an
extensive overview of the computer output regarding relative skill, also for pot limit and spread limit
variants, including possible variations in measuring a player type’s game result.
As argued before, the central issue in applying the methodology of relative skill is the selection of an
appropriate strategy of a beginner in each possible player role. In our quantification of a strategy of
the beginner’s type we are primarily led by general observations in the popular and more scientific
literature on poker and the findings in the Texas Hold’em experiment reported on in [3]. Roughly
speaking it can be concluded that a beginner in Texas Hold’em does not discriminate between roles
(positions at the table) and will base his decision at any time moment solely on his own (preflop)
5
cards, the available community cards (if applicable), the current pot size and the number of
opponents left. These considerations lead to a general parametrized format of a beginner’s strategy
applicable to all poker variants. To determine adequate values for the various parameters we used
an extrapolation procedure from the choices made for a basic cash poker variant. To motivate these
choices, we considered four possible options for a beginner’s strategy in the basic variant and, to be
on the safe side, we selected the option with the lowest relative level of skill.
A standard fixed limit game of Texas Hold’em typically is played with a complete deck of cards, hosts
between two to ten players, consists of 4 rounds (pre-flop, flop, turn, river), has a bet equal to two
times the ante, allows for a maximum of 3 raises (excluding bet) per round, and has no possibility of
all-in. From these aspects we selected 3 basic variables: the number of players, the number of rounds
and the maximum number of raises per round. Setting the number of players equal to 2, the number
of rounds equal to 2, and the maximum number of raises per round equal to 1 (while adopting the
remaining standard settings), our basic variant of Texas Hold’em is obtained as the starting point of
our analysis. Although less complex (but more tractable), this basic variant contains all important
strategic ingredients of Texas Hold’em itself.
At this point we briefly want to comment on the accuracy of the levels of relative skill derived below.
Our statistical approach will not use a random sample since, due to the characteristics of an optimal
an a fictive player type, the game result of an optimal player will be positively biased towards the
game result of a fictive player, requiring a random sample which from a computational perspective is
too large. This problem is circumvented in our analysis by using a different technique of
representative sampling as described in some detail in [1] to reduce the bias. The resulting accuracy
of the derived levels of relative skill is approximately 0.02.
We now present a compact overview of the quantitative findings obtained in [1]. To maintain a clear
focus, this overview is incomplete. The level of relative skill of the basic variant is at least 0.29. The
marginal impact of varying one of our three parameters while keeping the other two at the basic
level does not lower the level of relative skill. In particular, allowing for 2 or 3 raises per round, the
level of relative skill remains about the same. If, on the other hand, one increases the number of
players to 3, the level of relative skill increases substantially to 0.42, while, if one increases the
number of rounds to 4 the level of relative skill increases substantially too, in this case to 0.37.
Furthermore, a similar type of analysis in a Texas Hold’em no limit setting leads to a level of relative
skill of at least 0.52 for the basic variant while identical marginal effects as above can be observed
when varying the number of possible raises per round, the number of players and the number of the
rounds.
The overall conclusion is that the relative level of skill of Texas Hold’em cash variants will well exceed
the skill threshold level which, as argued before, should be set between 0.2 and 0.3.
4. On the level of relative skill for Texas Hold’em tournaments
Texas Hold’em tournaments are rapidly gaining popularity. Typically in such a tournament monetary
prizes are awarded on the basis of a final ranking on all participants as determined by the results (in
fiches or points) during several tournament rounds (knock-out or not) of the cash version of Texas
6
Hold’em. As argued in [2], almost any casino cash game can be organized in the form of tournament
with this casino game played in its various tournament rounds and moreover, that it is not clear a
priori if a tournament should be automatically legally classified in the same way as its underlying cash
variant. In particular, for the case of Texas Hold’em tournaments, the statement “since in each round
one plays a game of skill, the tournament itself is a game of skill too”, cannot be justified at all
without an explicit analysis of relative skill. The final qualification of a tournament will heavily
depend on the number of participants in relation to the number of tournament rounds (relatively
short knock-out systems and lengthy full competitions will differ in relative skill) and especially on the
exact prize scheme (relative skill will substantially differ in a prize scheme with one winner from a
prize scheme with a relatively flat prize scheme for e.g. the top 10%). In this sense the analysis of
tournaments has definite similarities with the study on management games as presented in [16] and
[17].
In [2] a detailed, tailor-made analysis has been made for concrete set-ups for Texas Hold’em
tournaments. From this analysis the general conclusion can be drawn that the level of relative skill of
Texas Hold’em tournaments as organized in the Netherlands varies between 0.3 and 0.5, depending
on the exact tournament specifications.
5. Conclusion
The mathematical analysis of realistic variants of both cash versions and tournament versions of
Texas Hold’em indicate levels of relative skill that well exceed the skill threshold level. It confirms
earlier more qualitative findings that Texas Hold’em in all its practical variants should be considered
as games of skill, based on a consistent application of the methodology of relative skill.
References
[1] van der Genugten, B. and P. Borm (2012). The skill of Texas Hold’em cash games: a mathematical
analysis. Available at
http://www.tilburguniversity.edu/webwijs/show/?uid=p.e.m.borm&uid=p.e.m.borm
[2] Borm. P. and B. van der Genugten (2009). Cash and tournament poker: games of skill. Report,
Department of Econometrics and OR, Tilburg University, Tilburg, The Netherlands (In Dutch).
Available at http://www.tilburguniversity.edu/webwijs/show/?uid=p.e.m.borm&uid=p.e.m.borm
[3] http://www.rechtspraak.nl/Pages/default.aspx (2011). Ruling about organizing poker
tournaments, national jurisprudence number LJN: BN 0013, Rechtbank ‘s Gravenhage, 09/867520-08.
[4] Suzuki, S. and P. Cizmar (1998). Poker tournament strategies. Two plus Two Publishing,
Henderson.
[5] Sklansky, D. (2004). The theory of poker. Two plus Two Publishing, Henderson, 6th printing.
[6] Sklansky, D. (2002). Tournament poker for the advanced player. Two plus Two Publishing,
Henderson.
7
[7] Graig, M. (2007). The full tilt poker strategy guide: tournament edition. Grand Central Publishing,
New York.
[8] Levitt, S.D. and T.J. Miles (2011). The role of skill versus luck in poker: evidence from the world
series of poker. Report, Department of Economics, University of Chicago, USA.
[9] Levitt, S.D., Miles, T.J., and A.W. Rosenfield (2012). Is Texas Hold’em a game of chance? A legal
and economic analysis. Report, Department of Economics, University of Chicago, USA.
[10] Some websites on how to play poker tournaments: http://www.mosesbet.com/mtt-strategy,
http://www.playwinningpoker.com/poker/tournaments, http://www.tightpoker.com/tournaments,
http://www.pokertournamentstrategy.org/
[11] Maaten, R., Borm, P., van der Genugten, B., and R. Hendrickx (2010). The relative skill of Texas
Hold ‘Em: an experiment in cooperation with VARA Nieuwslicht. Report, Department of Econometrics
and OR, Tilburg University, Tilburg, The Netherlands (In Dutch). Available at
http://www.tilburguniversity.edu/webwijs/show/?uid=p.e.m.borm&uid=p.e.m.borm
[12] Dreef, M., Borm, P., and B. van der Genugten (2004). A new relative skill measure for games with
chance elements. Managerial and Decision Economics, 25, 255-264.
[13] Hendrickx, R., Borm. P., and B. van der Genugten (2008). Measuring skill in more-person games
with applications to poker. CentER Discussion paper 2008-106, Tilburg University, Tilburg, The
Netherlands.
[14] van der Genugten, B., Das. M., and P. Borm (2001). Gambling skillfully in casinos. Academic
Service, Schoonhoven, The Netherlands, ISBN 90 395 1571 9 (In Dutch).
[15] http://www.rechtspraak.nl/Pages/default.aspx (2005). Ruling about organizing management
games, national jurisprudence number LJN: AS6590, Rechtbank Arnhem, 105364.
[16] van der Genugten, B., Borm, P., and M. Dreef (2004). The application of the Dutch gaming act on
the management games Competition manager and Grand Prix manager. Report, Department of
Econometrics and OR, Tilburg University, Tilburg, The Netherlands (In Dutch).
[17] van der Genugten, B., Borm, P., and M. Dreef (2005). The application of the Dutch Gaming Act
on the management games Competition Manager and Grand Prix manager (continued). Report,
CentER of Applied Research, Tilburg University, Tilburg, The Netherlands (In Dutch).
[18] van der Genugten, B. (1993). Blackjack in Holland Casinos: how to beat the dealer. Tilburg
University Press, Tilburg, The Netherlands. ISBN 90 361 9793 7.
[19] van der Genugten, B., Das, M., and P. Borm (1999). An analysis of Blackjack variants. Report,
CentER of Applied Research, Tilburg University, Tilburg, The Netherlands, by the direction of Moulin
Rouge Casino- und Gastronomiebetriebsgesellschaft M.B.H., Vienna (In Dutch).
[20] Borm, P. and B. van der Genugten (2005). Poker: a game of skill! Report, Department of
Econometrics and OR, Tilburg University, Tilburg, The Netherlands, by the direction of Concord Card
Casino, Vienna (In German).
8
THE SKILL OF TEXAS HOLD’EM CASH GAMES: A MATHEMATICAL
ANALYSIS
1. INTRODUCTION
In this paper we give the specific details about THM-cashgames, showing that their skill is larger than
the juridical threshold, somewhere between 0.2 and 0.3. The general methodology about relative
skill and the result of its application to THM is described in [1]. This paper gives a precise description
of the calculations as far as it concerns cash games. In the following we suppose that the rules of
THM are known. Therefore, for shortness, we refer hereafter to specific rules using only some
catchwords.
A THM-cashgame consists of 4 rounds: preflop, flop, turn, river (MaxRounds = 4). Betting starts with
a fixed Small Blind (SB) and Big Blind (BB). The real decisions of players can be Fold, Call, Check, Bet
and Raise, dependent on the stage of the game. The maximum number of raises per round is 3
excluding the bet (MaxRaise = 3). The bet amount is doubled from the turn on (DoubleRound = 3). If
at a certain stage of the game the chips of player are insufficient for a certain decision he can go all-
in. The number of players varies: 2 <= Players <= 10. Betting rules can vary: fixed-, spread-, pot- and
no-limit betting. We refer to games with all these rules as full games.
Full games have very large game trees and are too difficult to analyze with a direct approach.
Therefore in section 1 we start with a simplified fixed-limit game with Players = 2, MaxRounds = 2
(preflop, river), DoubleRound = 2 and with MaxRaise = 1. We refer to this game as the basic game.
This basic game contains all the specific strategic possibilities of full games. It is extensively analyzed
in detail by varying beginner strategies, looking especially for cases with a low skill for comparison
with the juridical threshold.
Thereafter, in section 2 we consider fixed-limit generalizations by varying separately Players,
MaxRounds and MaxRaise towards usual values, again paying attention to cases with low skill. From
these results we extrapolate conclusions to cases where these parameters are varied simultaneously.
9
We extend the fixed-limit basic games of section 1 in section 3 to spread-limit, in section 4 to pot-
limit and in section 5 to no-limit. Again the emphasis is on variants with low skill. This leads to the
final conclusion that the skill of all full versions of cash-THM are larger than the juridical threshold.
THM is played with a simple card-deck of 52 cards. But this leads to a tremendous number of card
distributions over players and community cards. So the analysis cannot be based on this population
but, necessarily, on samples. This raises special questions about dispersion and bias (see section 1.2).
For the analysis of cash-THM a computer program is developed under Matlab. It covers all variations
mentioned above by just filling in the appropriate parameters (see Appendix A2). But this variety has
only theoretical value. In practice the resulting game trees and sample sizes are always restricted by
computer time and computer memory. For the obtained results in this paper the calculations were
done with the 64-bit version of Matlab under a Windows-7 PC ( 64-bit, i7-processor) with an internal
memory of 16 Gb.
1. THE BASIC GAME (SIMPLIFIED FIXED-LIMIT THM)
1.1 Rules and game tree
The global rules are described in the introduction. The game tree of figure 1.1.1. reveals all details:
- no all-in situations: the start capital (amount of chips) of the players is always large
enough),
- maximal 1 raise per round (exclusive the bet) means no reraise,
- the BB-player in the first round gets always has a real move (see node 3 in the figure);
his decision Bet is sometimes called Raise and is always counted as a raise,
- after the first round (preflop) the last round comes immediately (the river with 5 cards),
- the bet is doubled in the last round (DoubleRound = 2),
- 2 players (headsup): player I is the SB, player II the BB,
- as usual with headsup, player I starts the betting in the first round, player II in the last
round.
10
In the figure the decisions and the corresponding betting chips are indicated at the leaves, the
amounts of betting chips of the 2 players are indicated at the nodes. The nodes themselves are
numbered in such a way that from the top on a higher node is never followed by a lower node.
Furthermore, the numbers of a specific round are always greater than the numbers of a foregoing
round. The numbers of endnodes are put within squares, the intermediate ones within circles. At
each intermediate node the corresponding player and the pot composition is indicated; the leaves
show the decisions and the corresponding bets.
11
12
Figure 1.1.1 The game tree of the basic game.
We have included this figure because it makes immediately clear the specific rules of the basic game.
Appendix A3 contains all details as they are generated by the computer program.
1.2 Sampling from the population of card distributions
Even with 2 players and 2 rounds (the basic game) the population of card distributions is too large to
work with:
C(52,2) x C(50,2) x C(48,5) = 2 781 381 002 400 .
We have to take a representative sample and act as if this is the whole population. It should be
sufficiently large for the desired accuracy of the skill ( 0.02, say).
A special problem for calculations with respect to the optimal player is that the gain result has a
positive bias (towards the fictive player) for any sample. However, the bias for a random sample can
be reduced considerably by taking a representative sample in a special way.
In section 1.3 we give the results for various beginner strategies. In every subsection we use at first a
very elementary sample of size 4 and give the detailed calculations which can be controlled by hand.
Thereafter, we use a representative sample of size 3 288 600, sufficiently large for getting the desired
accuracy. Of course results follow now from the computer program.
For sake of comparison with different rules these elementary and representative samples are
repeatedly used whenever the number of rounds and the number of players is 2. The samples
themselves are described in the sections 1.2.1 and 1.2.2.
1.2.1 Elementary sample
The elementary sample is shown in table 1.2.1.1. Cards are numered in the following way:
1 = 2c(lubs), 2 = 2d(iamonds), 3 = 2h(earts) , 4 = 2s(pades), . . . , 52 = A(spades).
CardComb Preflop I Preflop II River
1
2
1 8
1 8
42 46
42 46
26 36 43 48 49
3 4 5 33 38
13
3
4
42 46
42 46
1 8
1 8
26 36 43 48 49
3 4 5 33 38
Table 1.2.1.1 Card numbers of the elementary sample of 4 card combinations.
So in our analysis we act as if this set forms the whole population of card distributions. We do this
just for illustrating the concepts with detailed calculations. It is even not a random sample but
contrarily carefully chosen. We see that for each player there are only 2 preflops and also only 2
rivers. This leads to 4 carefully mixed card combinations. This specific small sample makes results not
completely trivial. Note that the knowledge of the optimal player is almost the same as that of the
fictive player. The difference is only in the first round. Player I with Preflop = [1 8 ] knows already that
Player II has Preflop [ 42 46 ] but he is only uncertain about which of the 2 possible sets of the River
applies. For him the probabilities are equal. The same holds for player I with Preflop [ 42 46]. Similar
considerations apply to Player II. If we would have chosen a random sample then, with a very high
probability, there would not have been such a pattern at all: the optimal player has exactly the same
information as the fictive player, making detailed analysis worthless.
The following table 1.2.1.2 gives the card names and the coded poker values at the river.The coded
poker value contains 5 numbers: nr 1 gives the main value (1-9) in decreasing value with
9 = SF (straight flush), 8 = 4K (four of a kind), 7 = FH (full house), 6 = F (flush),
5 = S (straight), 4 = 3K (three of a kind), 3 = 2P (two pairs), 2 = 1P (one pair).
1 = 0P ( no pair = high card),
and nrs 2-6 the other meaningful values in decreasing importance (0=no meaning). So, in the first
card combination at a possible showdown player I has a high card with kicker 14 = Ace and player II
has two pairs 13 = King, 12 = Queen with kicker 14 = Ace.
CardComb Player Preflop River Poker value
1
2
3
4
I
II
I
II
I
II
I
II
2c 3s
Qd Kd
2c 3s
Qd Kd
Qd Kd
2c 3s
Qd Kd
2c 3s
8d Ts Qh Ks Ac
2h 2s 3c Tc Jd
8d Ts Qh Ks Ac
2h 2s 3c Tc Jd
1 14 13 12 10 8
3 13 12 14 0 0
7 2 3 0 0 0
2 2 13 12 11 0
3 13 12 14 0 0
1 14 13 12 10 8
2 2 13 12 11 0
7 2 3 0 0 0
Table 1.2.1.2 Corresponding card names and poker values for table 1.2.1.
1.2.2 Representative sample
14
The discussion of the properties of the elementary sample in section 1.2.1 makes clear that we will
not use a random sample. The results for the beginner and fictive player are always unbiased but the
sample must be chosen large for a small standard deviation. The result for the optimal player is
always positively biased towards that of the fictive player. We need a (too) large random sample to
be sure that the optimal player has not much more knowledge compared with the whole population.
It is much better to choose another type of still a representative sample in order to reduce the bias.
The basic idea is already contained in the elementary sample: let the chosen card combinations be
such that for a specific player a preflop has a lot of other preflops of his opponents and also a lot of
rivers. Of course, the selection algoritm must be such that theoretically we get the whole population
by taking the sample sufficiently large.
In table 1.2.2.1 we give an example of the result of a construction of a representative sample in a
specific way.
At first we draw from the deck of 52 cards randomly SimRiver = 2 sets of 5 cards for the River (the
maximum would be MaxSimRiver = C(52,5) = 2 598 960). For each River there remain 47 cards in the
deck. Choose with such a River randomly SimCards = 4 (the minimum, the maximum would be
MaxSimCards = 47). From those 4 cards form all possible SimPlayers = 6 Preflop-combinations of
both players (MaxSimPlayers = C(47,2) x C(45,2) = 1 070 190). So the total numer of card
combinations is TotCombs = 2 x 6 = 12 (MaxTotCombs = 2 781 381 002 400). The final step for the
list is to randomize the order (not shown here).
15
CardComb Preflop I Preflop II River
1
2
3
4
5
6
7
8
9
10
11
12
30 35
30 38
30 43
35 38
35 43
38 43
9 23
9 27
9 46
23 27
46 9
27 46
38 43
35 43
35 38
30 43
30 38
30 35
27 46
23 46
23 27
9 46
9 27
9 23
1 3 10 47 48
1 3 10 47 48
1 3 10 47 48
1 3 10 47 48
1 3 10 47 48
1 3 10 47 48
21 24 36 38 45
21 24 36 38 45
21 24 36 38 45
21 24 36 38 45
21 24 36 38 45
21 24 36 38 45
Table 1.2.2.1 CardCombs with Players =2, MaxRounds =2, SimCards =4, SimRiver =2.
Of course the sample of table 1.2.2.1 is much to small. For moderate accuracy we take SimRiver = 20,
SimCards = 30 (=> SimPlayers = 164 430).This gives TotCombs =
3 288 600. A sketch of such a representative sample is given in table 1.2.2.2. We use this sample for
the whole analysis of games with MaxRounds = 2 and Players = 2. For this sake the sample has been
saved in a file.
CardComb Preflop I Preflop II River
1
2
..
3288600
37 50
7 46
.. ..
6 37
16 38
15 33
.. ..
8 26
10 11 24 29 36
10 11 24 29 36
.. .. .. .. ..
15 19 21 31 49
Table 1.2.2.2 Card numbers of representative sample of 3 288 600 card combinations.
The same procedure for getting representative samples can be followed in other cases. For example,
for a game with 3 rounds (Preflop, Flop, River) and 2 players the table 1.2.2.3 gives a (much too
small) representative sample with MaxRounds = 2, Players = 2, SimFlop = 2 and SimRiver = 2
(combined with each flop). Again we show the list without randomization.
CardComb Preflop I Preflop II Flop River
1
2
3
4
5
6
7
8
9
10
11
12
13
14
25 33 51 52 5 10 46 12 24
25 51 33 52 5 10 46 12 24
25 52 33 51 5 10 46 12 24
33 51 25 52 5 10 46 12 24
33 52 25 51 5 10 46 12 24
51 52 25 33 5 10 46 12 24
2 12 39 50 5 10 46 18 37
2 39 12 50 5 10 46 18 37
2 50 12 39 5 10 46 18 37
12 39 2 50 5 10 46 18 37
12 50 2 39 5 10 46 18 37
39 50 2 12 5 10 46 18 37
11 24 44 46 21 38 39 2 33
11 44 24 46 21 38 39 2 33
16
15
16
17
18
19
20
21
22
23
24
11 46 24 44 21 38 39 2 33
24 44 11 46 21 38 39 2 33
24 46 11 44 21 38 39 2 33
44 46 11 24 21 38 39 2 33
1 5 28 31 21 38 39 25 52
1 28 5 31 21 38 39 25 52
1 31 5 28 21 38 39 25 52
5 28 1 31 21 38 39 25 52
5 31 1 28 21 38 39 25 52
28 31 1 5 21 38 39 25 52
Table 1.2.2.3 CardCombs with MaxRounds= 3, SimCards=4, SimRiver=2, SimFlop =2.
1.3 Skill
For determining the (relative) skill we only have to specify te beginner strategies. From those the
optimal and fictive strategies follow, leading to the skill.
We assume that a beginner will base his decision on direct available card information while
comparing the pot odds with the probability odds (see [1]). Otherwise stated, at any stage of the
game he is able to calculate his expected gain based on his own cards and the available community
cards. In fact, this is a really ‘advanced’ beginner because that is only possible after studying some
poker literature. This seems to be the only way to get a simple formulation of beginner strategies
suitable for all poker variants. Of course, this will lead to an underestimate of skill. This fits in the
policy to search for poker variants with low skill. However, this should be kept in mind when judging
specific skill values.
We consider 4 types of beginner strategies: take the decision with
- maximal expected gain (section 1.3.1), a rather passive strategy,
- randomized decisions as variation (section 1.3.2) leading to bluffing and sandbagging,
- implied pot odds as variation (section 1.3.3),
- maximal bet under those decisions with positive max.exp.gain (section 1.3.4), a rather
aggressive strategy.
We conclude this section with a discussion about the accuracy of the results (section 1.3.5).
1.3.1 Maximal expected gain
17
1.3.1.1 Calculations for the elementary sample
Table 1.3.1.1 gives for both players the win- and equal probabilities at a possible showdown for the
card combinations of table 1.2.1.1 (or table 1.2.1.2):
CardComb Player Preflop
Win Equal
River
Win Equal
1
2
3
4
I
II
I
II
I
II
I
II
0.2924 0.0613
0.6241 0.0198
0.2924 0.0613
0.6241 0.0198
0.6241 0.0198
0.2924 0.0613
0.6241 0.0198
0.2924 0.0613
0 0.2030
0.7828 0.0040
0.9848 0.0020
0.3879 0.0091
0.7828 0.0040
0 0.2030
0.7828 0.0040
0.9848 0.0020
Table 1.3.1.1 Win-and equal probabilities for the players for table 1.2.1.1.
These probabilities have been obtained by enumarating all possible card combinations. They have
been saved in files and form the basis for the calculation of the beginner strategies.
Beginners:
We assume that beginners base their decisions in any decision node on the expected gain given the
pot size, their own preflop and the common river cards in the second round. In poker terminoligy:
they base their decisions by comparing the pot odds with the (probability) odds.
In this section we assume that they take always the decision with maximal expected gain, supposing
that they can actually calculate the desired probabilities accurately enough.We give the calculation
details for the beginner decisions for CardComb=1 in table 1.3.1.2. Its contents will be clear with
figure 1.1.1. E.g., for Node =1 the pot is [2 4] (PotVec with the SB and BB)) with sum 6. Player I is the
current player CurPlayer who has to decide between 3 decisions leading to the next nodes 2, 3, 4
(NextNodes) by Fold with bet 0, Call with bet 2 and Raise with bet 6. The corresponding expected
gains (ExpGainVec) can be calculated from this. E.g., for the decision Call we get the expectation (see
table 1.3.1.1):
[0.2924 + 0.0613 / 2) x (6 + 2)] – 2 = 0.5843.
The gain for Fold is 0 and the gain for Raise is -2.1236. So Call is the decision with the maximal
expected gain. This is represented as the randomized decision [0 1 0] (THM0DecsCell).
18
Node = 1 PotVec: [2 4] CurPlayer: 1
NextNodes: {[2 3 4]} (F=0 Ca=2 R=6)
ExpGainVec = 0 0.5843 -2.1236
THM0DecsCell{Node} = 0 1 0
Node = 3 PotVec: [4 4] CurPlayer: 2
NextNodes: {[7 8]} (Ch=0 B=4)
ExpGainVec = 5.0720 3.6080
THM0DecsCell{Node} = 1 0
Node = 4 PotVec: [8 4] CurPlayer: 2
NextNodes: {[5 6]} (F=0 Ca=4)
ExpGainVec = 0 6.1441
THM0DecsCell{Node} = 0 1
Node = 8 PotVec: [4 8] CurPlayer: 1
NextNodes: {[9 10]} (F=0 Ca=2)
ExpGainVec = 0 1.1685
THM0DecsCell{Node} = 0 1
Node = 11 PotVec: [8 8] CurPlayer: 2
NextNodes: {[12 13]} (Ch=0 B=8)
ExpGainVec = 12.5576 10.8364
THM0DecsCell{Node} = 1 0
Node = 12 PotVec: [8 8] CurPlayer: 1
NextNodes: {[14 15]} (Ch=0 B=8)
ExpGainVec = 1.6242 -5.5636
THM0DecsCell{Node} = 1 0
Node = 13 PotVec: [8 16] CurPlayer: 1
NextNodes: {[16 17 18]} (F=0 Ca=8 R=16)
ExpGainVec = 0 -4.7515 -11.9394
THM0DecsCell{Node} =1 0 0
Node = 15 PotVec: [16 8] CurPlayer: 2
NextNodes: {[21 22 23]} (F=0 Ca=8 R=16)
ExpGainVec = 0 17.1152 15.3939
THM0DecsCell{Node} = 0 1 0
Node = 18 PotVec: [24 16] CurPlayer:2
NextNodes: {[19 20]} (F=0 Ca=8)
ExpGainVec = 0 29.6727
THM0DecsCell{Node} = 0 1
Node = 23 PotVec: [16 24] CurPlayer: 1
NextNodes: {[24 25]} (F=0 Ca=8)
ExpGainVec = 0 -3.1273
THM0DecsCell{Node} = 1 0
Node = 26 PotVec: [4 4] CurPlayer: 2
NextNodes: {[27 28]} (Ch=0 B=8)
ExpGainVec = 6.2788 4.5576
THM0DecsCell{Node} = 1 0
Node = 27 PotVec: [4 4] CurPlayer: 1
NextNodes: {[29 30]} (Ch=0 B=8)
ExpGainVec = 0.8121 -6.3758
THM0DecsCell{Node} = 1 0
Node = 28 PotVec: [4 12] CurPlayer: 1
NextNodes: {[31 32 33]} (F=0 Ca=8 R=16)
ExpGainVec = 0 -5.5636 -12.7515
THM0DecsCell{Node} = 1 0 0
Node = 30 PotVec: [12 4] CurPlayer: 2
NextNodes: {[36 37 38]} (F=0 Ca=8 R=16)
ExpGainVec = 0 10.8364 9.1152
THM0DecsCell{Node} =0 1 0
Node = 33 PotVec: [20 12] CurPlayer: 2
NextNodes: {[34 35]} (F=0 Ca=8)
ExpGainVec = 0 23.3939
THM0DecsCell{Node} = 0 1
Node = 38 PotVec: [12 20] CurPlayer: 1
NextNodes: {[39 40]} (F=0 Ca=8)
ExpGainVec = 0 -3.9394
THM0DecsCell{Node} = 1 0
Node = 41 PotVec: [8 8] CurPlayer: 2
NextNodes: {[42 43]} ( Ch=0 B=8)
ExpGainVec = 12.5576 10.8364
THM0DecsCell{Node} = 1 0
Node = 42 PotVec: [8 8] CurPlayer: 1
NextNodes: {[44 45]} (Ch=0 B=8)
ExpGainVec = 1.6242 -5.5636
THM0DecsCell{Node} = 1 0
Node = 43 PotVec: [8 16] CurPlayer: 1
NextNodes: {[46 47 48]} (F=0 Ca=8 R=16)
ExpGainVec = 0 -4.7515 -11.9394
THM0DecsCell{Node} = 1 0 0
Node = 45 PotVec: [16 8] CurPlayer: 2
NextNodes: {[51 52 53]} (F=0 Ca=8 R=16)
ExpGainVec = 0 17.1152 15.3939
THM0DecsCell{Node} = 0 1 0
Node = 48 PotVec: [24 16] CurPlayer: 2
NextNodes: {[49 50]} (F=0 Ca=8)
ExpGainVec = 0 29.6727
THM0DecsCell{Node} = 0 1
Node = 53 PotVec: [16 24] CurPlayer: 1
NextNodes: {[54 55]} (F=0 Ca=8)
ExpGainVec = 0 -3.1273
THM0DecsCell{Node} = 1 0
Table 1.3.1.2 Calculation of beginners decisions for CardComb = 1
The calculations for CardComb = 2, 3, 4 are the same: see table 1.3.1.3. Ordered by the player nodes,
this leads to table 1.3.1.3.
19
Player I
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
1
8
12
13
23
27
28
38
42
43
53
0 1 0
0 1
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
1 0 0
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
0 1 0
0 1
1 0
1 0 0
0 1
1 0
1 0 0
0 1
1 0
1 0 0
0 1
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
Player II
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
3
4
11
15
18
26
30
33
41
45
48
1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
1 0
0 1 0
0 1
Table 1.3.1.3 Beginner strategies in randomized form for maximal exp. gain.
Note that for nodes in round = 1 the decisions for CardComb 1 and 2 have to be the same since the
preflop is the same. The same holds for CardComb 3 and 4.
Substitution of the beginner strategies of table 1.3.1.3 leads to the result of table 1.3.1.4:
CardComb EndNode Result Player I
Gain Bet
Player II
Gain Bet
1
2
3
4
29
29
29
29
II wins SD
I wins SD
I wins SD
II wins SD
-4 4
4 4
4 4
-4 4
4 4
-4 4
-4 4
4 4
Sum 0 16 0 16
Exp. 0 4 0 4
Table 1.3.1.4 Gains and total bets for all beginners.
In fact this result is rather trivial for decisions based on maximal expected gain. Let p be the
probability of winning the showdown, S the pot size before the decision and B the bet amount.Then
Check is always better than Bet since
E(Check) = pS > E(Bet)= p(S + B) – B.
Furthermore, let C be the call amount and R the raise amount (R > C). Then Call is always better than
Raise since
20
E(Call) = p (S + C) – C > E(Raise)= p(S + R) – R.
Finally, at node 1 Call is always better than Fold regardless of the cards. The minimal win probability
for a preflop is p = 0.2924 (for [2c 3d] ). So in this case
E(Call) = p (6 +2 ) – 2 = 0.3392 > E(Fold) = 0.
Therefore any game with beginners develops as 1 => 3=> 7 => 26 => 27 => 29.
Fictive players:
An (optimal) fictive player knows the cards of his opponent and the river cards before the betting
starts. Together with the knowlegde of the beginner strategy of his opponent the maximal gain can
be calculated for each CardComb. The first step is table 1.3.1.5, also needed for the optimal player.
We comment the result for CardComb = 1 given in table 1.3.1.5. At first consider the fictive or
optimal player I against the beginner II. Place arrows in figure 1.1.1 in the leaves of the decisions of
beginner II according to table 1.3.1.3. This determines 5 possible endnodes of the fictive or optimal
player I (EndNodeVec). From this and table 1.2.1.1 for the showdown the 5 corresponding gains and
bets follow (GainEndVec and BetEndVec). The results for the fictive or optimal player II follows in the
same way. For other CardCombs and tables of the beginner the results follow also in the same way.
CardComb OptFictPlayer I OptFictPlayer II
1 EndNodeVec
GainEndVec
BetEndVec
2 14 22 29 37
-2 -8 -16 -4 -12
2 8 16 4 12
29 31 44 46
4 4 8 8
4 12 8 16
2 EndNodeVec
GainEndVec
BetEndVec
2 14 22 29 37
-2 8 16 4 12
2 8 16 4 12
29 32 44 47
-4 -12 -8 -16
4 12 8 16
3 EndNodeVec
GainEndVec
BetEndVec
2 14 21 29 36
-2 8 8 4 4
2 8 16 4 12
29 32 44 47
-4 -12 -8 -16
4 12 8 16
4 EndNodeVec
GainEndVec
BetEndVec
2 14 22 29 37
-2 -8 -16 -4 -12
2 8 16 4 12
29 32 44 47
4 12 8 16
4 12 8 16
Table 1.3.1.5 Basic calculations for optimal and fictive players.
Now consider the fictive players. For CardComb = 1 and FictPlayer = I take the maximal gain (FictGain
= -2) with the corresponding Bet (FictBet = 2). This gives the first results in Table 1.3.1.6. For the
other CardCombs and FictPlayer = II the results follow in the same way.
CardComb OptFictPlayer I
Gain Bet
OptFictPlayer II
Gain Bet
1
2
3
4
-2 2
16 16
8 8
-2 2
8 8
-4 4
-4 4
16 16
21
Sum 20 28 16 32
Exp. 5 7 4 8
Table 1.3.1.6 Gains and bets for the fictive players.
Optimal players:
The set of all card distributions is the elementary sample of table 1.2.1.1. This small sample makes
that the knowledge of the optimal player is almost the same as that of the fictive player. The
difference is only in the first round. Player I with Preflop = [1 8 ] knows already that Player II has
Preflop [ 42 46 ] but he is only uncertain about which of the 2 possible sets of the River applies. For
him the probabilities are equal.
The same holds for player I with Preflop [ 42 46]. Similar considerations apply to Player II. We
describe in detail the calculation of the expected gains and bet for the optimal Player I against the
beginner II. The explained framework is also suitable for more complex cases.
Calculations are backwards in the game tree. We start with the endnode results for the optimal and
fictive player I as given in the table 1.3.1.5 and do the calculations backwards. It follows
automatically that betting rounds are also taken backwards. In a specific round we take the maximal
expected gain over the possible decisions for the optimal player I; for the other beginners we fill in
their decisions given in table 1.3.1.4. If we go back one round then we have to sum up the results
with the same information.
The following table with subsequent indices for different card combinations based on table 1.2.1.1 is
helpful. So we see for player I that for CardComb = 1 and = 2 the indeces are the same for the Preflop
(OptInx= 1) but different for the River.
Player I Player II
CardComb Preflop River
1 1 1
2 1 2
3 2 3
4 2 4
CardComb Preflop River
1 1 1
2 1 2
3 2 3
4 2 4
Table 1.3.1.7 The indices OptInx for CardCombs of the players.
Table 1.3.1.8 gives for the optimal player I the backward calculations. Substitute the beginner
strategy of player II in the game tree. Nodes which cannot be reached get a bet equal to 0. The first
node that can be reached is node 30 of player II. Calculate the expected gains (OptGain) and bets
(OptBet) for all different card combinations (OptInx). For this pure strategy this means simply
copying.
The next node is 27 of the optimal player I. Take the maximal expected gain over the decisions for all
different card combinations (OptGain) and denote the corresponding bet (OptBet) and the number
22
of the decision (OptDec). For equal expected gains we take the one with the lowest bet. Since
optimal strategies can always be taken pure a randomized form is not necessary.
So we continue up till the last node 11 for which round = 2 (River).
23
LastNode = 55, MaxRInx = 4
Round = 2 Node = 30 (Player = 2)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet
1
2
3
4
0 -12 0
0 12 0
4 0 0
0 -12 0
-12
12
4
-12
0 12 0
0 12 0
12 0 0
0 12 0
12
12
12
12
Round = 2 Node = 27 (Player = 1)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet OptDec
1
2
3
4
-4 -12
4 12
4 4
-4 -12
-4
12
4
-4
4 12
4 12
4 12
4 12
12
12
12
12
1
2
1
1
Round = 2 Node = 26 (Player = 2)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet
1
2
3
4
-4 0
12 0
4 0
-4 0
-4
12
4
-4
4 0
12 0
4 0
4 0
4
4
4
4
Round = 2 Node = 15 (Player = 2)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet
1
2
3
4
0 -16 0
0 16 0
16 0 0
0 -16 0
-16
-16
16
-16
0 16 0
0 16 0
16 0 0
0 16 0
16
16
16
16
Round = 2 Node = 12 (Player = 1)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet OpDec
1
2
3
4
-8 -16
8 16
8 8
-8 -16
-8
16
8
-8
8 16
8 16
8 16
8 16
8
16
8
8
1
2
1
1
Round = 2 Node = 11 (Player = 2)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet
1
2
3
4
-8 0
16 0
8 0
-8 0
-8
16
8
-8
8 0
8 0
8 0
8 0
8
16
8
8
Table 1.3.1.8 The calculations for Round = 2 of optimal player I.
Before we get to round 1 (Preflop) we have to sum up the results with the same information on the
base of table 1.3.1.7. Hence, for node 11 we sum the results for OptInx = 1 and 2 and put it in table
1.3.1.9 in node 6. The same has to be done for nodes 26 and 41.
LastNode = 10 (NextRound), MaxRInx = 2
OptInx 1 2 3 4 5 11
=>
6
26
=>
7
8 9 41
=>
10
1 Gain
Bet
0
0
-4
4
0
0
0
0
0
0
8
24
8
16
0
0
0
0
0
0
2 Gain
Bet
0
0
4
4
0
0
0
0
0
0
0
16
0
8
0
0
0
0
0
0
Table 1.3.1.9 The transformation of the results to Round = 1 of optimal player I
24
With table 1.3.1.9 we continue as before with round = 1. The result is shown in table 1.3.1.10. The
last step is always node 1.
25
Round = 1 Node = 4 (Player = 2)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet
1
2
0 8
0 16
8
16
0 24
0 16
24
16
Round = 1 Node = 3 (Player = 2)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet
1
2
8 0
0 0
8
0
16 0
8 0
16
8
Round = 1 Node = 1 (Player = 1)
OptInx NextOptGainVec OptGain NextOptBetVec OptBet OpDec
1
2
-4 8 8
-4 0 0
8
0
4 16 24
4 8 16
16
8
2
2
Sum 8 24
Exp. 2 6
Table 1.3.1.10 The calculations for Round = 1 of optimal player I.
Summation in tabel 1.3.1.10 gives SumOptGain = 8 and SumOptBet = 24. Since all 4 CardCombs have
equal probability this leads for the optimal Player I to OptGain = 8 / 4 = 2 and OptBet = 24 / 4 = 6.
For the optimal player II the calculations are the same.
Table 1.3.1.11 gives the (pure) optimal optimal strategies of both players. Of course, for nodes that
cannot be reached due to the beginner strategy of the opponent, the choice is arbitrary and has been
filled with 1 for the smallest bet. For comparison we have included the beginner strategies of table
1.3.1.3 as well.
Player I
Node OptInx = 1 OptInx = 2 OptInx = 3 OptInx = 4
Beginner Opt Beginner Opt Beginner Opt Beginner Opt
1
8
12
13
23
27
28
38
42
43
53
0 1 0 2
0 1 1
1 0 1
1 0 0 1
1 0 1
1 0 1
1 0 0 1
1 0 1
1 0 1
1 0 0 1
1 0 1
0 1 0 2
0 1 1
1 0 2
0 1 0 1
0 1 1
1 0 2
0 1 0 1
0 1 1
1 0 1
0 1 0 1
0 1 1
1 0 1
1 0 0 1
0 1 1
1 0 1
1 0 0 1
0 1 1
1 0 1
1 0 0 1
0 1 1
1 0 1
0 1 0 1
0 1 1
1 0 1
0 1 0 1
0 1 1
1 0 1
0 1 0 1
0 1 1
Player II
Node OptInx = 1 OptInx = 2 OptInx = 3 OptInx = 4
Beginner Opt Beginner Opt Beginner Opt Beginner Opt
3
4
11
15
18
26
30
1 0 1
0 1 1
1 0 1
0 1 0 1
0 1 1
1 0 1
0 1 0 1
1 0 1
0 1 1
1 0 1
0 1 0 1
0 1 1
1 0 1
0 1 0 1
1 0 1
1 0 0 1
1 0 1
1 0 1
1 0 0 1
1 0 1
0 1 0 1
0 1 1
1 0 2
0 1 0 1
26
33
41
45
48
0 1 1
1 0 1
0 1 0 1
0 1 1
0 1 1
1 0 1
0 1 0 1
0 1 1
1 0 1
1 0 1
1 0 0 1
1 0 1
0 1 1
1 0 2
0 1 0 1
0 1 1
Table 1.3.1.11 Beginner and optimal strategies for maximal exp. gain
The next table 1.3.1.12 gives an overview of the gains and bets for both players for the different
types beginner, optimal and fictive player.
Gain
I II
Bet
I II
Beginner 0 0 4 4
Optimal 2 2 6 6
Fictive 5 4 7 8
Table 1.3.1.12 The expected gains for all types of players.
Skill:
If we measure the game result simply as (expected) gain, then by taking the mean (or sum) of the
gains over the 2 players we get from table 1.3.1.11:
LE = 2, RE = 2 ½ => S = 4/9 = 0.4444.
This is the way the calculation of skill is explained in [1].
Of course there are other possibilities. We make a distiction in notation. Since the foregoing measure
is based on the gain and the sum of both players we use the notation Sgs ( = S). We can also take the
mean of the skill based on gain of the two players, leading to Sgm. Another possibility is to consider
the return rate as the game result. This leads to measures Srs and Srm. Table 1.3.1.13 gives an
overview of the calculations based on table 1.3.1.12.
EGainArr EBetArr Rarr=
I II gmean | I II mean rmean | I II
0 0 0 | 4 4 4 | 0 | 0 0
2 2 2 | 6 6 6 | 2/6 | 1/3 1/3
5 4 9/2 | 7 8 15/2 | 9/15 | 5/7 1/2
--- --- ---- ----- ---- ----
Sm 2/5 1/2 Sgs Srs | Sr 7/15 2/3
--------- -----------
Sgm Srm
=> Sgs Srs Sgm Srm
4/9 5/9 9/20 17/30
0.4444 0.5556 0.4500 0.5667
27
Table 1.3.1.13 Different forms of skill measures for maximal expectations.
If the order of magnitude of the bets is more or less the same we think that the measure Sgs is the
most simple one, and therefore the most appropriate. The 4 measures can differ substantially only
for special constructed theoretical games. In practical examples the difference is always small. For
signalyzing irregularities we compute always the 4 values and refer to S = Sgs as the relative skill.
1.3.1.2 Results for the representative sample
Table 1.3.1.14 gives the results for the representative sample (table 1.2.2.2). It has been put in a
general form, suitable for all variants to come later on.
At first the game parameters are shown. We take always SB = 2 and BB = 4. Pot-limit and no-limit are
disabled (values 0): see sections 7 and 8, respectively. BetVec = 4 means that the only size of the bet
is 4, the usual value for fixed-limit: see section 6 for spread-limit. DoubleRound = 2 is the first round
in which all bets are doubled. The starting chips of the players (ChipVec = [100 100]) have been
chosen in such a way that they are sufficient for both players and any decision in the game.
Then the beginner parameters follow. GainLabel = 1 means maximal expected gains (this section),
the meaning of the other value GainLabel = 0 is explained in section 1.3.4. Implied odds are disabled
for both players (FactDecVec = [ 0 0 ]): see section 1.3.2. Randomization is also disabled for both
players (RandProbDecVec = [0 0 ] ): see section 1.3.3.
The game parameters together with the beginner parameters generate the gametree: TotNodes = 55
(see figure 1.1.1 or appendix A3).
Thereafter the information about sampling follows. The data come from the file as sketched in table
1.2.2.2. The meaning of the sim-parameters is discussed already there.
The bias-test gives insight in the bias of the results of the optimal players. Their gains are calculated
for 10 intermediate sample sizes. So, based on the first part TotCombVec = 328 860 (of the whole
sample size of TotCards = 3 288 600) the gains of the optimal players are [1.6218 1.5937] (under
MGainArr) and the corresponding bets [8.9975 9.1165] (under MbetArr). We see that the gains
decrease with increasing sample sizes, indicating a decreasing positive bias. At the end this bias has
almost vanished; fluctuations are only due to dispersion (see section 1.3.4).
Then the gains and bets of all players are given and the skill values follow. In particular, S = Sgs =
0.3105, larger than the upper juridical threshold 0.3.
Finally, the computer time for the calculations is given: 37228.322697 seconds = 0.4309 days.
28
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0 0
TotNodes = 55
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
P1 P2 P1 P2
328860 1.6218 1.5937 8.9975 9.1165
657720 1.6088 1.5634 9.0747 9.1160
986580 1.5923 1.5620 9.0565 9.0893
1315440 1.5859 1.5581 9.0269 9.1284
1644300 1.5855 1.5517 9.0426 9.1047
1973160 1.5842 1.5481 9.0532 9.0591
2302020 1.5811 1.5495 9.0443 9.0937
2630880 1.5798 1.5492 9.0695 9.0999
2959740 1.5739 1.5525 9.0721 9.1040
3288600 1.5715 1.5527 9.0735 9.1016
ExpGainArr:
P1 P2
beg 0 0
opt 1.5715 1.5527
fict 5.4615 4.5999
ExpBetArr:
beg 4.0285 4.0285
opt 9.0735 9.1016
fict 7.9411 8.7314
Sgs,Srs,Sgm,Srm:
0.3105 0.2848 0.3126 0.2878
Elapsed time is 37228.322697 seconds = 0.4309 days.
Table 1.3.1.14 Results for maximal expectations (MaxRaise = 1).
1.3.2 Randomized decisions
29
The beginner strategies of table 1.3.1.3 can be modified from pure to mixed strategies. In poker
terminology: we incorperate sandbagging and bluffing. The main decision gets probability 0.9 and the
remaining probability 0.1 is equally distributed among the other decicions as far as they differ from
the decision Fold. We take these values for both players (RandProbDecVec = [0.1 0.1]. This leads to
the modified beginner strategies of table 1.3.2.1.
Player I
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
1
8
12
13
23
27
28
38
42
43
53
0 0.9 0.1
0 1
0.9 0.1
0.9 0.05 0.05
0.9 0.1
0.9 0.1
0.9 0.05 0.05
0.9 0.1
0.9 0.1
0.9 0.05 0.05
0.9 0.1
0 0.9 0.1
0 1
1 0
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0 0.9 0.1
0 1
1 0
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0 0.9 0.1
0 1
1 0
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
Player II
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
3
4
11
15
18
26
30
33
41
45
48
0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 1
0.9 0.1
0.9 0.05 0.05
0.9 0.1
0.9 0.1
0.9 0.05 0.05
0.9 0.1
0.9 0.1
0.9 0.05 0.05
0.9 0.1
0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
0.9 0.1
0 0.9 0.1
0 1
Table 1.3.2.1 Beginner strategies for randomization.
In case of randomized decisions for beginners the calculation of the gain for the fictive player can be
quite cumbersome, even if we do this in the most appropriate way.
We give as example the calculation of the expected gain for CardComb = 1 and Fictive Player = 2. At
first we sketch why things become more complicated than for pure strategies (see table 1.3.1.5 and
1.3.1.6) . As the first step we take the tree of figure 1.1.1 and fill in at the leaves the (conditional)
probabilities of the randomized strategy of the beginner. For each leave of the fictive player we
substitute the (conditional) probability 1, indicating that this node can be chosen by him. From that
we calculate the absolute probabilities for the endnodes. In the table 1.3.2.2 we restrict ourselves to
endnodes with positive probability.
Node Prob Gain Bet Node Prob Gain Bet
14 0.09 8 8 36 0.09 -4 4
30
16 0.09 8 16
17 0.005 16 16
19 0.005 -16 16
20 0.005 24 24
21 0.01 -8 8
22 0.01 16 16
24 0.009 16 24
25 0.001 24 24
29 0.81 4 4
31 0.81 4 12
32 0.045 12 12
34 0.045 -12 12
35 0.045 20 20
37 0.09 12 12
39 0.081 12 20
40 0.009 20 20
44 0.81 8 8
46 0.81 8 16
47 0.045 16 16
49 0.045 -16 16
50 0.045 24 24
51 0.09 -8 8
52 0.09 16 16
54 0.081 16 24
55 0.009 24 24
Table 1.3.2.2 Abs.probs. of the first step for fictive player II with CardComb = 1
Note that for pure strategies all nodes would have prob 1. We take the one with maximal gain:
node = 20, maxgain = 24, maxbet = 24, abs.prob. = 0.005.
For a pure strategy of the beginner we get always abs.prob.= 1 and we are ready. However, now in
only a fraction 0.005 (FreeProb) of the cases this maximal gain is attainable. Therefore we have to
decide what is optimal in the remaining cases with fraction 0.995 (ProbFict). We continue now in the
same way by recalculating the conditional probabilities given that node 20 is not chosen. This leads
to another chosen node. We repeat this with all chosen nodes untill there are no remaining cases any
longer (ProbFict = 0). We describe for this particular case the whole calculation process precisely in
the form of an iterative algoritm.
We start formally with FictGain = 0, FictBet = 0, ProbFict = 1 and the conditional probs of table
1.3.2.1. For the fictive player II we substitute for all cond.probs 1, indicating that all leaves can be
chosen.
Step 1:
We determine the set of nodes with positive abs. probs. We find that endnode 20 has maximal gain
MaxGain = 24 with corresponding bet MaxBet = 24.
Then we go one node backwards from this endnode 20. This gives node 13 with the 3rd leave. This
node has the cond.probs [0.9 0.05 0.05]. The 3rd element gives the current free probability:
FreeProb = 0.05 and not-normalized new vector [0.9 0.05 0]. Normalization to sum 1 gives the new
cond.probs [0.9474 0.0526 0].
Then we continue with the next backnodes with node 1 as the last one.
From node 13 we go backwards with leave 2 to node 11. This node belongs to the optimal player II,
so we do not change the ones as this leaves can still be chosen.
From node 11 we go backwards with leave 2 to node 6 of beginner I. This node has only one leave, so
nothing can change: FreeProb = 0.05 and cond.probs [1].
31
From node 6 we go backwards with leave 2 to node 4 of the optimal player II, so nothing changes.
From node 4 we go backwards with leave 3 to node 1 of the beginner I with cond.probs [0 0.9 0.1].
Then the current free prob becomes FreeProb = FreeProb x 0.1 = 0.05 x 0.1 = 0.005, leading to [0 0.9
0.1-0.005] = [0 0.9 0.0995] and after normalization to [0 0.9045 0.0955], the new cond.probs.
Finally, we update FreeProb and ProbFict:
FreeProb = FreeProb x ProbFict = 0.005 x 1 = 0.005,
ProbFict = ProbFict – FreeProb = 1 – 0.005 = 0.995,
and the gain and bet:
FictGain = FictGain + MaxGain x FreeProb = 0.12
FictBet = FictBet + MaxBet x FreeProb = 0.12.
Step 2:
With the new ProbFict and the new cond.probs we continue as under step 1.
Last Step:
The iteration stops at the step with ProbFict = 0. Then FictGain is the maximal gain of the fictive
player II. Table 1.3.2.3 shows all intermediate steps. In this particular case we have MaxBet =
MaxGain for all steps, implying the same for FictBet and FictGain. Therefore we have omitted the
bet-columns.
32
Step FreeProb Node MaxGain FictGain ProbFict
1 0.0050 20 24 0.12 0.9950
2 0.00095 25 24 0.1428 0.9940
3 0.0450 50 24 1.2228 0.9490
4 0.0086 55 24 1.4280 0.9405
5 0.0423 35 20 2.2745 0.8982
6 0.0080 40 20 2.4353 0.8901
7 0.0050 17 16 2.5145 0.8852
8 0.0081 22 16 2.6441 0.8771
9 0.0419 47 16 3.3145 0.8352
10 0.0686 52 16 4.4115 0.7666
11 0.0361 32 16 4.8445 0.7305
12 0.0590 37 12 5.5531 0.5715
13 0.0810 14 8 6.2011 0.5905
14 0.5905 44 8 10.9250 0
Table 1.3.2.3 Successive steps for FictGain for fictive player II with CardComb = 1.
The calculation of the expected gains of beginners and optimal players is done in the same way as in
section 1.3.1: randomized decisions of beginners give no new problems. So, this leads to the
expected gains and bets for all types of players for all CardCombs. Taking the means over the 4
CardCombs gives table 1.3.2.4.
Gain
I II
Bet
I II
Beginner 0.0172 -0.0172 6.2168 6.2330
Optimal 2.8700 2.5000 9.0700 10.5000
Fictive 6.3058 5.2498 8.3058 9.2498
Table 1.3.2.4 The expected gains for all types of players for randomization.
Finally, this leads to table 1.3.2.5 of the different forms of skill measures.
Sgs Srs Sgm Srm
0.4647 0.4169 0.4658 0.4185
Table 1.3.2.5 The expected gains for all types of players for randomization.
1.3.2.2 Results for the representative sample
Table 1.3.2.6 gives the results for the representative sample (table 1.2.2.2).
33
Game parameters:
Players = 2 - WithBB = 1
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 55
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 1.9522 1.7177 10.0645 9.8385
657720 1.9378 1.6847 10.1193 9.8089
986580 1.9199 1.6832 10.0850 9.8096
1315440 1.9134 1.6797 10.1001 9.8214
1644300 1.9132 1.6725 10.1273 9.7866
1973160 1.9112 1.6689 10.1430 9.7835
2302020 1.9078 1.6704 10.1118 9.7931
2630880 1.9068 1.6702 10.1320 9.7939
2959740 1.9007 1.6740 10.1060 9.8125
3288600 1.8979 1.6744 10.1066 9.7968
ExpGainArr:
0.0126 -0.0126
1.8979 1.6744
6.6880 5.7785
ExpBetArr:
6.1819 6.2048
10.1066 9.7968
9.1647 9.9290
Sgs,Srs,Sgm,Srm:
0.2865 0.2749 0.2869 0.2757
Elapsed time is 80454.379604 seconds = 0.9312 days.
Table 1.3.2.6 Results for randomized decisions (MaxRaise = 1).
1.3.3 Implied odds
The beginner strategies of table 1.3.1.3 can be modified in such a way that possible future gains are
added to the pot before expectations are taken. In poker terminoligy: we incorperate implied pot
odds. More specific, at each decision node for a specific player and a specific decision he takes the
product of the corresponding decision chips DecChips and the number CntDecisions of players with
34
bet decisions in the same round. He multiplies this number of chips with an optimistic factor FactDec
and adds this to the real pot. This gives a fictive pot. Then he takes his decision based on the maximal
expectations of this fictive pot. The factor FactDec may depend on the player: the higher the more
optimistic. In the special case that FactDec = 0 for all players, we get back table 1.3.1.3.
We work out the example that FactDec = 1 for both players. In table 1.3.3.1 we give the chips that
will be added to the pot dependent on the decisions to be taken. Only positive amounts are shown.
For example, in node 1 Player I has CntDecisions = 2 and DecChips = 2 for the possible decision Call
(product 4), and for the Raise decision CntDecision = 1 and DecChips = 6 (product 6). For all other
nodes in the table we have CntDecisions = 1.
Node = 1 PotVec: [2 4] CurPlayer: 1
Ca=4 R=6
Node = 3 PotVec: [4 4] CurPlayer: 2
B=4
Node = 11 PotVec: [8 8] CurPlayer: 2
B=8
Node = 12 PotVec: [8 8] CurPlayer: 1
B=8
Node = 13 PotVec: [8 16] CurPlayer: 1
R=16
Node = 15 PotVec: [16 8] CurPlayer: 2
R=16
Node = 26 PotVec: [4 4] CurPlayer: 2
B=8
Node = 27 PotVec: [4 4] CurPlayer: 1
B=8
Node = 28 PotVec: [4 12] CurPlayer: 1
R=16
Node = 30 PotVec: [12 4] CurPlayer: 2
R=16
Node = 41 PotVec: [8 8] CurPlayer: 2
B=8
Node = 42 PotVec: [8 8] CurPlayer: 1
B=8
Node = 43 PotVec: [8 16] CurPlayer: 1
R=16
Node = 45 PotVec: [16 8] CurPlayer: 2
R=16
Table 1.3.3.1 Fictive addional amounts to the pot for FactDec = 1.
Now expectations can be calculated exactly in the same way as it is done in table 1.3.1.2 and the
chosen decisions follow: table 1.3.3.2. Comparing with table 1.3.1.3 we see that now often decisions
with more chips are taken.
Player I
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
1
8
12
13
23
27
28
38
42
43
53
0 1 0
0 1
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
1 0 0
1 0
0 1 0
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1 0
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1 0
0 1
1 0
0 0 1
0 1
1 0
0 0 1
0 1
1 0
0 0 1
0 1
Player II
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
3
4
11
15
18
26
30
0 1
0 1
0 1
0 0 1
0 1
1 0
0 0 1
0 1
0 1
1 0
0 0 1
0 1
1 0
0 0 1
1 0
0 1
1 0
1 0 0
1 0
1 0
1 0 0
1 0
0 1
0 1
0 0 1
0 1
0 1
0 0 1
35
33
41
45
48
0 1
0 1
0 0 1
0 1
0 1
1 0
0 0 1
0 1
1 0
1 0
1 0 0
1 0
0 1
0 1
0 0 1
0 1
Table 1.3.3.2 Beginner strategies for implied odds.
With this starting point we can calculate the gains and bets of the optimal and fictive players and
thereafter the skill: see table 1.3.3.3.
36
Gain
I II
Bet
I II
Beginner 0 0 16 16
Optimal 4 4 10 8
Fictive 7 7 9 10
Sgs Srs Sgm Srm
0.5714 0.6032 0.5714 0.6143
Table 1.3.3.3 The expected gains, bets and skill for future gains
1.3.3.2 Results for the representative sample
Table 1.3.3.4 gives the results for this sample (table 1.2.2.2; GainLabel = 1).
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 1 1
RandProbDecVec = 0 0
TotNodes = 55
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 3.1152 2.4870 12.7424 11.2571
657720 3.0740 2.4257 12.8287 11.2473
986580 3.0419 2.4162 12.8766 11.2829
1315440 3.0277 2.4064 12.8965 11.2770
1644300 3.0282 2.3963 12.9139 11.2693
1973160 3.0234 2.3899 12.9108 11.2551
2302020 3.0138 2.3918 12.9192 11.2697
2630880 3.0104 2.3902 12.9227 11.2559
2959740 3.0040 2.3936 12.9235 11.2722
3288600 2.9999 2.3941 12.9317 11.2587
ExpGainArr:
-0.2741 0.2741
2.9999 2.3941
6.9779 6.2097
37
ExpBetArr:
13.8833 13.8687
12.9317 11.2587
9.5126 10.5484
Sgs,Srs,Sgm,Srm:
0.4090 0.3392 0.4043 0.3366
Elapsed time is 39233.546753 seconds = 0.4541 days.
Table 1.3.3.4 Results for implied odds (MaxRaise = 1).
1.3.4 Maximal bet under positive expectations
In section 1.3.1 beginners choose for the decisions with maximal expectations. Another more
aggressive way is to choose under all possible decisions with positive expectations that one with the
maximal bet. So, if Call and Raise have both positive expectations then the beginner chooses Raise.
The strategies follow easily with table 1.3.1.3:
Player I
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
1
8
12
13
23
27
28
38
42
43
53
0 1 0
0 1
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
1 0 0
1 0
0 1 0
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
1 0
0 1 0
0 1
0 1
0 1 0
0 1
Player II
Node CardComb = 1 CardComb = 2 CardComb = 3 CardComb = 4
3
4
11
15
18
26
30
33
41
45
48
0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 1
0 1
0 1 0
0 1
1 0
0 1 0
0 1
0 1
0 1 0
0 1
1 0
0 1
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
1 0 0
1 0
1 0
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
0 1
0 0 1
0 1
Table 1.3.4.1 Beginner strategies for maximal bet under positive exp. gain.
Gain
I II
Bet
I II
Beginner 2 -2 16 16
Optimal 4 4 10 10
Fictive 7 6 9 10
Sgs Srs Sgm Srm
38
0.6154 0.5846 0.5750 0.5727
Table 1.3.4.2 The expected gains, bets and skill for maximal bet
1.3.4.2 Results for the representative sample
Table 1.3.4.3 gives the results for the representative sample of table 1.2.2.2 (GainLabel = 0).
39
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 0
FactDecVec = 0 0
RandProbDecVec = 0 0
TotNodes = 55
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 2.3978 2.1658 13.3749 11.9443
657720 2.3702 2.0520 13.5788 11.9027
986580 2.3394 2.0319 13.5778 11.9248
1315440 2.3272 2.0177 13.6121 11.9424
1644300 2.3262 2.0023 13.6186 11.9436
1973160 2.3227 1.9939 13.5810 11.9502
2302020 2.3150 1.9938 13.5850 11.9669
2630880 2.3125 1.9914 13.5893 11.9588
2959740 2.3042 1.9945 13.5742 11.9660
3288600 2.2997 1.9939 13.5670 11.9760
ExpGainArr:
0.2796 -0.2796
2.2997 1.9939
6.8752 6.0921
ExpBetArr:
16.6369 16.6346
13.5670 11.9760
9.4241 10.4961
Sgs,Srs,Sgm,Srm:
0.3311 0.2582 0.3315 0.2606
Elapsed time is 38710.745513 seconds = 0.4480 days.
Table 1.3.4.3 Results for maximal bet under positive expected gains (MaxRaise = 1).
1.3.5 Standard deviations
In the foregoing sections 1.3.1 – 1.3.4 attention has been paid to the bias in the results of the optimal
player. The representative sample has been chosen so large that this bias is small. In this section it
will appear that this holds also for the standard deviation in the skill. We give details for the case
with the lowest skill: randomized decisions with skill S = 0.2865 (see sections 1.3.2.2).
40
Table 1.3.5.1 gives the result of 8 repetitions, of which the first one is table 1.3.2.6.
41
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 55
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Sample 1:
ExpGainArr:
0.0126 -0.0126
1.8979 1.6744
6.6880 5.7785
ExpBetArr:
6.1819 6.2048
10.1066 9.7968
9.1647 9.9290
Sgs,Srs,Sgm,Srm:
0.2865 0.2749 0.2869 0.2757
Sample 2:
ExpGainArr:
0.0136 -0.0136
1.9649 1.7127
6.6768 5.7568
ExpBetArr:
6.1961 6.2189
10.0226 9.7787
9.1024 9.8540
Sgs,Srs,Sgm,Srm:
0.2958 0.2832 0.2960 0.2837
Sample 3:
ExpGainArr:
0.0118 -0.0118
1.9831 1.7476
6.6453 5.7433
ExpBetArr:
6.1610 6.1840
9.9492 9.7153
9.1467 9.8994
Sgs,Srs,Sgm,Srm:
0.3011 0.2917 0.3014 0.2924
Sample 4:
ExpGainArr:
0.0126 -0.0126
1.8711 1.6452
6.7413 5.8218
ExpBetArr:
6.1628 6.1860
10.0132 9.7721
9.1246 9.9078
Sgs,Srs,Sgm,Srm:
0.2799 0.2692 0.2802 0.2699
Sample 5:
ExpGainArr:
0.0119 -0.0119
2.0414 1.7990
6.7336 5.8029
ExpBetArr:
6.1787 6.1997
10.0709 9.8736
9.1311 9.8935
Sgs,Srs,Sgm,Srm:
0.3063 0.2922 0.3067 0.2929
Sample 6:
ExpGainArr:
0.0123 -0.0123
1.9068 1.6549
6.6258 5.7195
ExpBetArr:
6.1618 6.1851
9.8512 9.7074
9.1811 9.9209
Sgs,Srs,Sgm,Srm:
0.2885 0.2818 0.2887 0.2822
Sample 7:
ExpGainArr:
0.0132 -0.0132
2.0202 1.7840
6.7148 5.7959
ExpBetArr:
6.1944 6.2168
10.1203 9.9432
9.1731 9.9266
Sgs,Srs,Sgm,Srm:
0.3041 0.2895 0.3044 0.2902
Sample 8:
ExpGainArr:
0.0137 -0.0137
1.9617 1.7300
6.7879 5.8734
ExpBetArr:
6.2106 6.2337
10.2491 9.9688
9.2098 9.9863
Sgs,Srs,Sgm,Srm:
0.2916 0.2768 0.2919 0.2776
Table 1.3.5.1 Result of 8 repetitions for the case of randomized decisions (MaxRaise = 1).
42
From the 8 samples in table 1.3.5.1 we get the following mean and (estimated) standarddeviation
(table 1.3.5.2):
Sgs Srs Sgm Srm
Mean = 0.2942 0.2824 0.2945 0.2831
Std = 0.0092 0.0084 0.0093 0.0084
Table 1.3.5.2 Means and standard deviations of 8 repetitions.
Taking accuracy = 2 x standarddeviation, we roughly get a range of 0.02 . This is enough for the
conclusions that will be based on it. The same holds for the simulation results for other skill measures
and beginner strategies.
1.3.6 Overview
Table 1.3.6.1 gives an overview of the results in sections 1.3.1 – 1.3.4. According to section 1.3.5 the
deviation in S can be at most 0.02.
Beginner strategies Table S=Sgs
Maximal exp.gain 1.3.1.14 0.31
Randomized decisions 1.3.2.6 0.29
Implied odds 1.3.3.4 0.41
Maximal bet 1.3.4.3 0.43
Table 1.3.6.1 Overview for skill (MaxRounds =2, Players = 2).
We repeat that the results are all based on the same sample of card distribution (table 1.2.2.2). The
strategy: maximal expectation with a little bit randomization gives the lowest skill. Implied odds and
maximal bet give a significant higher skill.
We are interested in comparisons of the skill with the juridical bound. Therefore in the next sections
with other game parameters we will only investigate the skill for beginner strategies of maximal
expected gain with randomization. We take as randomization the value 0.1. Other values lower or
higher increase the skill (results omitted).
1.4 Going all-in
In the foregoing sections the number of chips to start with was 100 for both players: ChipVec = [100
100]. This is large enough to take every desired decision. In fact this could have been reduced to 24
without changing figure 1.1.1. But if at a certain stage the number of chips is less than (or equal to) to
43
the chips needed for a decision, a player can still take this decision with all his chips. Then his
remaining chips are 0 and further decisions of him are not allowed: he is going all in. The rule is that a
player with no chips cannot take decisions further on. He takes part in the showdown only for the
total chips he has put into the pot. So we have to consider the influence of going all-in. The remaining
players continue to play for the remainder of the pot, the so called side pot. For Players = 2 this is
rather simple: if one player goes all in, the game ends at last when it would be his turn again.
However, for Players >=3 the other players can continue. This can become quite complicated and
more than one side pot can occur.
In any case, the game tree depends on the number of chips if the total maximal bet of decisions is
larger than the number of starting chips of a player.
In this section we illustrate this effect with the full game tree of figure 1.1.1 for the case that ChipVec
= [25 7]. So player I starts with 25 chips and II with 7. Figure 1.4.1 gives the reduction of the game
tree of figure 1.1.1.
Figure 1.4.1 The game tree for all-in with ChipVec = [ 25 7].
At node 3 player II bets all-in with the 3 remaining chips. Then at node 8 player I makes the last
decision. At nodes 4 and 15 player II calls all-in with the 3 remaining chips and has only 7 chips
44
contributed to the pot. So the sidepot for the showdown is [ 7 7]. The remaining pot is simply for
Player I. Table 1.4.2 gives for different ChipVecs the total nodes of the resulting game tree.
I | II 25 23 15 7 3 1
25
23
15
7
3
1
55 49 40 19 3 3
49 43 40 19 3 3
40 40 37 19 3 3 16 61 16 16 3 3
3 3 3 3 3 3
1 1 1 1 1 1
Table 1.4.2 Total nodes dependent on ChipVec.
Note that for Player I starting with 1 chip the game is reduced to pure gambling. Of course, such
games are never played in fixed-limit THM-cash games but can appear in a specific game in THM-
tournements.
1.5 Strategies from a game-theoretic perspective
In the foregoing section 1.3 no attention has been paid to the optimal and fictive strategies
themselves because only the corresponding expected gains and bets are needed for determining the
skill. But optimal play against a given (e.g. beginner) strategy is an interesting topic itself. Even more
interesting is the best play against unknown strategies of the opponents. This is a typical topic in
game theory. It has nothing to do with skill and so this section can be omitted if one is not interested
in strategies for optimal play.
We say that a strategy profile of the set of all players form a (Nash-)equilibrium if for each particular
player his strategy is optimal given the strategies of the other players. Every game has at least one
equilibrium. Clearly, if for a certain game the beginner strategies together form an equilibrium then
the skill of this game is 0. But this has only theoretical value because for practical games (e.g. full
THM, bridge and chess) equilibria cannot even be calculated and therefore will never be played.
For two-person zero-games a little bit more can be said (e.g. THM without rake). By definition, the
gain of player I, say, is always the loss of player II. An equilibrium consists of a so called maximin-
strategy of player I (maximizing his minimal expected gain taken over the strategies of player II), and
a minimax-strategy of player II (minimizing his maximal loss token over the strategies of player I). The
corresponding expected gain of I (loss of II) is the same for each equilibrium and is called the value of
the game.
In this section we can say a little bit about the value and corresponding strategies of the simplified
fixed-limit THM-version of section 1: the base game. The usual way to calculate minimax-strategies of
45
a game is to transform it to its normal form and use linear programming. But this gives only success
for very simple games, certainly not for the THM-basicgame. However, another possibility is to find
the solution by a modified form of fictitious play straightforward from the game tree (the extensive
form of the game). This iteration technique can be applied by extending the computer program for
skill in such a way that it gives also the optimal strategy of a player against given strategies of his
opponents. There remains one serious objection: the convergence of the iteration procedure is
extremely slow.
Table 1.5.1 gives for THM-basic the result of fictitious play. We start with the beginner strategies of
section 1.3.1.2: simply maximal expected gain. The iteration process has been stopped after TotRpts
= 482 iterations (about 250 days).
46
NoOptFictPl = 1 - Rpts = 24
Players = 2 - WithBB = 1
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
GainLabel = 0
FactDec = 0 0
RandProbDecVec = 0 0
TotNodes = 55
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
TotRpts = 1
ExpGainArr:
-0.3516 0.3516
1.8244 5.0168
ExpBetArr:
14.2835 13.9554
12.6737 19.4443
TotRpts = 2
ExpGainArr:
-0.6129 0.6129
2.1988 3.0855
ExpBetArr:
14.3756 15.0318
12.7908 18.5625
................................................................
TotRpts = 481
ExpGainArr:
0.5207 -0.5207
0.5697 -0.4718
ExpBetArr:
10.7187 10.1288
10.5151 9.9656
TotRpts = 482
ExpGainArr:
0.5207 -0.5207
0.5696 -0.4719
ExpBetArr:
10.7182 10.1283
10.5219 9.9647
Table 1.5.1 Fictitious play after 482 rounds
Although there is hardly a difference between TotRpts = 481 and TotRpts = 482, there is still not a
sharp convergence. For that we need the equality between the ‘beginners’ and the ‘optimal’ players:
for player I the difference is still 0.5696 - 0.5207 = 0.0489, and for player II 0.0488. This prevents
47
sharp conclusions but gives more than only a qualitative description. A rough estimate for the
minimax-value of the game is + 0.52. This means that player I has the advantage over player II. Of
course this is due to the fact that the SB player I has to invest less in the pot than the BB player II.
Convergence of the corresponding optimal maximin- and minimax strategies of the players has been
far from reached. Table 1.5.2 gives a rough impression on the base of equities. The Eq(uity) of a card
combination is the expected gain of the pot of size 1, in this case Eq = WProb + EProb/2 with WProb
the winning probability and EProb the equal probability. Decisions for card combinations with more
ore less the same Eq have been averaged. In the table equities and probs in strategies are in
percentages (i.e. multiplied by 100).
Node 1: Eq Strat
30 33 2 65
40 32 2 66
50 12 2 86
60 7 1 91
70 6 1 93
80 0 1 99
90 0 1 99
Node 3: Eq Strat
30 67 33
40 56 44
50 40 60
60 23 77
70 13 87
80 0 100
90 0 100
Node 4: Eq Strat
30 31 69
40 32 68
50 12 88
60 7 93
70 5 95
80 0 100
90 0 100
Node 8: Eq Strat
30 31 69
40 32 68
50 14 86
60 8 92
70 7 93
80 0 100
90 0 100
Node: 11
Eq Strat
0 51 49
10 56 44
20 80 20
30 91 9
40 100 0
50 99 1
60 94 6
70 78 22
80 48 52
90 42 58
100 45 55
Node: 12
Eq Strat
0 2 98
10 11 89
20 68 32
30 87 13
40 99 1
50 97 3
60 83 17
70 36 64
80 8 92
90 0 100
100 0 100
Node: 13
Eq Strat
0 96 0 4
10 96 0 4
20 86 9 5
30 61 28 11
40 43 51 6
50 35 58 7
60 31 65 5
70 7 92 2
80 1 85 14
90 0 25 75
100 0 0 100
Node: 15
Eq Strat
0 100 0 0
10 99 0 1
20 94 4 2
30 71 24 6
40 45 51 4
50 33 63 4
60 29 68 3
70 4 94 2
80 0 82 17
90 0 15 85
100 0 0 100
Node: 18
Eq Strat
0 100 0
10 100 0
20 99 1
30 83 17
40 89 11
50 80 20
60 64 36
70 48 52
80 31 69
90 3 97
100 0 100
Node: 23
Eq Strat
0 100 0
10 100 0
20 100 0
30 87 13
40 96 4
50 86 14
60 73 27
70 50 50
80 24 76
90 2 98
100 0 100
Node: 26
Eq Strat
0 95 5
10 97 3
20 99 1
30 99 1
40 100 0
50 99 1
Node: 27
Eq Strat
0 6 94
10 33 67
20 87 13
30 80 20
40 99 1
50 91 9
Node: 28
Eq Strat
0 98 0 2
10 96 1 3
20 85 10 5
30 78 13 9
40 66 27 7
50 45 50 5
48
60 98 2
70 95 5
80 89 11
90 84 16
100 84 16
60 76 24
70 31 69
80 5 95
90 0 100
100 0 100
60 33 64 3
70 4 91 5
80 1 73 27
90 0 25 75
100 0 0 100
Node: 30
Eq Strat
0 100 0 0
10 99 0 1
20 90 9 2
30 76 20 4
40 61 37 2
50 41 56 3
60 32 65 2
70 4 91 4
80 1 72 27
90 0 14 86
100 0 0 100
Node: 33
Eq Strat
0 100 0
10 100 0
20 99 1
30 91 9
40 88 12
50 79 21
60 68 32
70 52 48
80 26 74
90 1 99
100 0 100
Node: 38
Eq Strat
0 100 0
10 100 0
20 99 1
30 86 14
40 95 5
50 87 13
60 74 26
70 48 52
80 24 76
90 2 98
100 0 100
Node: 41
Eq Strat
0 5 95
10 18 82
20 68 32
30 85 15
40 99 1
50 98 2
60 89 11
70 68 32
80 40 60
90 31 69
100 38 62
Node: 42
Eq Strat
0 43 57
10 52 48
20 82 18
30 86 14
40 99 1
50 97 3
60 81 19
70 37 63
80 7 93
90 0 100
100 0 100
Node: 43
Eq Strat
0 98 0 2
10 98 0 2
20 87 9 4
30 59 31 9
40 42 52 6
50 32 62 6
60 28 68 5
70 5 93 2
80 0 85 14
90 0 22 78
100 0 0 100
Node: 45
Eq Strat
0 99 0 1
10 98 0 2
20 86 11 3
30 67 27 6
40 48 48 4
50 40 56 4
60 32 65 3
70 5 92 3
80 0 81 18
90 0 16 84
100 0 0 100
Node: 48
Eq Strat
0 100 0
10 100 0
20 99 1
30 90 10
40 86 14
50 79 21
60 64 36
70 47 53
80 30 70
90 3 97
100 0 100
Node: 53
Eq Strat
0 100 0
10 100 0
20 100 0
30 89 11
40 96 4
50 85 15
60 69 31
70 41 59
80 21 79
90 2 98
100 0 100
Table 1.5.2 Approximate optimal (maximin- and minimax) strategies.
So, the preflop 2c3d gives the minimal equity 0.3203, or rounded off 30%. In the table we see that in
node 1 the decision of player I is [ 33 2 65], randomization with 33% Fold, 2% Call and 65%
Raise.The preflop AcAd gives the maximal equity 0.8520, or rounded of 90%.
The strategy profile of table 1.5.2 can be applied in practice if we can make a rough guess of the
winning probability of the card combination. The effect of the equal probability is almost always
negligible. In the table it is interesting to observe the role of bluffing and sandbagging reflected in the
decision probabilities.
49
2. GENERAL FIXED-LIMIT THM
2.1 Introduction
In section 1 we have analyzed the THM-basic game: a simplified form of fixed-limit THM with the
simplifying parameters Players = 2, MaxRounds = 2, MaxRaises = 1. It appeared that beginner
strategies with randomized decisions give the lowest skill. Therefore we continue only with this
choice and analyze what will happen with the skill for larger values of these parameters. The rules
that are used for full THM are Players >=2, MaxRounds = 4 and MaxRaises =3. The difficulty with
these parameters is that the game tree becomes very large. Table 2.1.1 gives a specification.
MaxRounds Pl = 2 Pl = 3 Pl = 4 Pl = 5 ...
2 MaxRaise = 1
MaxRaise = 2
MaxRaise = 3
3 MaxRaise = 1
MaxRaise = 2
MaxRaise = 3
4 MaxRaise = 1
MaxRaise = 2
MaxRaise = 3
55 509 3902 27307 ...
121 2579 41927 ...
211 11171 ...
280 5634 86287 ...
856 59322 ...
1912 ...
1405 ... ...
6001 ...
17221 ...
Table 2.1.1 Number of nodes for different values of parameters.
Furthermore, for MaxRounds > 2 a representative sample must be taken in a different way to
emphasize the increasing information the optimal player has in each round. This makes that the
needed computer time to do the required calculations becomes too large. Therefore we vary these
parameters independently of each other and analyze the change in skill. In section 2.2 we do this for
MaxRaises, in section 2.3 for Players and in section 2.4 for MaxRounds. In section 2.5 we will
extrapolate to the effect of changing these parameters simultaneously.
So starting point in the analysis is section 1 for Players = 2, MaxRounds = 2, MaxRaises = 1. For the
beginner strategies we use the randomization as explained in section 1.3.2.
2.2 More raises
2.2.1 MaxRaises = 2
50
We change MaxRaises = 1 (table 1.3.2.6) to MaxRaises = 2. Figure 2.2.1.1 gives a part of the game
tree: only the first round. The whole game tree counts 121 nodes (see table 1.2.1). Note that the bet
at node 3 is counted as a raise. So at node 14 there is no choice for another raise.
Figure 2.2.1.1 Gametree for MaxRaises = 2, MaxRounds = 2 and Players = 2.
2.2.2 MaxRaises = 3
Table 2.2.1.1 gives the result if we change MaxRaises = 1 (table 1.3.2.6) to MaxRaises = 3, the number
used in practice.
51
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 3 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 211
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 2.2280 2.0248 10.8533 10.8160
657720 2.2112 1.9875 10.9248 10.7803
986580 2.1913 1.9855 10.8920 10.7883
1315440 2.1841 1.9812 10.8809 10.7989
1644300 2.1837 1.9729 10.9260 10.7723
1973160 2.1817 1.9689 10.9520 10.7474
2302020 2.1777 1.9706 10.9207 10.7710
2630880 2.1768 1.9703 10.9378 10.7668
2959740 2.1700 1.9747 10.9096 10.7720
3288600 2.1669 1.9750 10.9333 10.7812
ExpGainArr:
0.0127 -0.0127
2.1669 1.9750
7.9633 7.0967
ExpBetArr:
6.2798 6.3027
10.9333 10.7812
10.4492 11.2473
Sgs,Srs,Sgm,Srm:
0.2750 0.2748 0.2753 0.2753
Elapsed time is 402027.074354 seconds = 4.6531 days.
Table 2.2.1.1 Result for randomized decisions (MaxRaises = 3).
Comparing the results with MaxRaises = 1 we see that there is not much difference between the two
settings. Our conclusion is that the number of allowed raises in THM does not change the skill
significantly. However, there is one important practical difference: the needed computer time for
MaxRaises = 3 is 5 x as large as for MaxRaises = 1.
2.3 More players
52
At first we consider the generalization of the simplified fixed-limit THM-game from 2 players (section
1.1) to 3 players. Note that in this case player I has the SB, II the BB and that III starts the real betting.
Assuming that their starting chips are large enough to avoide going all-in, the game tree grows from
55 to 509 nodes. This is too large to make a complete sketch like figure 1.1.1. In the figures 2.3.1 we
have only displayed the first betting round.
53
Figure 2.3.1 Main Figure 2.3.1a (Node 2)
Figure 2.3.1b (Node 3) Figure 2.3.1c (Node 4)
54
Player I has the SB, player II the BB and player III starts the real betting. Therefore in the second
round player I starts again the betting unless he has already folded (e.g. node 56).
As beginner strategies for the 3 players we consider only the case of randomized decisions of section
1.3.2 for all of them. The calculation of expectations is based on winning probabilities. More
specifically, for a specific player we need now 3 probabilities: the highest card value (Win = 1), equal
card values with one other player (Win = 2) and all equal cardvalues (Win = 3). Enumerating and
saving all possible card combinations takes too much computer time and space. So, as an
approximation we calculate the probs under the assumption that the preflops of all opponents are
drawn with replacement given the cards of the specific player and other open cards. Then we can
calculate the desired probablities PW from P2 of 2 players, which were saved in files (section
1.3.1.1). The general formula is (P = Players, W = Win):
PW(W) = C(P-1,W-1) x P2(2)^(W-1)... x P2(1)^(P-W), 1 <= W <= P.
Take the following example:
P2(1) = 0.2924 (win), P2(2)= 0.0613 (equal).
Then for 3 players we get:
PW(1) = 0.0855, PW(2) = 0.0358, PW(3) = 0.0038
and for 4 players:
PW(1) = 0.0250, PW(2) = 0.0157, PW(3) = 0.0033, PW(4) = 0.0002 .
With these approximations of probabilities we can calculate gain expectations and proceed as in
section 1.3 to obtain the beginner strategies. In particular this holds for randomized decisions as in
section 1.3.2.
2.3.1 Elementary sample
We start with an extension of the elementary sample for 2 players of section 1.2.1: see table 2.3.1.1.
Again this is just intended to explain the analysis and has no practical value.
CardComb Prefl.I Prefl.II Prefl.III River
1
2
3
4
5
6
7
8
9
10
1 8
1 8
1 8
1 8
42 46
42 46
42 46
42 46
29 40
29 40
42 46
42 46
29 40
29 40
1 8
1 8
29 40
29 40
1 8
1 8
29 40
29 40
42 46
42 46
29 40
29 40
1 8
1 8
42 46
42 46
26 36 43 48 49
3 4 5 33 38
26 36 43 48 49
3 4 5 33 38
26 36 43 48 49
3 4 5 33 38
26 36 43 48 49
3 4 5 33 38
26 36 43 48 49
3 4 5 33 38
55
11
12
29 40
29 40
42 46
42 46
1 8
1 8
26 36 43 48 49
3 4 5 33 38
Table 2.3.1.1 Card numbers of an elementary sample of 12 card combinations.
Table 2.3.1.2 gives the relevant card names and poker values.
Card numbers Card names Poker Values
1 8 26 36 43 48 49
1 8 3 4 5 33 38
42 46 26 36 43 48 49
42 46 3 4 5 33 38
29 40 26 36 43 48 49
29 40 3 4 5 33 38
2c 3s 8d Ts Qh Ks Ac
2c 3s 2h 2s 3c Tc Jd
Qd Kd 8d Ts Qh Ks Ac
Qd Kd 2h 2s 3c Tc Jd
9c Js 8d Ts Qh Ks Ac
9c Js 2h 2s 3c Tc Jd
1 14 13 12 10 8
7 2 3 0 0 0
3 13 12 14 0 0
2 2 13 12 11 0
5 14 0 0 0 0
3 11 2 10 0 0
Table 2.3.1.2 Card names and poker values.
For CardComb = 1 and Player =1 the preflop-winning probs are just the probs of the example in the
foregoing introduction: 0.2924 (Win = 1), 0.0613 (Win = 2). The river-probs are: 0 (Win = 1 ),
0.2030 (Win = 2). From these probs the exected gain can be calculated. This leads for this CardComb
to the best decision (maximal exp. gain). Results for the beginner strategies of the other players and
card combinations follow in the same way. From that we can calculate the gains of the beginners,
optimal players and fictive players. We omit the lenghtly details now. The final result is given in table
2.3.1.2. Of course, since the sample is very small the gain of the optimal player has a high positive
bias and this explains the high unrealistic skill value.
ExpGainArr:
I II III
B 0 1.3333 -1.3333
O 7.3333 7.0000 8.5000
F 9.1667 8.0000 10.6667
ExpBetArr:
B 3.6667 4.0000 1.3333
O 9.3333 10.0000 9.3333
F 9.5000 10.3333 8.6667
Sgs: Srs: Sgm: Srm:
0.8204 0.8156 0.8231 0.8342
Table 2.3.1.2 Results for the elementary sample for 3 players (max.exp.gain).
56
2.3.2 Representative sample
Table 2.3.2.1 gives the results for a representative sample. The parameters SimCards = 30 and
SimRiver = 20 determine the size TotCombs = 1 663 200.
Game parameters:
Players = 3
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0 0
RandProbDecVec = 0.1000 0.1000 0.1000
TotNodes = 509
SimCards = 12 MaxSimCards = 47
SimPlayers = 83160 MaxSimPlayers = 966381570
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 1663200
Bias test:
TotCombVec MGainArr MBetArr
166320 2.5358 2.4774 4.2955 8.8241 9.8405 7.4507
332640 2.5303 2.4546 4.2377 8.8259 9.8536 7.4061
498960 2.5110 2.4453 4.2319 8.8427 9.8480 7.3714
665280 2.5170 2.4310 4.2354 8.8090 9.8353 7.4169
831600 2.5149 2.4278 4.2262 8.8306 9.8534 7.3926
997920 2.5202 2.4207 4.2221 8.8272 9.8383 7.3407
1164240 2.5222 2.4198 4.2247 8.8250 9.8186 7.2991
1330560 2.5202 2.4233 4.2209 8.8287 9.8139 7.3609
1496880 2.5234 2.4198 4.2226 8.8322 9.8158 7.3088
1663200 2.5212 2.4192 4.2279 8.8881 9.8319 7.3455
ExpGainArr:
-0.1375 -0.0914 0.2289
2.5212 2.4192 4.2279
6.8373 5.7183 9.4738
ExpBetArr:
6.1112 6.3903 1.2239
8.8881 9.8319 7.3455
8.8652 9.6575 7.5774
Sgs,Srs,Sgm,Srm:
0.4162 0.4167 0.4153 0.3935
Elapsed time is 901234.627429 seconds = 10.4310 days.
Tabel 2.3.2.1 Result for 3 players (MaxRaises = 1, MaxRounds = 2).
We see that the skill for 3 players is substantially larger than that for 2 players (see table 1.3.2.6).
Furthermore, note that the result for beginner player III is positive, due to the fact that he is not
involved in prepaying blinds.
57
2.4 MaxRounds = 4
We consider the general case that the number of rounds is 4: preflop, flop, turn, river and that the
betting is doubled from the turn on. Tabel 2.4.1 gives the result.
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 4 - DoubleRound = 3
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
SecMakeTree = 0.283
TotNodes = 1405
SimCards = 15 MaxSimCards = 47
SimPlayers = 8190 MaxSimPlayers = 1070190
SimRiver = 3 MaxSimRiver = 48
SimFlop = 8 MaxSimFlop = 22100
SimTurn = 3 MaxSimTurn = 49
SecGenCards = 75.730
TotCombs = 589680
Bias test:
TotCombVec MGainArr MBetArr
58968 6.0775 5.4849 16.8760 16.2886
117936 5.7336 5.2150 16.8186 16.2835
176904 5.6179 5.1063 16.7041 16.2450
235872 5.5684 5.0526 16.8005 16.2469
294840 5.5242 5.0179 16.7995 16.1880
353808 5.5294 4.9761 16.7494 16.1751
412776 5.5282 4.9488 16.7763 16.1392
471744 5.5307 4.9261 16.8360 16.1114
530712 5.5104 4.9288 16.8000 16.1480
589680 5.4993 4.9204 16.7950 16.1705
ExpGainArr:
0.0216 -0.0216
5.4993 4.9204
14.7062 13.8642
ExpBetArr:
8.3733 8.4103
16.7950 16.1705
17.4587 18.2325
Sgs,Srs,Sgm,Srm:
0.3647 0.3949 0.3645 0.3945
Elapsed time is 1423045.281676 seconds = 16.4704 days.
Table 2.4.1 Result for 2 players (MaxRaises = 1, MaxRounds = 4).
58
The step from MaxRounds = 2 (table 1.3.2.6) to MaxRounds = 4 leads to a substantial increase of
skill. This should be kept in mind when judging other variations.
2.5 Combining results
Table 2.5.1 gives an overview of the foregoing variations (S = Sgs).
59
Basic example Table 1.3.2.6 S = 0.2865
MaxRaises = 3 Table 2.2.1.1 S = 0.2750
Players = 3 Table 2.3.2.1 S = 0.4162
MaxRounds = 4 Table 2.4.1 S = 0.3647
Table 2.5.1 Individual variations on the basic example.
Note that in this table possible deviations are about 0.02 ( see section 1.3.4). We conclude that the
increase of the number of raises has a negligible influence. Contrarily, the increase of the number of
players and the number of rounds is substantial. The general save conclusion can be that the skill of
full fixed-limit THM is larger than the juridical upper bound 0.3.
3. SPREAD-LIMIT THM
We take a simple example of spread-limit as a variation of the basic example: see figure 3.1. BetVec =
[4 8] means that a player has the choice between a bet of 4 (as in section 1) and bet of 8. RaiseVec
=[6] means a choice between the usual raise of 2 x the bet or an additional amount 6 extra.
60
Figure 3.1 Spread-limit with BetVec=[4 8] and RaiseVec = [8].
Table 3.2 gives the result.
61
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4 8
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = 8
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 211
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 4.2353 3.4490 17.7965 15.5967
657720 4.1980 3.3811 17.9027 15.5014
986580 4.1604 3.3745 17.8630 15.5671
1315440 4.1434 3.3635 17.8935 15.5370
1644300 4.1407 3.3521 17.8928 15.5181
1973160 4.1404 3.3439 17.9374 15.4891
2302020 4.1320 3.3467 17.9037 15.4907
2630880 4.1304 3.3458 17.8677 15.5163
2959740 4.1192 3.3512 17.8561 15.4996
3288600 4.1130 3.3514 17.8730 15.5303
ExpGainArr:
0.0206 -0.0206
4.1130 3.3514
16.7779 14.4178
ExpBetArr:
7.5541 7.5419
17.8730 15.5303
20.2158 19.2316
Sgs,Srs,Sgm,Srm:
0.2393 0.2826 0.2389 0.2827
Elapsed time is 426384.754621 seconds = 4.9350 days.
Tabel 3.2 Spread-limit result for 2 players (MaxRaises = 1, MaxRounds = 2).
Compared with the basic example (table 1.3.2.6) we see a decrease of the skill.
The spread in table 3.2 can be described qualitatively as high-low. Table 3.3 gives the result of a more
refined spread. The skill decreases even more.
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0
BetVec = 4 6 8 10
62
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = 8
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 547
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 4.7263 4.1650 19.2616 17.8413
657720 4.6804 4.0732 19.3370 17.7838
986580 4.6384 4.0614 19.1948 17.7907
1315440 4.6200 4.0462 19.2606 17.8227
1644300 4.6147 4.0320 19.3092 17.7854
1973160 4.6140 4.0203 19.3238 17.7380
2302020 4.6042 4.0240 19.2795 17.7087
2630880 4.6022 4.0231 19.2957 17.7016
2959740 4.5893 4.0287 19.2541 17.7035
3288600 4.5824 4.0290 19.2565 17.7498
ExpGainArr:
0.0257 -0.0257
4.5824 4.0290
20.6804 22.1695
ExpBetArr:
7.9803 7.9837
19.2565 17.7498
24.2830 27.3132
Sgs,Srs,Sgm,Srm:
0.2010 0.2802 0.2016 0.2796
Elapsed time is 1947561.203295 seconds = 22.5412 days
Tabel 3.3 Refined spread-limit result for 2 players (MaxRaises = 1, MaxRounds = 2).
Note that the beginner strategies are based on maximal expected gain. This requires that winning
probabilities must be calculated more accurately with increasing spread. This is too much to require
for a beginner. We have no simple substitute for describing beginner strategies for such cases.
Therefore we omit spread limit results in drawing general conclusions.
4. POT-LIMIT THM
We take a simple example of potlimit-THM as a variation of the basic example.
Comparing the gametree for this pot-limit with that of section 3 for spread-limit, we get that the
part of it shown in figure 3.1 is the same as for pot-limit. The only difference is node 5 with leave R =
63
8 (instead of R = 14) followed at node 9 with Ca = 6 (instead of Ca = 12). Further up in the tree the
structure deviates. Table 4.2 gives the result.
64
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 1 - NoLimit = 0
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 193
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 4.0219 4.0419 17.3508 18.4438
657720 3.9825 3.9534 17.2447 18.3481
986580 3.9496 3.9476 17.2050 18.3240
1315440 3.9344 3.9397 17.2000 18.3837
1644300 3.9314 3.9264 17.2435 18.3592
1973160 3.9265 3.9193 17.2767 18.3031
2302020 3.9187 3.9227 17.2371 18.3423
2630880 3.9136 3.9208 17.2258 18.2786
2959740 3.9012 3.9289 17.2335 18.3228
3288600 3.8953 3.9304 17.2572 18.3245
ExpGainArr:
0.0157 -0.0157
3.8953 3.9304
14.8799 15.3307
ExpBetArr:
6.7199 6.7496
17.2572 18.3245
18.0368 20.3203
Sgs,Srs,Sgm,Srm:
0.2590 0.2792 0.2591 0.2790
Elapsed time is 371862.046053 seconds = 4.3040 days.
Tabel 4.2 Pot-limit result for 2 players (MaxRaises = 1, MaxRounds = 2).
Compared with the basic example (table 1.3.2.6) we see a small decrease of the skill. This decrease is
lower than for the spread-limit cases in section 4. However, the same objections for beginner
strategies apply.
5. NO-LIMIT THM
We take a simple example of potlimit-THM as a variation of the basic example: see figure 5.1. Note
that the game stops immediately when a player goes all in sice there are only 2 players.
65
66
Figure 5.1 No-limit with ChipVec = [100 100] (MaxRaises = 1, MaxRounds = 2).
Table 5.2 gives the result.
67
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 1
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 97
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 24.6636 24.5636 51.2680 51.6521
657720 24.4308 24.2429 51.6546 51.5529
986580 24.3080 24.1809 51.4212 51.3944
1315440 24.2942 24.1559 51.1874 51.3353
1644300 24.2802 24.0843 51.3043 51.1924
1973160 24.2800 24.0529 51.1955 51.0645
2302020 24.2455 24.0651 51.1790 50.8304
2630880 24.2468 24.0589 51.0647 50.8448
2959740 24.2061 24.0900 50.9056 50.9111
3288600 24.1887 24.1030 50.9077 50.9228
ExpGainArr:
0.0107 -0.0107
24.1887 24.1030
46.9817 46.1221
ExpBetArr:
22.8776 22.8886
50.9077 50.9228
51.6570 52.2286
Sgs,Srs,Sgm,Srm:
0.5187 0.5292 0.5187 0.5292
Elapsed time is 97485.810493 seconds = 1.1283 days.
Tabel 5.2 No-limit result with ChipVec = [100 100] (Randomization).
Compared with the basic example (table 1.3.2.6) we see a large increase of the skill. It is a surprise
that there is such a big difference in skill with potlimit (table 4.2). The reason is not the
randomization as table 5.3 shows. Here the beginner is just based on maximal expectation.
68
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 1
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0 0
TotNodes = 97
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 25.6402 25.4974 50.4747 50.6842
657720 25.3989 25.1677 50.8627 50.7435
986580 25.2764 25.1019 50.7210 50.6572
1315440 25.2665 25.0781 50.5023 50.6950
1644300 25.2505 25.0051 50.5472 50.4925
1973160 25.2505 24.9727 50.5232 50.3981
2302020 25.2158 24.9860 50.5033 50.2244
2630880 25.2173 24.9798 50.3738 50.2213
2959740 25.1762 25.0110 50.1466 50.2702
3288600 25.1596 25.0249 50.2157 50.2676
ExpGainArr:
-0.0000 0.0000
25.1596 25.0249
47.0545 46.1928
ExpBetArr:
4.2986 4.2986
50.2157 50.2676
51.1758 52.0375
Sgs,Srs,Sgm,Srm:
0.5382 0.5528 0.5382 0.5529
Elapsed time is 49660.943425 seconds = 0.5748 days.
Tabel 5.3 No-limit result with ChipVec = [100 100] (No randomization)
We can combine PotLimit and Nolimit. This leads to an extension of the gametree. Tabel 5.4 gives the
result.
69
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 1 - NoLimit = 1
BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = [ ]
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 283
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 24.8141 24.7936 51.1117 51.6377
657720 24.5785 24.4688 51.4883 51.5297
986580 24.4560 24.4058 51.2905 51.3542
1315440 24.4430 24.3813 51.0622 51.3044
1644300 24.4288 24.3094 51.1661 51.1470
1973160 24.4287 24.2776 51.0888 51.0267
2302020 24.3940 24.2899 51.0815 50.7900
2630880 24.3949 24.2837 50.9259 50.8019
2959740 24.3544 24.3148 50.7459 50.8589
3288600 24.3371 24.3280 50.7797 50.8749
ExpGainArr:
0.0126 -0.0126
24.3371 24.3280
47.4227 46.6709
ExpBetArr:
20.1154 20.1314
50.7797 50.8749
52.1775 53.2952
Sgs,Srs,Sgm,Srm:
0.5172 0.5366 0.5172 0.5367
Elapsed time is 330035.137109 seconds = 3.8199 days
Tabel 5.4 Combined Pot- and No-limit result (No randomization)
Surprisingly, there is no much difference in skill with NoLimit alone. If we take a more spread pattern
of bets, the result is the same as shown in table 5.5.
70
Game parameters:
Players = 2
Small Blind = 2 - Big Blind = 4
PotLimit = 1 - NoLimit = 1
BetVec = 4 6 8
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec = 4 8
ChipVec = 100 100
Beginner parameters:
GainLabel = 1
FactDecVec = 0 0
RandProbDecVec = 0.1000 0.1000
TotNodes = 1087
Data from file:
SimCards = 30 MaxSimCards = 47
SimPlayers = 164430 MaxSimPlayers = 1070190
SimRiver = 20 MaxSimRiver = 2598960
TotCombs = 3288600
Bias test:
TotCombVec MGainArr MBetArr
328860 24.9518 24.9367 51.0909 51.6746
657720 24.7089 24.6099 51.4591 51.5763
986580 24.5862 24.5464 51.2229 51.3894
1315440 24.5730 24.5227 51.0065 51.3408
1644300 24.5582 24.4505 51.0924 51.1795
1973160 24.5568 24.4185 51.0585 51.0601
2302020 24.5222 24.4311 51.0577 50.8241
2630880 24.5228 24.4249 50.9033 50.8322
2959740 24.4821 24.4561 50.7353 50.8809
3288600 24.4652 24.4696 50.7028 50.8958
ExpGainArr:
0.0185 -0.0185
24.4652 24.4696
48.1735 46.9229
ExpBetArr:
15.3142 15.3209
50.7028 50.8958
53.5390 53.3861
Sgs,Srs,Sgm,Srm:
0.5146 0.5416 0.5147 0.5416
Elapsed time is 1729939.300924 seconds = 20.0224 days
Tabel 5.5 Combined Spread, Pot- and No-limit result (randomization)
The general conclusion is that the skill is very high as soon as no-limit is allowed.
6. CONCLUSIONS
Table 6.1 gives an overview of the foregoing game variations in sections 3 - 5 (S = Sgs).
Basic example Table 1.3.2.6 S = 0.2865
71
No-limit Table 5.2 S = 0.5187
Combined Table 5.5 S = 0.5146
Table 6.1 Game variations on the basic example.
In this tables the results for spread-limit (section 3) and pot-limit (section 4) have been omitted. As
already explained there, the required ability for calculating accurately winning probabilities is too
heavy for beginners.
Noe that in table 6.1 all game variations have the same number of raises and rounds: MaxRaises = 1
and MaxRounds = 4. We have seen in section 2.5 that especially the usual number of MaxRounds = 4
and the number of players do increase the skill substantially. Therefore, we conclude that not only
for full fixed-limit but also for other usual full game variations the skill is larger than the upper
juridival bound 0.3.
72
REFERENCES
[1] Texas Hold ‘Em: a game of skill
For other references we refer to [1].
73
APPENDIX: MATLAB-PROGRAM
In this appendix we describe roughly the MATLAB-program with which all results were obtained.
Note that the program only works on a 64-bit Windows- 7 PC with at least an internal memory of 10
Gb and with a 64-bit MATLAB-version of at least version 7.12 (R2011a).
A1. Files of the program THM.
All files have been put on the disc THMDVD. They should be copied all to the same directory.
There are 3 types of files: m-files with the program code, mat-files as part of the program and mat-
files generated by the program. The following overview does not contain generated files.
m-files with program code:
THM the main program, reads all parameters;
calls CalcTHM.
CalcTHM Used by THM;
does the calculations, loads and saves files of the type CardCombArr**;
eventually saves the file THMStratsFile for strategies and displays results;
calls MakeTreeTotTHM, GenTHMCardCombArr, CalcTHMProbs,
CalcTHMOptCardInx, CalcTHMStrats, CalcSkill.
MakeTreeTotTHM Used by CalcTHM;
generates the game tree.
GenTHMCardComArr used by CalcTHM;
generates the card combinations CardCombArr.
CalcTHMProbs used by CalcTHM;
calculates for CardCombArr the probs;
calls GetProbPreflop, GetProbFlop, GetProbTurn, GetProbRiver
internal function InitGlobalVars uses the following mat-files:
RiverCountsFile, NumComm5CardsVecFile,
TurnCountsFile, NumComm4CardsFile,
FlopCountsFile, NumComm3CardsFile,
PreflopCountsFile, NumPreflopVecFile.
GetProbPreflop used by CalcTHMProbs;
calculates probs for the preflop;
calls TransformPreflopComm, Hand2Nr, BinSearch.
GetProbFlop used by CalcTHMProbs;
calculates probs for the flop;
calls TransformPreflopComm, Hand2Nr, BinSearch.
GetProbTurn used by CalcTHMProbs;
calls TransformPreflopComm, Hand2Nr, BinSearch;
calculates probs for the turn.
GetProbRiver used by CalcTHMProbs;
calculates probs for the river;
calls TransformPreflopComm, Hand2Nr, BinSearch.
TransformPreflopComm used by GetProbPreflop, GetProbFlop, GetProbTurn, GetProbRiver;
transforms for suits in preflop.
74
Hand2Nr used by GetProbPreflop, GetProbFlop, GetProbTurn, GetProbRiver;
returns for a hand the corresponding numerical value.
BinSearch used by GetProbPreflop, GetProbFlop, GetProbTurn, GetProbRiver;
the binary search for numerical values in ordered mat-files.
CalcTHMOptCardInx used by CalcTHM;
administrates card combinations and corresponding probs.
CalcTHMStrats used by CalcTHM;
calculates exp. Gains and bets for all players;
eventually loads, modifies and saves the file THMStratsFile for strategies;
calls MakeTHM0Decs, CalcFictGainBet, CalcTHMOptGainBet, GetGainBetDistr.
MakeTHM0Decs used by CalcTHMStrats;
calculates the beginner strategies.
CalcFictGainBet used by CalcTHMStrats;
calculates the gain and bet of the fictive players.
CalcTHMOptGainBet used by CalcTHMStrats;
calculates the gain and bet of the optimal players and eventually the corresponding optimal
strategies.
GetGainBetDistr used by CalcTHMStrats;
calculates the gain and bets of the endnodes;
internal function calls GetValPokerHands.
GetValPokerHands used by GetGainBetDistr;
returns the poker values of hands.
CalcSkill used by CalcTHM;
calculates the 4 measures of skill from gains and bets.
75
A2. Description of the parameters of the program THM.
We will only describe the the various parameters of the program. In section A2.1 we just indicate
how the results of the tables in the foregoing sections can be reproduced. In section A2.2 we discuss
some other settings for more detailed and additional output.
A2.1 Reproducing tables
We will describe how to set the parameters in the program in order to obtain the results as descibed
in the various sections.
The main program is the script file THM.m in which the game parameters get their values. All other
parts are functions (including subfunctions). The main function is the file CalcTHM.m which calls all
other functions and generates the main output.
% THM
% last update 2011-12-21
% Program to calculate for THM
% skill, strategies for beginners and optimal players
% and do fictitious play dependent on parameters.
% The main program only sets parameters,
% calculations and output is made with CalcTHM.
% external m-function:
% CalcTHM (results are displayed by this procedure)
clear, clear global
Rpts=1; % >1 for fictitious play
NoOptFictPl=0; % =2 for no calc. optimal & fictive player
% =1 for no calc. for fictive player
% =0 for all calculations
WithBB=1; % =1 if BB always makes a decision in 1st round
SaveCards=0; % =1 creates CardCombFile** (MaxRounds Players) for rep.use
SaveText=0; % =1 creates the file THM.txt of results
% set game parameters
SB=2;
BB=4;
BetVec=[4]; % fixed bet amounts
Players=2; % 2<=Players<=9
ChipVec=100*ones(1,Players); % length =Players !!!
RaiseVec=[]; % fixed raise amounts additional above MinRaise
MaxRounds=2; % max.number of rounds <=4
% game rules depending on MaxRounds:
% MaxRnd 1 2 3 4
76
% Round
% 1 River=4 Prefl=1 Prefl=1 Prefl=1
% 2 - River=4 Flop=2 Flop=2
% 3 - - River=4 Turn=3
% 4 - - - River=4
% DoubleRound: double Bet and Raise from Round>=DoubleRound>=2
NoLimit=0; % =1 allowed (only AllIn)
PotLimit=0; % =1 allowed
MaxRaises=1; % max.number of raises
% set beginners parameters
RandProb=0.1; %0 0.1 0.2 0.3
FactDec=0; % 0.2 0.4 0.6 0.8
GainLabel=1; % =1 for max exp.gain, =0 for max dec.chips
% can be set separately:
FactDecVec=FactDec*ones(1,Players); % FactDecVec >= 0
RandProbDecVec=RandProb*ones(1,Players); % 0<=RandProbDecVec<=1
% set up simulation if no CardCombFile
% dummy values
SimFlop=2; % dummy value, only used if MaxRounds>=3
SimTurn=2; % dummy value, only used if MaxRounds>=4
switch Players+10*MaxRounds
case 12
SimCards=30; % always used
SimRiver=20; % always used
% TotCombs = 3,288,600
case 22
SimCards=30;
SimRiver=20;
% TotCombs = 3,288,600
case 32
SimCards=25;
SimRiver=5;
SimFlop=10;
% TotCombs = 3,795,000
case 42
SimCards=15;
SimRiver=2;
SimFlop=10;
SimTurn=2;
% TotCombs = 589,680
case 13
SimCards=12; % always used
SimRiver=20; % always used
% TotCombs = 1,663,200
case 23
SimCards=12;
SimRiver=20;
% TotCombs = 1,663,200
case 33
SimCards=10;
SimRiver=5;
SimFlop=10;
% TotCombs = 945,000
case 43
SimCards=10;
SimRiver=2;
SimFlop=10;
SimTurn=2;
77
case 14 % not used
case 24 % not used
case 34 % not used
case 44 % not used
case 15 % not used
case 25
SimCards=10;
SimRiver=10;
% TotCombs = 1,134,000
case 35
SimCards=10;
SimRiver=2;
SimFlop=5;
% TotCombs = 1,134,000
case 45
SimCards=10;
SimRiver=2;
SimFlop=5;
SimTurn=2;
otherwise
disp('error in simulation parameters'),pause
end
CalcTHM(SB,BB,BetVec,ChipVec,PotLimit,NoLimit,RaiseVec,...
MaxRaises,MaxRounds,WithBB,GainLabel,FactDecVec,RandProbDecVec,...
SimCards,SimRiver,SimFlop,SimTurn,...
SaveCards,SaveText,NoOptFictPl,Rpts)
Table A2.1 The main script file THM.m.
If the program THM is run with the script in table A2.1 as it is, then another sample like those in table
1.3.5.1 is taken and analyzed.
Text written to the screen can be saved to the text-file THM.txt if we change SaveText = 0 to
SaveText = 1.
At first we describe the effect if we only change the parameter SaveCards = 0 to 1. Then the program
checks whether there is a file CardCombFile22.mat ( MaxRounds=2, Players = 2) present in the
current directory. If not, then a new sample with SimCards=30 and SimRiver=20 is taken (see case
22) and at the end of the program this sample is saved with this name. However, if already such a file
exists then the sample in it is taken. So this gives the possibilty to use the same sample again and
again.
So, check that CardCombFile22.mat is present and change to SaveCards = 1. Then we get exactly the
result of table 1.3.2.6.
Now, with SaveCards = 1, change RandProb = 0.1 to RandProb = 0. Then we get table 1.3.1.14.
After that, with SaveCards = 1 and RandProb = 0, change FactDec = 0 to FactDec = 1. Then we get
table 1.3.3.4.
78
After that, with SaveCards = 1, RandProb = 0 and FactDec = 0, change GainLabel = 1 to GainLabel = 0.
Then we get table 1.3.4.3.
Finally, with SaveCards = 1, RandProb = 0.1, FactDec = 0 and GainLabel = 1, change MaxRaises = 1 to
MaxRaises = 3. Then we get table 2.2.1.1.
So, as far as MaxRounds = 2 and Players = 2, all results of section 1 can be checked exactly since the
file CardCombFile22 is saved and read again if SaveCards = 1.
For other values of (MaxRounds, Players) than (2, 2) other samples are used. With
CardCombFile23.mat present, table 2.3.2.1 can be reproduced: take Players = 3.
With CardCombFile42.mat we can reproduce table 2.4.1: take MaxRounds = 4.
Tables 3.2, 3.3, 4.2, 5.2 and 5.3 use again CardCombFile22. For all those tables set RandProb=0.1. For
table 3.2 (spread-limit) set BetVec = [4 8] and RaiseVec = [8], for table 4.2 set PotLimit = 1 and for
table 5.2 set NoLimit = 1.
A2.2 Settings for different and additional output
In the foregoing section A2.1 always Rpts = 1, NoOptFictPl = 0 and WithBB = 1. Other settings of
these parameters leads to all kind of options.
Changing WithBB = 1 to WithBB = 0 leads to game trees for the simplifying THM-rule that the BB-
player gets not necessarily a move in the first round if MaxRaises >=1.
With Rpts = 1, changing NoOptFictPl = 0 to NoOptFictPl = 1 has two effects:
- the calculations for the fictive players are skipped,
- the strategies for the optimal players are also calculated: if more decisions lead to the same
expected gain, then the decision with the lowest bet is taken. These optimal strategies
(THMOptStratsCell) together with the beginner strategies (THM0StratsCell) are saved in the file
THMStratsFile. This file can be used for further analysis of strategies.
If THM.m is executed again then beginner strategies are not calculated again but read from this file.
This file contains also a variable TotRpts with unchanged value 1.
With Rpts = 1, changing NoOptFictPl = 0 to NoOptFictPl = 2 has also two effects:
- the calculations for the fictive and optimal players are skipped,
79
- for the beginners not only expected gains and bets are calculated but also the complete
simultaneous prob.distribution of gains and bets of all players.
The most complex option is the setting Rpts > 1 and NoOptFictPl = 1. Then a loop Rpt = 1: Rpts is
executed. In each step TotRpts = TotRpts + 1 and is saved in THMStratsFile. As long as Rpt < Rpts the
beginner strategies and the optimal strategies are mixed to new beginner strategies: at each decision
node the weigthed sum of the beginner- and optimal decision with weights Weight and 1 – Weight
respectively. The new beginner strategy is saved in the file THMStratsFile. The variable Weight is
defined in the function file CalcTHMStrats. It has the value Weight = TotRpts / (1+TotRpts), leading to
a kind of fictitious play in behaviour game form (instead of the usual normal game form). This whole,
somewhat difficult, procedure with a repeatedly saved file makes it possible to interrupt the
execution of THM.m at any time without loosing the essential information: if the program is
restarted it continues with the last step. At the stage where expected gains of the beginners are
equal to those of the optimal players, a Nash-equilibrium is reached. This has been applied in section
1.5
Of course options are created easily. E.g. if at each new step we can give beginners the just
calculated optimal strategies by changing Weight = TotRpts / (1+TotRpts) to Weight = 0.
80
A3. The game tree function for THM.
The procedure TestMakeTreeTotTHM is an important test function which only generates the game
tree. It only calls MakeTreeTotTHM (see the list in A1). The procedure works also with a 32-bit
version of MATLAB. Below we display the result for the tree in figure 1.1.1. All these data are used in
the calculations for skill.
Players = 2 - Small Blind = 2 - Big Blind = 4
PotLimit = 0 - NoLimit = 0 - BetVec = 4
MaxRaises = 1 - MaxRounds = 2 - DoubleRound = 2
RaiseVec =[ ]
ChipVec = 100 100
Nodes = 55 Node = 1
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [2 4]
ChipVec: [98 96]
CurPlayer: 1
DecType: ' '
DecChips: 0
CntDecisions: 1
EndState: 0
BackNode: 0
NextNodes: {[2 3 4]}
MinRaise: 4
Raises: 0
Round: 1
Node = 2
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [2 4]
ChipVec: [98 96]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 1
NextNodes: {[]}
MinRaise: 4
Raises: 0
Round: 1
Node = 3
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 4]
ChipVec: [96 96]
CurPlayer: 2
DecType: 'Call'
DecChips: 2
CntDecisions: 2
EndState: 0
BackNode: 1
NextNodes: {[7 8]}
MinRaise: 4
Raises: 0
Round: 1
Node = 4
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 4]
ChipVec: [92 96]
CurPlayer: 2
DecType: 'Raise'
DecChips: 6
CntDecisions: 1
EndState: 0
BackNode: 1
NextNodes: {[5 6]}
MinRaise: 4
Raises: 1
Round: 1
Node = 5
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [8 4]
ChipVec: [92 96]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 4
NextNodes: {[]}
MinRaise: 4
Raises: 1
Round: 1
Node = 6
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 1
DecType: 'Call'
DecChips: 4
CntDecisions: 0
EndState: 0
BackNode: 4
NextNodes: {[11]}
MinRaise: 4
Raises: 1
Round: 1
Node = 7
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 4]
ChipVec: [96 96]
CurPlayer: 1
DecType: 'Check'
DecChips: 0
CntDecisions: 0
EndState: 0
BackNode: 3
NextNodes: {[26]}
MinRaise: 0
Raises: 0
Round: 1
Node = 8
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 8]
ChipVec: [96 92]
CurPlayer: 1
DecType: 'Bet'
DecChips: 4
CntDecisions: 1
EndState: 0
BackNode: 3
NextNodes: {[9 10]}
MinRaise: 4
Raises: 0
Round: 1
Node = 9
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [4 8]
ChipVec: [96 92]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 8
NextNodes: {[]}
MinRaise: 4
Raises: 0
Round: 1
Node = 10
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
Node = 11
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
Node = 12
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
81
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 2
DecType: 'Call'
DecChips: 4
CntDecisions: 0
EndState: 0
BackNode: 8
NextNodes: {[41]}
MinRaise: 4
Raises: 0
Round: 1
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 2
DecType: 'NextR'
DecChips: 0
CntDecisions: 2
EndState: 0
BackNode: 6
NextNodes: {[12 13]}
MinRaise: 0
Raises: 0
Round: 2
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 1
DecType: 'Check'
DecChips: 0
CntDecisions: 1
EndState: 0
BackNode: 11
NextNodes: {[14 15]}
MinRaise: 0
Raises: 0
Round: 2
Node = 13
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 16]
ChipVec: [92 84]
CurPlayer: 1
DecType: 'Bet'
DecChips: 8
CntDecisions: 1
EndState: 0
BackNode: 11
NextNodes: {[16 17 18]}
MinRaise: 8
Raises: 0
Round: 2
Node = 14
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 2
DecType: 'Check'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 12
NextNodes: {[]}
MinRaise: 0
Raises: 0
Round: 2
Node = 15
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 8]
ChipVec: [84 92]
CurPlayer: 2
DecType: 'Bet'
DecChips: 8
CntDecisions: 1
EndState: 0
BackNode: 12
NextNodes: {[21 22 23]}
MinRaise: 8
Raises: 0
Round: 2
Node = 16
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [8 16]
ChipVec: [92 84]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 13
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 17
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 16]
ChipVec: [84 84]
CurPlayer: 2
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 13
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 18
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [24 16]
ChipVec: [76 84]
CurPlayer: 2
DecType: 'Raise'
DecChips: 16
CntDecisions: 1
EndState: 0
BackNode: 13
NextNodes: {[19 20]}
MinRaise: 8
Raises: 1
Round: 2
Node = 19
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [24 16]
ChipVec: [76 84]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 18
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 20
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [24 24]
ChipVec: [76 76]
CurPlayer: 1
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 18
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 21
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [16 8]
ChipVec: [84 92]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 15
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 22
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 16]
ChipVec: [84 84]
CurPlayer: 1
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 15
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 23
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 24]
ChipVec: [84 76]
CurPlayer: 1
DecType: 'Raise'
DecChips: 16
CntDecisions: 1
EndState: 0
BackNode: 15
NextNodes: {[24 25]}
MinRaise: 8
Raises: 1
Round: 2
Node = 24
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [16 24]
ChipVec: [84 76]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 23
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
82
Node = 25
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [24 24]
ChipVec: [76 76]
CurPlayer: 2
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 23
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 26
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 4]
ChipVec: [96 96]
CurPlayer: 2
DecType: 'NextR'
DecChips: 0
CntDecisions: 2
EndState: 0
BackNode: 7
NextNodes: {[27 28]}
MinRaise: 0
Raises: 0
Round: 2
Node = 27
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 4]
ChipVec: [96 96]
CurPlayer: 1
DecType: 'Check'
DecChips: 0
CntDecisions: 1
EndState: 0
BackNode: 26
NextNodes: {[29 30]}
MinRaise: 0
Raises: 0
Round: 2
Node = 28
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 12]
ChipVec: [96 88]
CurPlayer: 1
DecType: 'Bet'
DecChips: 8
CntDecisions: 1
EndState: 0
BackNode: 26
NextNodes: {[31 32 33]}
MinRaise: 8
Raises: 0
Round: 2
Node = 29
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [4 4]
ChipVec: [96 96]
CurPlayer: 2
DecType: 'Check'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 27
NextNodes: {[]}
MinRaise: 0
Raises: 0
Round: 2
Node = 30
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [12 4]
ChipVec: [88 96]
CurPlayer: 2
DecType: 'Bet'
DecChips: 8
CntDecisions: 1
EndState: 0
BackNode: 27
NextNodes: {[36 37 38]}
MinRaise: 8
Raises: 0
Round: 2
Node = 31
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [4 12]
ChipVec: [96 88]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 28
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 32
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [12 12]
ChipVec: [88 88]
CurPlayer: 2
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 28
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 33
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [20 12]
ChipVec: [80 88]
CurPlayer: 2
DecType: 'Raise'
DecChips: 16
CntDecisions: 1
EndState: 0
BackNode: 28
NextNodes: {[34 35]}
MinRaise: 8
Raises: 1
Round: 2
Node = 34
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [20 12]
ChipVec: [80 88]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 33
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 35
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [20 20]
ChipVec: [80 80]
CurPlayer: 1
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 33
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 36
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [12 4]
ChipVec: [88 96]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 30
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 37
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [12 12]
ChipVec: [88 88]
CurPlayer: 1
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 30
Node = 38
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [12 20]
ChipVec: [88 80]
CurPlayer: 1
DecType: 'Raise'
DecChips: 16
CntDecisions: 1
EndState: 0
BackNode: 30
Node = 39
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [12 20]
ChipVec: [88 80]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 38
83
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
NextNodes: {[39 40]}
MinRaise: 8
Raises: 1
Round: 2
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 40
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [20 20]
ChipVec: [80 80]
CurPlayer: 2
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 38
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 41
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 2
DecType: 'NextR'
DecChips: 0
CntDecisions: 2
EndState: 0
BackNode: 10
NextNodes: {[42 43]}
MinRaise: 0
Raises: 0
Round: 2
Node = 42
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 1
DecType: 'Check'
DecChips: 0
CntDecisions: 1
EndState: 0
BackNode: 41
NextNodes: {[44 45]}
MinRaise: 0
Raises: 0
Round: 2
Node = 43
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 16]
ChipVec: [92 84]
CurPlayer: 1
DecType: 'Bet'
DecChips: 8
CntDecisions: 1
EndState: 0
BackNode: 41
NextNodes: {[46 47 48]}
MinRaise: 8
Raises: 0
Round: 2
Node = 44
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [8 8]
ChipVec: [92 92]
CurPlayer: 2
DecType: 'Check'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 42
NextNodes: {[]}
MinRaise: 0
Raises: 0
Round: 2
Node = 45
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 8]
ChipVec: [84 92]
CurPlayer: 2
DecType: 'Bet'
DecChips: 8
CntDecisions: 1
EndState: 0
BackNode: 42
NextNodes: {[51 52 53]}
MinRaise: 8
Raises: 0
Round: 2
Node = 46
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [8 16]
ChipVec: [92 84]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 43
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 47
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 16]
ChipVec: [84 84]
CurPlayer: 2
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 43
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 48
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [24 16]
ChipVec: [76 84]
CurPlayer: 2
DecType: 'Raise'
DecChips: 16
CntDecisions: 1
EndState: 0
BackNode: 43
NextNodes: {[49 50]}
MinRaise: 8
Raises: 1
Round: 2
Node = 49
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [24 16]
ChipVec: [76 84]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 48
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 50
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [24 24]
ChipVec: [76 76]
CurPlayer: 1
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 48
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 51
FoldVec: [0 1]
AllinVec: [0 0]
PlayVec: [1 0]
PotVec: [16 8]
ChipVec: [84 92]
CurPlayer: 1
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 45
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
Node = 52
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 16]
Node = 53
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [16 24]
Node = 54
FoldVec: [1 0]
AllinVec: [0 0]
PlayVec: [0 1]
PotVec: [16 24]
84
ChipVec: [84 84]
CurPlayer: 1
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 45
NextNodes: {[]}
MinRaise: 8
Raises: 0
Round: 2
ChipVec: [84 76]
CurPlayer: 1
DecType: 'Raise'
DecChips: 16
CntDecisions: 1
EndState: 0
BackNode: 45
NextNodes: {[54 55]}
MinRaise: 8
Raises: 1
Round: 2
ChipVec: [84 76]
CurPlayer: 2
DecType: 'Fold'
DecChips: 0
CntDecisions: 0
EndState: 1
BackNode: 53
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
Node = 55
FoldVec: [0 0]
AllinVec: [0 0]
PlayVec: [1 1]
PotVec: [24 24]
ChipVec: [76 76]
CurPlayer: 2
DecType: 'Call'
DecChips: 8
CntDecisions: 0
EndState: 1
BackNode: 53
NextNodes: {[]}
MinRaise: 8
Raises: 1
Round: 2
85
Cash and tournament poker: games of skill?
Report commissioned by W.M.C. van den Berg, examining magistrate in charge of criminal matters in
the district of Amsterdam.
Prof. Peter Borm
Prof. Ben van der Genugten
Department of Econometrics and Operations Research
Tilburg University
Tilburg, September 2009
86
Table of contents
1. Introduction.………………………………………………………………………………. 4
2. Conclusions.……………………………………………………………………………..... 6
3. Dutch case law………………………………………………………………………. 7
3.1. State of affairs regarding cash poker…………………………………….… 7
3.2. Management games……………………………………………………………….... 8
3.3. Review and skill threshold level………………………………………………... 9
4. Method of relative skill..................………………………………………………... 9
4.1. General…………………………………………………………………………. . 10
4.2. Single-player games……………………………………………………………... 11
4.3. Multiple-player games…………………………………………………………….... 12
4.4. Theoretical and practical validation ……………………………………….… 12
4.5. Tournaments..…………………………………………………………………. 13
5. Poker…………………………………………………………………………………….. 14
5.1. Cash poker: general…………………………………………………………….. 14
5.2. Cash poker: a game of skill......…………………………………………….. 16
5.3. A skill analysis of tournament poker………………………………............. 16
References…………………………………………………………………………………. 22
87
Authors’ note
The original 2009 Dutch version of this report also contained a CD-ROM with 15 appendices to
facilitate its practical use at the time. Most of these appendices are listed in the references but are
unfortunately only available in Dutch. In any case they are not essential for understanding the main
line of argument of the report. In order not to deviate too much from the original version however
we have opted to maintain all references to the appendices in the current English version.
88
1. Introduction
The direct reason for this report was the request by Mr. W.M.C. van den Berg, examining magistrate
in charge of criminal matters in the district of Amsterdam, to answer the following question as part
of the BKB case (Blaas, Kurver and Blaas):
Why, in your view, should poker be treated as a game of skill, rather than as a game of chance?
First we provide a brief outline of our expertise. Ben van der Genugten is professor of Probability and
Statistics and Peter Borm is professor of Mathematics and Game Theory. We both work at the
Department of Econometrics and Operations Research of Tilburg University. Since 1990 we have
acted as expert witnesses in several court cases concerning the distinction between games of skill
and games of chance in the Netherlands and Austria. Several projects were performed jointly with
Marcel Das, professor of Econometrics and Data Collection and director of CentERdata, a research
institute affiliated with Tilburg University. Professor Das has reviewed a draft version of the current
report. We have also studied this topic intensively at a more theoretical level, through scientific
publications as well as through the supervision of PhD and Master students.
The current report builds on the conceptual framework of the general method of relative skill as
introduced in Van der Genugten & Borm (1994a), and explicitly draws on the simplifications and
refinements that this method has undergone over the past 15 years through both practical
experience and theoretical deliberations. In essence it has not changed, however; the method
provides a tool for the objective and consistent classification of games in terms of relative skill.
The classification method of relative skill has been developed and elaborated to serve Dutch law as
set forth in Article 1 of the Dutch Gaming Act: a game of chance is understood as
“an opportunity to compete for prizes or premiums, if the indication of the winners is subject to any form of chance, on which the participants generally cannot exercise any predominant influence.” In this description, the term opportunity stands for game in the widest sense of the word. Games in which money does not play a role are not relevant to the law on games of chance. We will therefore consistently assume that prizes and premiums are awarded in the form of money, and that the sums of money depend on the results of the game. The legal definition of a game of chance also alludes to chance elements. Apparently, to qualify as a game of chance, the game must contain some chance element that designates one possibility among a series of possibilities as the winning one, while this element generally cannot be influenced by the participants in any meaningful way. Such an element is usually referred to as a chance element or an uncertainty element, but often simply as “chance”. A recurrent theme in relevant verdicts is moreover the idea that a game is more skillful if the players can improve their
89
ability through study or through frequent practice. The game is then less likely to be qualified as a game of chance. We therefore designate a game as a game of skill if it is not a game of chance. Every game involving money is therefore either a game of chance or a game of skill.
Our general method regarding the level of relative skill of a game produces a number between 0 and
1, whereby the value of 0 corresponds with a pure game of chance (skill plays no role whatsoever)
and the value of 1 denotes a pure game of skill (in which chance plays no role whatsoever). This
specifically allows for a comparison and classification of games in terms of skill level, and hence to
determine the minimally required skill level (the skill threshold level), which will determine the
game’s classification as a game of skill.
In this report we choose to present our method and earlier conclusions in a fairly compact manner.
For a more extensive and detailed description and analysis, in several places we refer to the
appendixes on the enclosed CD-ROM containing integral versions of some of our earlier reports and
publications. In the text we therefore concentrate specifically on the main lines of argument leading
up to our earlier conclusions that a regular cash poker variant like Texas Hold‘em should be
considered a game of skill.
This report furthermore focuses on tournament poker, as this is the most frequently occurring form
of poker in the BKB case. It is not at all obvious to us that any tournament version of a game of skill
will automatically qualify as a game of skill as well. Accordingly, we have conducted a new and
specific analysis for a tournament model that is suitable both for the specific tournament involved in
the BKB case, and for the tournaments held regularly at Holland Casino. Given the time restrictions
under which this report was produced, it is a non-complex tournament model that should be viewed
as a reasonable, initial approach, which can be elaborated further through additional research. For
this reason we can only draw conclusions from this analysis with due caution. In paragraph 5 we
again keep to the main lines of argument. One of the appendixes on the CD-ROM contains a detailed
overview of the analyses and computer simulations of tournament poker, performed specifically for
this report.
Regarding poker variants, as far as we know only the fixed-limit cash poker variant has been assessed
under the Gaming Act. This is the first time that tournament poker is under assessment. One typical
feature of a tournament is that the prizes are awarded on the basis of an ultimate ranking of
participants, which is determined by the game results (generally in the form of tokens or points)
achieved during the rounds that make up the tournament. In this respect tournaments can well be
compared to the so-called management games, which have been legally assessed previously with a
view to the Gaming Act. In this study we were therefore able to benefit from our previous experience
with management games. It moreover turns out that tournament poker and management games are
comparable not only regarding the issue at hand, but also regarding the essence of our conclusions.
90
The structure of the remainder of this report is as follows. Paragraph 2 presents our main conclusions. Paragraph 3 outlines existing Dutch case law regarding the Gaming Act, with particular attention for the state of affairs concerning poker. The situation concerning management games is also considered explicitly. Paragraph 4 describes the method of relative skill for single-player and multiple-player games, with reference to the scientific and practical validation of the method. The paragraph furthermore discusses on a general level how the relative skill of a cash game compares to the relative skill of this same game in a tournament context. Paragraph 5 sketches the considerations based on previous research to qualify common multiple-player cash variants of poker, such as Texas Hold’em, as a game of skill. Here we furthermore present a separate, quantitative analysis of tournament poker.
91
2. Conclusions
Our conclusions issue from the method of relative skill. This method in principle enables a
classification of a large number of practical games in terms of their relative skill levels. However, it
does not answer the question exactly where to draw the skill threshold level. We have argued
previously that, with a view to consistency in case law, this threshold should be located between 0.1
and 0.3.
The classification of a game as one of chance or one of skill cannot automatically be made to apply to tournament versions of the same game. In analyzing the relative skill of tournaments, the number of participants in relation to the number of tournament rounds and the prize structure co-determine the outcome. Cash poker variants of Texas Hold’em should be classified as a game of skill, as their level of relative skill exceeds the 0.3 threshold. Our analysis of the Texas Hold’em tournament in the BKB case, explicitly taking into account the number of participants, the number of tournament rounds and the prize structure, does not give reason to classify this tournament version as a game of chance. The same conclusion applies to Texas Hold’em tournaments with around 50 participants, as organized for example by Holland Casino.
92
3. Dutch case law
For a detailed overview of the drafting of the Gaming Act, and our commentary on the rulings and
verdicts regarding specific games until approximately the year 2000, please see Appendix 1 on the
CD-ROM.
3.1. The state of affairs regarding cash poker
Fixed-limit Texas Hold’em, along with other cash variants such as Five card draw poker, Omaha
Hold’em and Seven card stud poker, has been the subject of court cases. The first time that a
multiple-player game was made subject to the Gaming Act occurred in the district court of
Amsterdam, on 7 May 1996. The judge exonerated the organizers for a lack of evidence. As expert
witnesses in this trial, we presented our assessment system based on relative skill and recommended
that the poker variants be classified as games of skill. We emphatically pointed out the wide range of
strategic aspects involved in such multiple-player games, particularly where it concerns games that
involve incomplete information, as in this instance. This advice was followed by the court: “It is after
all conceivable that starting players quickly develop a certain measure of skill with which, combined
with other variables such as the strategy of other players, they can develop a personal strategy to
such an extent that it cannot be ruled out that they can generally exercise predominant influence
over the role of chance”. The court also adopted the view that not only chance elements such as the
cards one is dealt are essential to the poker variants, but also and especially the application of
randomized strategies by the players: “The in itself correct observation by the public prosecutor that
in the poker variants offered by the accused, a maximum of 41 to 47 (of the 52) cards may be
unknown at the end of the fifth round (just before the showdown), does not contradict the
foregoing. It does not force the conclusion that the participants cannot in general exercise
predominant influence over the chances of winning.”
This line of argument stands in stark contrast to that followed by the Court of Appeal and the
Supreme Court (in which we were no longer directly involved). The Supreme Court followed the
argumentation of the Court of Appeal. The Court of Appeal based its opinion on the argument that
the game rules directly determine that the impact of chance in (these variants of) the game of poker
is such that the players cannot have any influence. The game could only be considered a game of skill
if it can be demonstrated that this impact can be “overcome” through the use of some form of
probabilistic calculations. With this argument, the Court of Appeal ruled that the impact of chance
cannot be overcome. Here, the court followed the advice of an expert witness who stated that the
average player is not prepared to develop his skill as he only plays for the purpose of relaxation. This
is not a sound argument. What exactly does “overcome” mean? To overcome what exactly? The
argument presumes a type of player who simply performs a random lottery over all his possible
93
actions at any possible decision moment. Yet no player will do so, also not in card games like
blackjack and bridge: given the structure of the game he will always pursue some form of strategy
with considerably better game results than by using a random lottery. In that sense alone he or she
more or less overcomes the impact of pure chance. To overcome should be replaced by “to do
significantly better than a beginner”. Here, the beginner is definitely another person than the above
described (non-existent) player using random lotteries all the time. The Court of Appeal relies on the
– to put it cautiously: -- non-verifiable judgment of an expert witness. We hold a different view. In
practice, poker players frequently play with a variety of other players. Our observation is that they
always strive to achieve a good result. This is quite typical, incidentally, for many multiple-player card
games, regardless of whether cash or game points are at stake. Finally, the Court’s general
deliberations are wholly directed at poker games, without considering the implications of any
comparative application of these deliberations to other card games such as bridge. In fact, it would
result in the erroneous conclusion that bridge is a game of chance rather than a game of skill.
The district court’s argumentation regarding the wealth of strategic aspects of poker variants is
annulled by subsequent imprecise deliberations by higher judges, who were possibly also
insufficiently cognizant of the difference in game characteristics between single-player games and
multiple-player games.
3.2. Management games
A recent ruling in proceedings concerns the so-called ‘management games’ of Competitie manager
and Grand Prix manager GPM (2 February 2005, no.105364, included as Appendix 2 on the CD-ROM).
These management games are operated via the internet and are the subject of a detailed analysis in
Van der Genugten, Borm & Dreef (2004), included as Appendix 3 on the CD-ROM, and in Van der
Genugten, Borm & Dreef (2005), Appendix 4 on the CD-ROM.
A typical feature of management games is that the final prize structure is determined by the final
ranking of all participants as determined by the game results (in points) achieved over the course of
the game rounds. In terms of structure, these games are thus comparable to tournaments. Since in
this case, the court basically adopted the arguments and conclusions of our reports in full, we shall
briefly discuss a few relevant details with regard to GPM here.
The goal of GPM is to compile a Formula 1 team in terms of car parts and staffing that will achieve
the best performance in a simulated season of Formula 1 competitions. It was possible to quantify
the role of chance in GPM on the basis of extensive data collection combined with statistical
techniques. In this way, the relative skill level was determined for various GPM variants that only
94
differ with respect to the actual prize structure. Our first report concluded that, given a fairly
horizontal or gradually increasing prize structure (in which the prizes are not only awarded to a small
number of highly placed players in the final ranking), the relative skill level of GPM can be set at
around 0.3, so that the final verdict on the game is: game of skill. For less gradual prize structures (as
occurred in actual practice), the relative skill level came to a maximum of 0.1, so that the final verdict
on the game is: game of chance. Following the recommendations in our second report, the prize
structure of GPM was modified to create a more gradual structure, putting the relative skill level at
round 0.3 and assuring its legal classification as a game of skill.
How can these conclusions be explained in qualitative terms? The scores in GPM vary widely, but
there is a relatively large group of players who achieve scores that fall only slightly short of those
achieved by advanced players in top ranking positions. The small difference in scores amply remains
within the margin generated through chance. So if the prize structure is restricted to the players in
the top ranking segment, then chance plays an important role in the awarding of prizes. For if the
game is played repeatedly, then this relatively large group of players will often wind up as winners,
while an advanced player will not. This implies that the advanced players achieve a low game result
on average, with a relatively low learning effect as a result. In case of a more gradual or horizontal
prize structure, the results achieved by an advanced player will vary across repeated plays, but
distributed evenly at a high level. The learning effect will thus be greater in case of a gradual prize
structure.
3.3. Review and skill threshold level
If we combine the data regarding management games with previous legal rulings on explicit games of
chance and skill slot machines (cf. Van der Genugten, 1997a), then the designation of the skill
threshold level would reasonably be located between 0.1 and 0.3. Games with a level of relative skill
above 0.3 should in any case be classified as game of skill, and games with a relative skill level below
0.1 as games of chance. This method is in any case consistent with court rulings so far. The only
exception to this is the Supreme Court ruling following the Court of Appeal with regard to a number
of fixed-limit cash poker variants. It would have been preferable if they had followed the well-argued
judgment of the district court.
95
4. Method of relative skill
The method of relative skill has been described extensively and meticulously in the general
publications by Hilbers, Hendrickx, Borm & Van der Genugten (2008), Dreef, Borm & Van der
Genugten (2004a, 2004b), Van der Genugten, Das & Borm (2001), Borm & Van der Genugten (2001),
Borm & Van der Genugten (1998), Van der Genugten (1997a), Van der Genugten & Borm
(1996a,1996b,1996c), and Van der Genugten & Borm (1994a).
Our method has been developed particularly with a view to studying so-called strategy games with
monetary rewards. Strategy games are about mental dexterity: the ability to make sensible decisions
systematically, which boils down to choosing a comprehensive game plan or strategy. A strategy thus
does not correspond to a general game attitude or approach, but it provides a detailed specification
of actions to be taken at any conceivable decision point in the game.
Strategy games can be classified according to a number of characteristic game features. A detailed
classification is provided in Van der Genugten and Borm (2005), as a contribution to a book written
for a legally versed audience. This contribution is included in full as Appendix 5 on the CD-ROM.
Game features are: the presence or absence of chance elements, the degree of complexity (e.g.
frequent or few decision moments), complete or incomplete information, equal or unequal
information among different players, and of course the number of players. As in more advanced card
games such as bridge, poker games typically involve multiple players, they contain chance elements,
they are highly complex, the information is incomplete, and there is an information disparity among
the players. Perhaps needless to say, but when referring to poker games we mean multiple-player
games such as Texas Hold’em, Seven card stud poker or Five card draw poker, and not single-player
games such as Caribbean stud poker or American poker.
In Paragraph 4.1 we describe in general terms the method of relative skill for strategy games with
chance elements. In Paragraph 4.2 we discuss the elaboration of this method for single-player
games. The main characteristic of a single-player game is that, although more players may be
engaged in playing the same game, any one player’s game result depends only on his or her own
chosen strategy, and not on any actions taken by other participants. Roulette, Golden Ten and
blackjack are typical examples. In multiple-player games such as Texas Hold’em and bridge, the result
achieved by each player typically depends partly on the decisions taken by other players. This makes
such games intrinsically much more complex. This interaction between players must be incorporated
in the method of relative skill for multiple-players games in an adequate and consistent manner.
Paragraph 4.3 explains how this is done. Paragraph 4.4 offers a practical and theoretical validation of
this method, drawing on relevant literature. Paragraph 4.5 offers a more general discussion of the
relationship between the relative skill of a strategy game and tournament versions of the same
game.
96
4.1. General
Classification as a “game of chance” or a “game of skill’ depends on the relative skill of a game, as
determined by weighing two effects on the game result:
The learning effect (LE), due to the skill elements involved in the game, The random effect (RE), attributable to the chance elements involved in the game.
By “game result” we mean, in this context, the (probabilistic) expected gain: that is, the average gain
over an, in principle, infinite number of repeated game plays. This concretely concerns an amount of
money that is itself no longer dependent on chance. For games in which the stake amount is not an
intrinsic component of the strategy or game plan but more of a random choice in advance (as in
Roulette), we standardize the expected gains by dividing by the expected stake amount. Obviously,
the classification as game of chance or game of skill does not depend on the exact value for the game
result or on whether the game result is positive or negative for a certain strategy. We have opted for
the term learning effect since skill is achieved through study or experience. The random effect can
then be associated with any further improvement in terms of game result that could be achieved, if
the effects of the random factors were known in advance. The latter is of course a fictional situation.
The random effect offers a very elucidating means of measuring the variation in game results, solely
attributable to the chance elements involved in the game.
The relative skill of a game is expressed as a number, say S (“Skill”), which is large (maximum of 1) if the learning effect is dominant, and is small (minimum of 0) if the random effect dominates. A simple formula to express this, incorporating the terms of the underlying learning effect and random effect, is as follows:
S = LE / ( LE + RE). Thus, no learning effect (LE =0) yields S = 0 and no random effect (RE = 0) yields S = 1. Every game will always have a relative skill level of between 0 and 1, so that all games can be ranked in terms of their relative skill. 4.2. Single-player games The concepts of LE and RE have been made operational for single-player games by distinguishing three types of players: 1) the beginner, who plays in naïve manner, knows and understands the rules of the game but
lacks experience in actual play, with game result R(0), 2) the advanced player, who has mastered every aspect of the game, with game result R(m),
and
97
3) the fictional (advanced) player, an advanced player in the theoretical situation that he knows the outcome of all chance elements during play (but cannot influence these), with game result R(f).
A specific player can learn the game and can become increasingly proficient at it, until he reaches the level of an advanced player. That is why the learning effect equals:
LE = R(m) – R(0). The extra result that the fictional player can attain over the result of the advanced players is solely attributable to the fictional knowledge of the outcome of the random factors. That is why the random effect equals:
RE = R(f) – R(m). For many casino games (such as Roulette and Golden Ten), the fictional player’s game result can simply be equated with the maximum result possible per game. For games with a small chance of a big win, only the fictional player will attain a high game result, so that the random effect RE is large and the relative skill, therefore, is small.
We wish to point out that the precise specification of the beginner’s strategy is a freely definable
parameter in our model. The skill level analysis can therefore be performed with any specification of
the beginner’s strategy. This means, specifically, that the fairly vague discussion about possible skill
elements in a game can be reduced to the more concrete question of what population of beginners
should function as reference point in determining the learning effect and the random effect.
When applying relative skill to specific games, it is necessary in each instance to adequately
characterize the three types of players. Experience shows that the application of the method of
relative skill to specific games requires a tailor-made analysis of the learning effect and random
effect, based on the particular features of the game. Also in the event of analytical restrictions, the
conceptual framework outlined above in any case provides a way of consistently comparing the
learning effect and random effect.
To get a better feeling for the method of relative skill for single-player games we refer to Appendixes
6 and 7 on the CD-ROM, in which the skill level analysis is illustrated with reference to the well-
known casino game blackjack and the less known but simpler game Spiel 21.
98
4.3 Multiple-player games
In multiple-player games, the different player roles are generally asymmetrical. For poker, think for
instance of each player’s specific position at the table. We shall therefore start by examining a single
player in a specific player role, in the midst of several beginners as opponents.
We will portray this player as three different types: as a beginner, as an advanced player, and as a
fictional player, but each time playing against a group of beginners. This yields a specific game result
for each type. Both the advanced and the fictional type are presumed to be aware of the beginner’s
strategy for each player role. They will apply this knowledge to determine their own strategy.
We will then consider all the player roles for each of these three types of player, enabling us to
calculate the average game result across the player roles for each type of player: R(0) for the
beginner type, R(m) for the advanced type, and R(f) for the fictional type. The resulting figures serve
as input for the learning effect LE and the random effect RE, with which the relative skill S can be
determined.
It is a typical feature of card games that players do not know each other’s cards in many phases of
the game. In that sense they have incomplete and different information. This also applies to many
other multiple-player games. It can therefore be to one’s benefit to vary the decisions one takes in a
certain game situation. This can formally be portrayed as a randomized choice between certain
possible actions with individually chosen probabilities. This internal decision-making mechanism, not
perceptible to other players, we describe as randomized. A strategy containing such randomized
decision-making mechanisms is accordingly referred to as a randomized strategy. For a closer
examination of the role and meaning of randomization, we refer to the discussion on bluffing and
sandbagging in the specific context of multiple-player poker variants in Paragraph 5. Players that
randomize in this internal way add their own chance elements to the game. In our analysis we
therefore assume that the fictional player also knows the outcome of these internal lottery
mechanisms.
To determine the skill level, the point of departure is again the beginner’s strategy. This strategy
must be established for every player role and may in principle be randomized. This specification can
again be considered as a freely definable parameter in our model. The analytical restrictions for
concrete multiple-player games are much larger than for single-player games. In any case, the
methodical approach again offers a useful conceptual framework for the consistent weighing of the
learning effect and random effect.
For a more detailed conceptual discussion on relative skill in multiple-player games we refer to
Hilbers, Hendrickx, Borm & Van der Genugten (2008), included on the CD-ROM as Appendix 8.
99
4.4. Theoretical and practical validation
First developed in Van der Genugten and Borm (1994a) and subsequently refined over the next 15 years, the method has amply proved its scientific value and practical applicability as an objective and consistent means of classifying games based on their relative skill.
Scientific output
Relative skill is the subject of the dissertations “Skill and strategy in games” (Dreef, 2005) and
“Golden ten and related trajectory games” (De Vos, 1997), which were both successfully defended at
Tilburg University. The method is furthermore addressed in two recent graduation theses, by Maaten
(2009) and by Hilbers (2007). Scientific reports and publications relating to the method of relative
skill are: Van der Genugten (2008, 2003, 1997c, 1993), Hilbers, Hendrickx, Borm & Van der Genugten
(2008), Dreef & Borm (2006), Dreef, Borm & Van der Genugten (2004a, 2004b, 2003), Borm & Van
der Genugten (2001, 1998), Van der Genugten, Das & Borm (2001) and Van der Genugten & Borm
(1996c).
Practical applications
The method of relative skill has been applied to a large number of actual games, on the request of
the public prosecutor, defense lawyers and/or the examining magistrate, as part of judicial
proceedings: Blackjack: Van der Genugten & Das (1999b), Fruitcard, Carribean Studpoker en
American poker 2: Van der Genugten & Das (1999a, 1999c, 1998), ROTA-roulette: Van der Genugten,
De Vos & Das (1999), Dromus-24: Van der Genugten & De Vos (1998), Jokeren, Jahtzee en
Eurobsgame: Van der Genugten (1997b), Concard Aces: Van der Genugten (1996), Piramidesystemen:
Van der Genugten & Borm (1994b), Spot the balll: Van der Genugten & Brekelmans (2006), Random
Flashback: Van der Genugten & Das (2005), Managementspelen: Van der Genugten, Borm & Dreef
(2005, 2004), and, finally, Poker: Borm & Van der Genugten (2005) and Van der Genugten, Borm &
Grossmann (1997).
100
4.5. Tournaments
So far, tournaments have not formally been the subject of Dutch cases of law. There has been a case
of Swedish case of law in which a tournament version of Texas Hold'em was classified as a game of
skill. We quote the key formulation: “…in this version of poker it is the players’ actual aptitude and
analytical capabilities that are the winning point and not the actual luck factor”. For further
information we refer to Appendix 9 on the CD-ROM.
Without performing an explicit skill level analysis, in this paragraph we wish to already discuss a few
general aspects that play a role in the classification of tournaments. Virtually every game can be
organized as a tournament, with the actual game functioning as separate game rounds. There are no
grounds, however, for equating “game of chance (or of skill) game round” with “game of chance (or
of skill) tournament”. The classification of a tournament very much depends on its design.
Take, as an example, a game in which each player’s goal is simply to win. The winning probability of the advanced player is greater than the winning probability of a beginner. The qualification as a game of chance or of skill also depends on the winning probability of the fictional player. These probabilities determine the relative skill S of a single play of the game. Now suppose this game were to be played in a tournament form, in which all players consistently participate in each game round, and all the game rounds are independent of each other. The tournament winner is the player who wins the largest number of game rounds, and only this person is awarded a prize. In this tournament form, the (tournament) winning probability of the advanced player increases in tandem with the number of game rounds. The game may continue to be a game of chance if the number of game rounds is limited. However, if the tournament consists of a large number of game rounds, then the tournament prize ultimately will be won by the advanced player, and the tournament will be (almost purely) a game of skill. This always applies, regardless of how small the difference between the winning probabilities per tournament round of the advanced player and the beginner. Let us now adapt the rules of the tournament such that, after a fixed number of rounds, only those players that have won a certain number of the game rounds may progress to the next round. This introduces the possibility that an advanced player will be eliminated after a certain number of game rounds, and thus does not win a prize. This person’s probability of winning the tournament is therefore limited, even if the tournament consists of a very large number of game rounds. This design of the tournament means that the level of relative skill remains limited with an increasing number of game rounds. Finally, we return to the original tournament design (no eliminations), but we introduce a prize structure that corresponds strictly with the number of game rounds won. Compared to ‘only the tournament winner gets the prize’, for the same number of tournament rounds the relative skill of this new type of tournament will be higher. After all, the advanced player will receive a prize more often than the beginner, even if he does not end up as the tournament winner. For a game in tournament form, the number of game rounds and the prize structure thus play an important role. In fact we already encountered this phenomenon in Paragraph 3.2, with respect to management games.
101
In practice, game rounds are usually not independent of each other. Maximizing the result per game
and achieving the highest possible final ranking in a tournament version of the same game are
different goals, then. This implies, in particular, that good or advanced strategic behavior in a game
and in a tournament version of the same game can be an intrinsically different matter, and hence
that a game and its tournament version can differ in terms of relative skill. A straightforward example
of this is provided in Maaten (2009), showing that the relative skill levels of a two-player cash poker
game and a tournament version (aimed at winning more tokens than the opposite player and a fixed
sum for the one who wins most tokens) can differ considerably.
102
103
5. Poker
5.1. Cash poker: general
Among the more common multiple-player cash poker variants such as Seven card stud, Five card
draw poker and Texas Hold’em, the latter game is most popular in the Netherlands. Each game has
different betting structures: fixed-limit, pot-limit or no-limit.
In earlier, general publications such as Van der Genugten (2008, Appendix 10 on the CD-ROM) and
Van der Genugten & Borm (2005, Appendix 5 on the CD-ROM), it was shown how skill in poker
variants is closely tied to practical skills regarding the calculation of probabilities, which play a role in
evaluating the quality of one’s own cards in combination with information from the flop, or in
evaluating the quality of other players’ cards. This calculation of probabilities is described in the
many books written about cash poker, discussing the various rules of thumb to determine one’s
expected gains. The quality of these calculations can be appraised through computer simulations.
The Pokerstove program, included as Appendix 11 on the CD-ROM, is a good example.
Besides skills in terms of evaluating one’s options through calculations of probabilities, more
psychological techniques such as bluffing (bidding on a bad hand, instead of folding) and sandbagging
(not bidding on a good hand, but going along or checking) form an essential part of good game
strategy. Both techniques aim to make the best possible use of the fact that all of the players,
without exception, are in a situation of incomplete information (nobody knows another player’s
cards); this is something that inexperienced poker (but also bridge) players tend to realize
insufficiently. The best way to apply these techniques can, in theory, be calculated mathematically.
Naturally, players should not always bluff or use a sandbagging type of strategy, as this would
neutralize the uncertainty factor. This leads to randomization, in quite a natural manner: a good
bluffing or sandbagging strategy uses an internal randomization mechanism. As noted above, the
odds to apply here (how often should I bluff with this type of hand?) can be exactly calculated in
principle, but players can also develop a good feel for this simply through frequent play. In that sense
they learn to play the game, and skill is acquired. There is an extensive literature on this subject.
There is even special software to help develop such strategies through self-study; see for instance
the computer program “Turbo Texas Hold‘em for Windows”, as described in Wilson (2005).
To characterize a beginner’s strategy, it seems reasonable to assume that a beginner does not apply
the techniques described in various poker handbooks. After all, studying and mastering these
techniques requires a considerable effort, which the beginner has yet to make. An advanced player
can of course be assumed to have made this effort. In accordance with this line of reasoning, the
typical ingredients of a beginner’s strategy would be: to stay in the game for too long with relatively
poor cards (no or wrong estimation of probabilities), and not (successfully) applying the techniques
104
of bluffing and sandbagging (too little game experience). The general import of these ingredients has
been confirmed to some extent by the results of the Texas Hold’em experiment, as described in
Maaten, Borm, Van der Genugten & Hendrickx (2008), included as Appendix 12 on the CD-ROM. This
experiment clearly demonstrated that many beginners in any case wish to stay in the game until the
flop, that they do not use sandbagging techniques, and that they attempt to bluff sporadically,
without careful timing. In our skill level analysis concerning cash poker, as reported in various
publications, these ingredients will always, in one way or another, form the foundation for the
choices made in a beginner’s strategy, which, as argued before, serves as reference point for the
determination of relative skill.
We conclude this paragraph with a general comment on the possible difference in relative skill
between fixed-limit, pot-limit and no-limit variants of multiple-player poker. The results by Hilbers
(2007) and Hilbers, Hendrickx, Borm & Van der Genugten (2009) concerning the variation in the
height of the possible fixed-limit bidding from low to high, indicate that the difference in relative skill
between the three variants is not significant. Restricting the skill level analysis to fixed-limit variants
thus appears justified.
5.2. Cash poker: a game of skill
The report entitled “Poker: ein Geschicklichkeitsspiel!” (Borm & Van der Genugten, 2005, Appendix
13 on the CD-ROM) extensively argues that, with a view to consistent jurisprudence, common
multiple-player cash poker variants such as Texas Hold’em should be classified as games of skill, on
the basis of a relative skill level analysis. This conclusion was based on both an exact quantitative
analysis of relatively simple stylized poker variants that did however incorporate key skill aspects of
real poker variants, and on a more qualitative analysis based on a comparison of so-called skill
indicators.
More recent studies into complex poker variants that more closely approximate the real poker game,
by means of straight poker in Hilbers (2007) and Hilbers, Hendrickx, Borm & Van der Genugten
(2009), once again confirm this conclusion. Depending on the exact specifications of the game
(number of players, bet heights, number of bidding rounds, the rake, choice of beginner’s strategy),
the relative skill level varies consistently between 0.3 and 0.5. A further confirmation that the skill
level is at this level is provided by the simulation results regarding Texas Hold’em, with two players
and two phases (“pre-flop” directly followed by “river”); see Appendix 14 on the CD-ROM.
In realistic poker variants, there is at least one additional source that demonstrates skill, compared to
straight poker or the just-indicated two-phase poker. The point is that the ultimate composition of a
player’s hand of five cards is built up in a larger number of phases, possibly with open own cards or
105
shared (community) cards. Generally speaking, in a complex game with various phases an advanced
player will have more scope to obtain information – information which is freely available to the
fictional player. In a global sense we could thus say that, in a more complex game, the difference in
information between an advanced and a fictional player plays a comparatively smaller role, which
means that the relative skill level is greater.
5.3. A skill analysis of tournament poker
The study into the relative skill level of multiple-player cash poker reveals that, depending on the precise specifications, this game has a skill level ranging between 0.3 and 0.5. Tournament poker differs from cash poker in the following essential manner. Players purchase tokens, which are the only stake throughout the tournament, and the ultimate number of tokens won determines players’ final ranking. Payouts are subsequently awarded on the basis of this ranking only. The relative skill of tournament poker therefore comes to depend in part on the prize structure, and as argued in paragraph 4, this skill level can differ from that of the underlying cash poker game. To obtain insight into the relative skill level of tournament poker, we assume a stylized “base form” of a single game round. If such a single game round is only played for cash, then this game corresponds with cash poker. On account of the limited time available to us, the chosen base form of a single play round is a somewhat crude model, but this could be refined in further research. Tournament poker consists of a series of such stylized game rounds, with tokens at stake. The tokens determine the final ranking, which then translates into cash payouts in accordance with the prize structure. Given our choice for this base form, we studied a number of base tournament variants which explicitly incorporate the characteristic aspects that play an effective role in practice. The step taken in our model from a game round to a tournament is thus an accurate reflection of actual practice. The model particularly takes into account the number of players, the anticipated duration of the tournament, the number of game rounds, the changes of table, the variation in the number of players per table due to possible quitters, the initial amount of tokens and the stake increases across the game rounds. The only aspect not factored in is the dependency between game rounds, generated by a game strategy that stretches across the entire tournament. To incorporate this dependency would make the analysis hugely complex and time-consuming. The so-called BKB base tournament models the BKB poker tournament with 45 participants. Unfortunately, the information available to us did not indicate the exact prize structure. However, it appears from several witness accounts that this structure does not differ significantly from the prize structure used in comparable poker tournaments in Holland Casino, and thus we have used this structure in our analysis. We additionally performed a specific analysis with regard to the so-called HC base tournaments that model the poker tournaments as organized by Holland Casino, based on the Master Classics of Poker. In our analysis we specifically vary the number of players within a realistic duration of play. A more detailed description of the simplified form of a single game round and the corresponding tournaments is offered in the second part of Appendix 15 (poker tournaments). We shall restrict our discussion to the main points here. An important quantity in the analysis concerns the winning percentage (Win%) of the bet that an advanced player will achieve during a base game round with a single beginner as opponent. In our simplified model, this winning percentage should correspond with the winning percentage that
106
an advanced player would achieve at a real cash poker table with a single beginner as opponent. This winning percentage may possibly be derived as well from the real winning percentage achieved by an advanced player at a cash poker table with multiple beginners. Ideally, an accurate estimate of the winning percentage Win% should be derived from empirical material. However, given the nature of this quantity – an advanced player among beginners – such material is not immediately available. The only usable data available derives from the Poker Experiment as documented in Maaten, Borm, Van der Genugten & Hendrickx (2008, Appendix 12). However, given the statistical limitations of the experiment on account of the small number of observations, the analysis contained in the first part of Appendix 15 (bids and gains) only indicates a lower limit for the winning percentage Win%. This percentage will in any case be higher than 5%, and will probably be considerably higher, but we wish to be cautious about drawing any clear-cut conclusions on the basis of this limited experiment. If we take the following rule of thumb (see Meinert (2007), page 270, a standard reference book on poker): “A skilled, solid player places approximately two big bets per hour. So if you play 3 / 6 Euro fixed-limit poker, the hourly wage comes to … around 12 Euros” -- and through some mathematical operations translate this to correspond with our base model, then this means a percentage of 3.5%. It should be noted that the quoted rule of thumb assumes a regular, real-life poker table; so instead of having one “skilled, solid player” facing only beginners, there will be other experienced or even advanced players at the table. The 3.5% derived from the rule of thumb thus signifies a lower limit for the winning percentage Win%. This percentage will in fact be considerably higher than 3.5%, and the percentage of 5% determined through the poker experiment would appear to be a cautious lower limit, based on the quoted rule of thumb. In our analysis we therefore only consider winning percentages starting from 5%. Further, the relative skill level (S-cash) of the corresponding cash poker should be located between 0.3 and 0.5, to remain in conformity with our previous findings for cash poker variants. For all base forms that meet the conditions above, we subsequently calculate the relative skill level of the associated base tournament in four different ways: S-1, S-2, S-3 and S-4. The difference between them is due to the different choices with regard to the prize structure. We have opted to do so in order to make visible the effect of the prize structure on the relative skill. We will now explain the differences between the four calculation methods. S-1: the level of relative skill that is based on a prize structure in which only the final tournament winner receives a payout. S-2: the level of relative skill that is based on a gradual prize structure in which all players receive a payout proportional to their final ranking. With 45 participants, this leads to the structure [45 44 43 … 3 2 1 0 … 0]. S-3: the level of relative skill that is based on a prize structure in which all players at the final table receive a payout proportional to their final ranking while the other players do not receive any payout. With 9 participants at the final table, this leads to the structure [9 8 7 6 5 4 3 2 1 0 … 0]. S-4: the level of relative skill that is based on the prize structure as used by HC and BKB. The 6 best players at the final table in the final ranking receive payouts in a less gradual way (than with respect to S-3) to their final ranking while the other players do not receive any payout. This leads to the structure [37 23 15 11 8 6 0 … 0]. For the classification of the BKB and HC tournaments, the relative skill S-4 is essential given the fact that the associated prize structure most resembles the prize structure that is actually used in these tournaments. The relative skill levels S-1, S-2 and S-3 not only serve for comparison but also to verify the qualitative arguments with respect to the form of the prize structures, as indicated in paragraph 3.2 with respect to management games. The relative skill level S-1 corresponds with the most extreme prize structure imaginable: only the tournament winner wins a prize. On the opposite side of the spectrum, S-2 corresponds with a very gradual prize structure that is fully
107
proportionate to a player’s final position in the final ranking. We shall see that, also for base tournaments, the relative skill level S-1 (corresponding with an extreme prize structure) turns out significantly lower that skill level S-2 with a gradual prize structure. Relative skill levels S-3 and S-4 correspond with prize structures that might be considered intermediate forms, with the prize structure that corresponds with S-3 having a more gradual structure than the structure for S-4. To give an idea of our simulation results for the BKB base tournament, a representative amount of data is summarized below in Table 1. The full version is given in the second part of Appendix 15 (poker tournaments).
Table 1. Relative skill levels of poker tournaments, depending on the winning percentage (Win%) and the prize structure.
Win% S-cash S-1 S-2 S-3 S-4
5 0.50 0.35 0.51 0.44 0.40
5 0.45 0.30 0.44 0.38 0.35
5 0.42 0.26 0.41 0.35 0.31
5 0.38 0.23 0.39 0.32 0.28
5 0.36 0.19 0.34 0.27 0.24
5 0.33 0.17 0.32 0.25 0.22
5 0.31 0.15 0.31 0.24 0.20
10 0.50 0.28 0.47 0.39 0.34
10 0.45 0.24 0.43 0.35 0.30
10 0.43 0.22 0.41 0.32 0.27
10 0.37 0.17 0.36 0.27 0.22
10 0.35 0.15 0.34 0.25 0.20
10 0.32 0.14 0.32 0.23 0.19
10 0.30 0.13 0.31 0.23 0.18
13 0.50 0.28 0.48 0.39 0.34
13 0.46 0.25 0.45 0.36 0.31
13 0.43 0.23 0.43 0.34 0.28
13 0.38 0.20 0.40 0.30 0.25
13 0.36 0.19 0.39 0.29 0.24
13 0.32 0.18 0.37 0.28 0.23
13 0.30 0.17 0.36 0.27 0.22
15 0.50 0.30 0.53 0.43 0.37
15 0.45 0.27 0.50 0.40 0.33
15 0.43 0.26 0.48 0.38 0.32
15 0.39 0.24 0.46 0.36 0.30
15 0.35 0.23 0.44 0.34 0.29
15 0.33 0.23 0.44 0.34 0.28
15 0.31 0.22 0.43 0.34 0.28
17 0.50 0.33 0.53 0.45 0.39
17 0.46 0.31 0.51 0.43 0.37
17 0.44 0.30 0.50 0.41 0.36
17 0.39 0.29 0.49 0.40 0.35
17 0.35 0.28 0.48 0.39 0.34
17 0.33 0.28 0.47 0.39 0.34
17 0.30 0.28 0.47 0.39 0.33
20 0.50 0.40 0.52 0.46 0.43
20 0.45 0.38 0.50 0.44 0.41
20 0.43 0.37 0.48 0.42 0.39
20 0.40 0.36 0.47 0.41 0.38
20 0.35 0.35 0.45 0.40 0.37
20 0.33 0.35 0.45 0.40 0.37
20 0.30 0.35 0.45 0.39 0.37
108
For illustration purposes, we shall examine the following row in this table: Win% S-cash S-1 S-2 S-3 S-4
15 0.39 0.24 0.46 0.36 0.30
This row indicates that, when the fixed winning percentage equals 15% and the fixed relative skill level S-cash = 0.39 in the base form of the cash poker game, the relative skill level S-4 of the corresponding BKB base tournament can be determined as 0.30. A more gradual prize structure for the final table results in a higher relative skill level of S-3 = 0.36, and a fully gradual prize structure results in an even higher skill level of S-2 = 0.46. However, with a payout for only the winner, the relative skill level would result in a lower figure: S-1 = 0.24. To analyze the influence of the winning percentages, in Table 1 we consider 6 different winning percentages: 5%, 10%, 13%, 15%, 17% and 20%. For each of these percentages we consider seven values for the relative skill level S-cash for the base form of cash poker, which adequately cover the potential range from 0.3 and 0.5. From the last column in Table 1, listing the relative skill level S-4, we may conclude that our analysis does not support the classification of the BKB poker tournament as a game of chance. The relative skill S-4 is mostly above 0.2 for all winning percentages, and mostly above 0.3 for a winning percentage of 15% or higher. Remember also that, in choosing the winning percentage, we assumed a real minimum level of 5%. More theoretical considerations such as the analysis for two-phase Texas Hold’em in Appendix 14 suggest that Win% may well exceed 20%, so that it seems entirely justified to classify this a game of skill. However, exercising due caution with respect to our rough modeling, we prefer to restrict our conclusion to saying that there is “no reason for classification as a game of chance.” If we compare the level of relative skill associated with the different prize structures, we see that the BKB base tournament, in which only the tournament winner is awarded a payout with a winning percentage of up to approximately 15%, mostly results in skill levels (see column beneath S-1) of above 0.1 and below 0.3, so that the skill classification remains unclear. However, with a winning percentage of 20% the skill levels would amply exceed 0.3 for this prize structure as well. If the fully gradual prize structure that corresponds with the relative skill levels in the column beneath S-2 were to be applied, then the classification as a game of skill is clear. Mainly on account of the low number of participants of 45, this would also apply in case of the more gradual prize structure at the final table, corresponding with the relative skill levels as listed in the column beneath S-3. The above implies that, for the BKB poker tournament to be classified even more clearly as a game of skill, it would be advisable to adopt a somewhat more gradual prize structure than currently applied. An interesting ancillary conclusion from the above results is that the statement “tournament poker involves more skill than cash poker”, or “cash poker involves more skill than tournament poker”, both prove to be incorrect. With a winning percentage of 17% and a skill level for cash poker of 0.39, the relative skill level S-4 of the corresponding tournament equals 0.35, whereas, with a relative skill level for cash poker of 0.30, the relative skill level of the corresponding tournament equals 0.33. The analysis with respect to the HC base tournaments yields a picture comparable to that for BKB base tournaments, also regarding the important effect of the prize structure on the level of relative skill. On the basis of our findings, there is no reason to classify an HC poker tournament with 50 participants as a game of chance. Our analysis does show very clearly, however, that the number of participants in a tournament in relation to the duration of the tournament plays a crucial role in the skill classification. Already with 100 participants, the prize structures that only award prizes to players on the final table result in relative skill levels (S-3 and S-4) that predominantly fall below 0.3. With 250 or 500 participants, this effect is only amplified. Where such variants are concerned, classification as “game of chance” would seem more appropriate.
109
References
Borm, P.E.M. & B.B. Van der Genugten (1998): On the exploitation of casino games: how to
distinguish between games of chance and games of skill. In: F. Patrone, I. Garcia-Jurado, and S.H. Tijs
(Eds.), Game practice: Contributions from applied game theory, Theory and Decision Library,
Dordrecht: Kluwer Academic Publishers, 19-33.
Borm, P.E.M. & B.B. Van der Genugten (2001): On a relative measure of skill for games with chance
elements. Trabajos de Investigación Operativa (TOP), 9, 91-114.
Borm, P.E.M. & Van der Genugten (2005): Poker: Ein Geschicklichkeitsspiel! Report (in German), Tilburg University. Dreef, M.R.M. (2005): Skill and strategy in games. PhD Thesis, Tilburg University. Dreef, M.R.M. & P.E.M. Borm (2006): On the role of chance moves and information in two-person games. Trabajos de Investigación Operativa (TOP), 14, 75-98.
Dreef, M.R.M., P.E.M. Borm & B.B. Van der Genugten (2003): On strategy and relative skill in poker.
International Game Theory Review (IGTR), 5, 83-103.
Dreef, M.R.M., P.E.M. Borm & B.B. Van der Genugten (2004a): Measuring skill in games: several
approaches discussed. Mathematical Methods of Operations Research (MMOR), 59, 375-391.
Dreef, M.R.M., P.E.M. Borm & B.B. Van der Genugten (2004b): A new relative skill measure for games
with chance elements. Managerial and Decision Economics, 25, 255-264.
Van der Genugten, B.B. (1993): Blackjack in Holland Casino’s: hoe de dealer te verslaan (in Dutch).
Tilburg University Press.
Van der Genugten, B.B. (1996): Eine Spielanalyse des Kartenspiels Concord Aces. Report (in German),
Tilburg University.
Van der Genugten, B.B. (1997a): Een keuringsmethode voor behendigheidsautomaten, EIT-report (in
Dutch), Tilburg University.
Van der Genugten, B.B. (1997b): De toepassing van de wet op de kansspelen op het jokerspel,
jahtzee en eurobsgame. EIT-report (in Dutch) , Tilburg University.
110
Van der Genugten, B.B. (1997c): Blackjack in Holland Casino’s: basic, optimal and winning strategies.
Statistica Neerlandica, 51, 318-344.
Van der Genugten, B.B. (2008): Pokeren: bluffen met wiskunde (in Dutch). Euclides, 4, 234-237.
Van der Genugten, B.B. & P.E.M. Borm (1994a): Een vergelijking van de behendigheid in spelen met
een kanselement. Report (in Dutch), Tilburg University.
Van der Genugten, B.B. & P.E.M. Borm (1994b): Het financieel piramidesysteem: kans of
behendigheidsspel? (& repliek). Reports (in Dutch), Tilburg University.
Van der Genugten, B.B. & P.E.M. Borm (1996a): On a measure of skill for games with chance
elements. Report FEB 721, Tilburg University.
Van der Genugten, B.B. & P.E.M. Borm (1996b): Het behendigheidsniveau van spelen en de
diagnostische criteria. Report (in Dutch), Tilburg University.
Van der Genugten, B.B. & P.E.M. Borm (1996c): Gokken op een kansspelkwalificatie:
een behendigheidsspel? (in Dutch). Justitiële Verkenningen, 6, 104-108.
Van der Genugten, B.B. & P.E.M. Borm (2005): Kansspelen. In: Het onzekere bewijs; gebruik van
statistiek en kansrekening in het strafrecht (eds. Sjerps & Coster van Voorhout) (in Dutch), Kluwer,
Deventer, 153-183.
Van der Genugten, B.B., P.E.M. Borm & M.R.M. Dreef (2004): De toepassing van de Wet op de Kansspelen op de managementspelen Competitie Manager en Grand Prix Manager. Report (in Dutch), Tilburg University.
Van der Genugten, B.B., P.E.M. Borm & M.R.M. Dreef (2005): De toepassing van de wet op de kansspelen op de managementspelen Competitie Manager en Grand Prix Manager (vervolg). Report (in Dutch), Tilburg University.
111
Van der Genugten, B.B., P.E.M. Borm & W. Grossmann (1997): Addendum to two reports on games
of skill. Report, Tilburg University.
Van der Genugten, B.B. & R.C.M. Brekelmans (2006): Spot the ball: game of chance or skill? Report,
Tilburg University.
Van der Genugten, B.B. & M. Das (1998): An analysis of American Poker 2. EIT-report, Tilburg
University.
Van der Genugten, B.B. & M. Das (1999a): An analysis of Fruitcard. EIT-report, Tilburg University.
Van der Genugten, B.B. & M. Das (1999b): Een analyse van Blackjack varianten. CentER-AR report (in
Dutch), Tilburg University.
Van der Genugten, B.B. & M. Das (1999c): Een analyse van behaalde winsten bij Caribbean Stud
Poker in Holland Casino’s. EIT-report (in Dutch), Tilburg University.
Van der Genugten, B.B. & M. Das (2005): Random Flashback: a detailed analysis. CentER-AR report,
Tilburg University.
Van der Genugten, B.B., M. Das & P.E.M. Borm (2001): Behendig gokken in en rond het casino (in
Dutch). Academic Press.
Van der Genugten, B.B. & J.C. De Vos (1998): De toepassing van de Wet op de Kansspelen op het spel
Dromus 24. EIT-report (in Dutch), Tilburg University.
Van der Genugten, B.B., J.C. De Vos & M. Das (1999): Het potentieel behendigheidsniveau van ROTA-
roulette. EIT-report (in Dutch), Tilburg University.
Hilbers, P. (2007): Skill in poker. Research master thesis. CentER and Department of Econometrics
and OR, Tilburg University.
112
Hilbers, P., R. Hendrickx, P. Borm & B.B. Van der Genugten (2008): Measuring skill in more-person
games with applications to poker. CentER Discussion paper 2008-116, Tilburg University.
Maaten, R. (2009): Poker: a game of chance. MSc Thesis, Department of Econometrics and OR,
Tilburg University.
Maaten, R., P. Borm, B.B. Van der Genugten & R. Hendrickx (2008): Behendigheidsniveau van Texas
Hold’Em: een experiment uitgevoerd in samenwerking met VARA-nieuwslicht. Report (in Dutch),
Tilburg University.
Meinert, J. (2007): Pokerschool (in Dutch). A.W. Bruna Uitgevers BV, Utrecht.
De Vos, H. (1997): Golden ten and related trajectory games. PhD Thesis, Tilburg University.
Wilson, B. (2005): Turbo Texas Hold’em for windows: user’s guide. Wilson software, Inc, Green
Valley, USA.
113
Behendigheidsniveau van Texas Hold’Em
Een experiment uitgevoerd in samenwerking met VARA-Nieuwslicht
Concept-verslag
Rogier Maaten
Peter Borm, Ben van der Genugten, Ruud Hendrickx
Universiteit van Tilburg, 27 maart 2008
Inhoudsopgave
1. Inleiding 2
2. Fixed Limit Texas Hold’Em 2
2.1 Notatie 2
2.2 Pokerhanden 2
2.3 Spelverloop 3
2.4 Behendigheidsniveau 5
3. Opzet en uitvoering experiment 5
3.1 Doel experiment 5
114
3.2 Globale opzet experiment 6
3.3 Details van de opzet van het experiment 6
3.4 De voorbereidende fase 6
3.5 De aanloopfase 7
3.6 De speelfase 7
3.7 Tafelindeling 7
3.8 Uitvoering 8
4. Analyse experiment 9
4.1 Spelverlopen 9
4.2 Resultatenlijst 9
4.3 Variantieanalyse resultaten 10
4.4 Behendigheidsniveau 10
4.5 Beginnersstrategie 12
4.6 De ervaren speler 17
5. Conclusies 17
6. Bijlage: Spelverlopen 18
7. Bijlage: Resultaten per positie 68
115
1. Inleiding
Volgens de Hoge Raad is poker een kansspel. Hierdoor mag poker officieel uitsluitend gespeeld
worden in vestigingen van Holland Casino; commerciële exploitatie op andere locaties is niet
toegestaan. Is deze classificatie terecht, of is poker eigenlijk een behendigheidsspel?
In een experiment, gehouden in samenwerking met de VARA-Nieuwslicht, is getracht om het
behendigheidsniveau van poker aan het licht te brengen. De details van het experiment, en de
resultaten ervan zijn in dit rapport te vinden.
De opbouw van dit rapport is als volgt. Allereerst worden de spelregels en het spelverloop van Texas
Hold’Em beschreven, gevolgd door een beschrijving van het pokerexperiment. In het hoofdstuk
“Analyse experiment” wordt een schatting van het behendigsheidsniveau van poker gegeven aan de
hand van de resultaten van het experiment. Ook is er een karakterisering van de strategie van de
beginners te vinden. Afgesloten wordt met enkele conclusies betreffende de nauwkeurigheid van de
resultaten van dit experiment en het behendigheidsniveau van Texas Hold’Em.
2. Fixed Limit Texas Hold-Em
2.1 Notatie
De verschillende kaarten in het spel worden als volgt aangeduid: A,K,Q,J,T,9,8,7,6,5,4,3,2. Hierbij
staat A voor ace (aas), K voor king (koning), Q voor queen (vrouw), J voor jack (boer) en T voor ten
(tien). De getallen in deze reeks spreken voor zich.
Een pokerspel bevat vier kleuren: schoppen (s van spades), harten (h van hearts), ruiten (d van
diamonds) en klaveren (c van clubs).
Een speler die AhJs ontvangt heeft dus de aas van harten en de boer van schoppen.
Wanneer een hand twee kaarten van dezelfde kleur bevat, dan noemen we deze hand “suited”.
Dit wordt genoteerd als A9s, 74s, etc. Handen die niet suited zijn, noteren we als T9o, 52o (o staat
voor “off-suit”).
2.2 Pokerhanden
Texas Hold-Em lijkt de meest populaire vorm van poker. Het spel wordt gespeeld met minimaal twee
en maximaal tien spelers. Iedere speler krijgt twee kaarten, die niet aan de andere spelers getoond
worden. Deze kaarten worden gecombineerd met vijf kaarten, die open op tafel komen te liggen
gedurende het spel (genaamd “community cards”). Van de twee kaarten in de hand en de vijf
community cards gebruikt iedere speler in totaal vijf kaarten om daarmee een zo sterk mogelijke
pokerhand te maken.
116
De pokerhanden van hoog naar laag zijn:
Royal Flush: AHVBT in een kleur. In poker zijn alle kleuren van gelijke hoogte. In andere varianten van
poker (bijvoorbeeld 7-Card Stud) is het mogelijk dat twee spelers een royal flush halen in een spel.
Straight Flush: Vijf opvolgende kaarten in een kleur, bijvoorbeeld 98765 van harten. Een aas mag
gebruikt worden als hoogste kaart in het spel, maar ook als laagste kaart in het spel. 5432A van
ruiten is dus ook een straight flush, 32AKQ van klaveren is geen straight flush. Tussen straight flushes
onderling is de hoogste kaart de beslissende factor. 98765 wint dus van 5432A.
Four of a kind: Vier kaarten van dezelfde rang, zoals vier vrouwen. Ook wel bekend onder de naam
“quads”. De rang beslist in het geval dat meerdere spelers four of a kind hebben, dus KKKK3 wint van
5555A. Als meerdere spelers een vierkaart van dezelfde rang hebben, dan is de hoogste bijkaart
cruciaal. Dit komt voor wanneer van de vijf community cards er vier dezelfde rang hebben. Als TTTT6
op tafel ligt, speler A heeft QJ in de hand en speler B heeft A7, dan heeft speler B de beste hand
(TTTTA).
Full House: Combinatie van drie kaarten van dezelfde rang plus twee kaarten van een andere rang,
bijvoorbeeld JJJ44. De rang van de driekaart is beslissend indien er meerdere spelers een full house
hebben gemaakt. 999QQ wint van 666KK.
Flush: Vijf kaarten in een kleur, zoals K9632 van schoppen. De hoogste kaart is bepalend bij flush
tegen flush, daarna de één na hoogste, etc.
Straight: Vijf opvolgende kaarten, zoals QJT98. Kleur is niet van belang. De hoogste kaart zorgt weer
voor het verschil als een andere speler ook een straight heeft gemaakt.
Three of a Kind: Drie kaarten van dezelfde rang plus twee andere kaarten, bijvoorbeeld 888KT. De
rang bepaalt weer bij meerdere three of a kinds, bij gelijke rang beslist de hoogste bijkaart. Andere
benamingen voor three of a kind zijn “trips” of een “set”.
Two Pair: Een pair bestaat uit twee kaarten van dezelfde rang. Two pair is een combinatie van twee
verschillende pairs, aangevuld met met de hoogst overgebleven kaart. Het hoogste pair is cruciaal in
geval van two pair tegen two pair. Daarna beslist het tweede paar, en daarna de bijkaart. AA882
verslaat KK55Q, en 9944K verslaat 99226.
Pair: Dit is een combinatie van een pair en drie overige kaarten. Zoals gebruikelijk is 77K83 beter dan
55AKQ, en wint KKQ65 van KKJ95.
High Card: Hieronder vallen alle overige handen. De hoogste kaart is cruciaal als het gaat om high
card versus high card.
2.3 Spelverloop
De spelers zijn beurtelings deler (of “dealer”). Met zes spelers aan tafel is iedere speler dus om de zes
spellen dealer. De speler direct links van de dealer plaatst een verplichte inzet van 1 fiche aan het
begin van het spel. Deze inzet wordt de small blind genoemd (kortweg SB). De volgende speler, twee
117
plaatsen van de dealer, plaatst de big blind (BB) van 2 fiches. Vervolgens worden de kaarten gedeeld,
en krijgt iedere speler zijn twee kaarten in de hand. Aan de hand hiervan, en aan de hand van de
acties van de andere spelers beslissen zij welke actie ze zullen ondernemen. Ze hebben de volgende
opties:
Fold: Terugtrekken uit het lopende spel. De kaarten worden teruggegeven aan de dealer, en de
speler speelt verder geen rol meer in het spel.
Call: De call-optie gebruik je indien je door wilt spelen in de hand. Je investeert 2 fiches in dit spel
(evenveel als de big blind). Mocht er al door een andere speler zijn verhoogd, dan kost callen meer
fiches (evenveel als hoeveel de speler die heeft verhoogd heeft ingezet).
Bet/Raise: De huidige inzet verhogen. De eerste verhoging wordt altijd aangeduid met “bet”, alle
volgende verhogingen zijn “raises”. Aangezien we Fixed Limit Texas Hold’Em spelen, kunnen spelers
uitsluitend verhogen met 2 fiches. Het maximaal aantal verhogingen in deze spelronde is vier, spelers
kunnen in deze fase dus niet meer dan 8 fiches investeren.
Check: De check-optie is in deze fase alleen relevant voor de big blind in het geval dat niemand
verhoogt. Hij heeft al zijn verplichte 2 fiches geinvesteerd, en heeft nu de optie om zonder extra inzet
door te spelen.
Een voorbeeld van mogelijke acties staat weergegeven in Tabel 1. De vetgedrukte speler, A6, is de
dealer. Speler A1 heeft de small blind geplaatst en speler A2 de big blind. Spelers A4 en A5 hebben
zich teruggetrokken uit de hand, de overige spelers hebben ieder 2 fiches geïnvesteerd, en spelen nu
verder. Niemand heeft verhoogd.
A1 A2 A3 A4 A5 A6
SB BB Call Fold Fold Call
Call Check
Tabel 1: Acties op de preflop
Als alle acties zijn afgerond, komen de eerste drie open kaarten (van de vijf community cards) op
tafel te liggen. Samen heten deze drie kaarten de “flop”. Analoog aan de vorige betronde hebben de
overgebleven spelers nu weer de mogelijkheid om in te zetten. De speler op de positie van de small
blind is weer als eerste aan de beurt. Hij heeft de volgende opties: check en bet. Tot er een speler
een bet plaatst, hebben de spelers dezelfde opties: check en bet. Als een speler een bet heeft
geplaatst, hebben de volgende spelers deze opties: fold, call en raise. Weer geldt dat vier bets het
maximale aantal is, nadat iemand een bet heeft geplaatst mag er dus nog maximaal drie keer worden
geraised. Een bet bedraagt nog altijd 2 fiches.
118
Een mogelijk biedverloop op de flop is weergegeven in Tabel 2 (waarbij spelers A4 en A5 zich al voor
de flop, oftewel preflop, hebben teruggetrokken). Speler A2 houdt het nu ook voor gezien, de andere
drie actieve spelers hebben allen 4 fiches geinvesteerd in deze fase.
A1 A2 A3 A4 A5 A6
Check Check Bet Call
Raise Fold Call Call
Tabel 2: Acties op de flop
Nu komt de vierde open kaart op tafel, dit wordt de “turn” genoemd. Deze fase is identiek aan de
“flop”, behalve dat iedere bet nu 4 fiches kost in plaats van 2. Het voorbeeldspel gaat verder als
aangegeven in Tabel 3.
A1 A2 A3 A4 A5 A6
Check Bet Call
Fold
Tabel 3: Acties op de turn
Speler A1, die op de flop nog verhoogd heeft, geeft hier de hand op. Wellicht was de turnkaart voor
hem ongunstig, of hij was aan het bluffen en verloor de hoop dat zijn tegenspelers de hand op
zouden gaan geven. Spelers A3 en A6 gaan verder naar de “river”.
De “river” is de laatste fase in het spel. De laatste van de vijf community cards wordt nu getoond. De
resterende spelers zien nu welke hand ze gemaakt hebben. De betronde is hetzelfde als op de turn,
een bet kost 4 fiches. Speler A3 heeft nogmaals verhoogd. Speler A5 erkent zijn verlies en is niet
bereid om nog 4 extra fiches in deze hand te investeren, zoals weergegeven in Tabel 4.
119
A1 A2 A3 A4 A5 A6
Bet Fold
Tabel 4: Acties op de river
Aangezien speler A3 de enige overgebleven speler is, ontvangt hij nu de inzetten van alle spelers in
dit spel. Hij wint dus de pot. Indien er meedere spelers over zijn in deze fase, worden de kaarten in
de hand open gedraaid, en ontvangt de speler met de beste combinatie de pot.
Vervolgens schuift de dealer positie een plaats op. Ook de positie van de small blind en big blind
verplaatsen nu, en het volgende spel kan beginnen.
2.4 Behendigheidsniveau
Het behendigheidsniveau van een spel kan met de volgende formule worden uitgerekend (Borm en
Van der Genugten, 1998)1:
leereffect
Behendigheidsniveau =
leereffect + random effect
Het leereffect wordt in het kader van dit experiment weergegeven door het verschil in winst tussen
de ervaren speler en de beginner. Het verschil in winst tussen de fictieve speler en de ervaren speler
is het random effect, ofwel het toevalseffect. De ervaren speler heeft geen voorkennis over de
toevalselementen van het spel (hij weet niet welke kaarten de andere spelers hebben, en welke
kaarten er op tafel zullen komen te liggen). De fictieve speler is wel op de hoogte van deze
informatie. Samengevat:
Leereffect = winst ervaren speler - winst beginner,
Random effect = winst fictieve speler - winst ervaren speler.
1 Borm, P. and B. van der Genugten (1998). On the exploitation of casino games: how to
distinguish between games of chance and games of skill. In: F. Patrone, I. García-Jurado, and
S. Tijs (Eds.), Game Practice: Contributions from Applied Game Theory, pp. 19-33.
Dordrecht: Kluwer Academic Publishers.
120
Kansspelen, waarbij het toevalseffect het leereffect domineert (zoals roulette), hebben een
behendigheidsniveau dat dicht bij 0 ligt. Behendigheidsspelen, waarbij het toevalseffect klein is in
vergelijking met het leereffect (zoals schaken), hebben een behendigheidsniveau dicht bij 1. Zoals
beargumenteerd in Hilbers et al. (2008)2 bedraagt het theoretische behendigheidsniveau van Texas
Hold’Em tenminste 0,30.
3. Opzet en uitvoering experiment
3.1 Doel experiment
Hoofddoel: experimenteel vaststellen van het behendigheidsniveau in Fixed Limit Texas Hold’Em, en
te zien of deze schatting in de buurt komt van het theoretische behendigheidsniveau van Texas
Hold’Em.
Nevendoel: inzicht te krijgen in de vorm van een beginnersstrategie.
3.2 Globale opzet experiment
Tijdens het experiment wordt Fixed Limit Texas Hold’Em gespeeld. Er wordt gespeeld aan drie tafels,
door zes spelers per tafel. In totaal zijn er dus 18 spelers nodig, hiervan zijn 16 spelers beginners, één
speler is een ervaren speler en één speler is een fictieve speler.
In totaal worden er 36 spellen gespeeld, verdeeld over drie spelrondes van 12 spellen. Na iedere
spelronde wordt er gepauzeerd, en wisselen de spelers van plaats.
Alle spelers ontvangen 500 fiches en deze fiches vertegenwoordigen echt geld. Aan het einde van het
experiment wisselen de spelers hun fiches in voor geld. De spelers hebben voldoende fiches om niet
voor het einde van de laatste spelronde blut te raken.
Na afloop wordt de precieze opzet en het doel van het experiment, alsmede een voorlopige schatting
van het behendigheidsniveau bekendgemaakt.
3.3 Details van de opzet van het experiment
Er worden 36 spellen per tafel gespeeld. In totaal zijn dit 108 spellen, de 108 benodigde kaartspellen
zijn voorafgaand aan het experiment voorbereid. Aan elk van de drie tafels worden dezelde spellen
synchroon gespeeld door andere spelers. Met behulp van de computer zijn de corresponderende
kaartvolgorden random gegenereerd. De kaartvolgordes van ieder spel aan de eerste tafel zijn dus
hetzelfde als de kaartvolgordes aan de andere tafels.
De spelers krijgen spelersnummers: 1,2,...,19. Speler 19 is een invalspeler, hij zal gedurende het
experiment invallen voor beginners die op een bepaald moment niet beschikbaar zijn. Speler 6 is de
2 Hilbers, P., R. Hendrickx, P. Borm and B. van der Genugten (2008). Skill in poker. Working
paper.
121
fictieve speler, speler 12 de ervaren speler. Alle andere nummers worden door loting toegewezen
aan de beginners. Hiervan wordt een deelnemerslijst gemaakt.
De plaasten aan tafels zijn ook genummerd: bij tafel A (met de klok mee) de nummers A1,A2,...,A6,
soortgelijk bij tafel B en tafel C.
In de eerste fase krijgt de fictieve speler tafelpositie A6, de ervaren speler krijgt positie B6. De
overige posities worden door loting toegewezen, zo ook de belangrijke positie C6. De resultaten van
de beginner die op positie C6 zit worden namelijk vergeleken met de resultaten van de fictieve speler
en de ervaren speler.
In de tweede fase krijgt de fictieve speler tafelpositie B6, de ervaren speler tafelpositie C6, en de
belangrijke beginner (een andere speler dan in de eerste fase) positie A6.
In de derde fase krijgt de fictieve speler tafelpositie C6, de ervaren speler tafelpositie A6 en de
belangrijke beginner positie B6.
Het loten voor de verschillende posities geschiedt met de computer.
3.4 De voorbereidende fase
De 36 verschillende lotingsvolgorden worden met de computer gemaakt. De 108 kaartspellen
worden samengesteld als boven aangegeven.
De fictieve speler wordt aangezocht. Hij krijgt het eigenlijke doel van het experiment te horen. Hij
weet dat hij alleen beginners als tegenstander heeft. Voorts krijgt hij een lijst die voor elk van de 36
spellen alle kaartvolgorden geeft en die bovendien de (vaste) tafelposities vermeldt van de spelers
die aan het eind (na de river) respectievelijk een betere hand, een gelijke hand en een slechtere hand
gemaakt hebben dan hij.
De ervaren speler wordt aangezocht. Hij krijgt als onderdeel van het experiment te horen dat het
doel is na te gaan hoeveel beter een ervaren speler presteert dan een beginner. Hij weet dus ook dat
hij uitsluitend tegen beginners zal spelen. Hij krijgt niets over het bestaan van een fictieve speler te
horen.
De 17 beginners worden aangezocht. Hen wordt verteld dat ze niets hoeven voor te bereiden en dat
speluitleg op de dag zelve wordt gedaan, en dat het doel van het experiment het nagaan is wat het
effect is van een pokerface op het spelresultaat. Zij krijgen niets te horen over een ervaren speler en
een fictieve speler. Van de beginners wordt een beknopt profiel van hun pokerervaring gemaakt.
122
3.5 De aanloopfase
De 19 spelers arriveren en krijgen eerst speluitleg (ook de ervaren en de fictieve speler zijn aanwezig
maar houden hun mond over alles wat met Texas Hold’Em te maken heeft, desgevraagd zijn ook zij
beginners).
Bij de speluitleg moet er wel een aantal oefenrondjes gespeeld worden om de spelregels te leren en
enigszins te verwerken, maar de instructeur moet er voor waken geen strategieadviezen te geven.
Dus beginners wordt wel duidelijk dat het fijn is met de beste hand in de showdown te zitten (dit is
het moment waarop de handkaarten open gaan, en de winnaar van de hand bekend wordt) en dat
met een slechte hand dit wel heel toevallig zou zijn. Er worden geen aanwijzingen gegeven over hoe
te handelen in specifieke situaties.
Vervolgens wordt verteld dat er in drie rondes gespeeld gaat worden. De spelers krijgen fiches en de
bijbehorende geldwaarde wordt bekend gemaakt. Daarna worden de spelernummers uitgeloot en
nemen de spelers hun eerste tafelposities in.
3.6 De speelfase
De eerste fase bestaat uit 12 spellen. De spelers ontvangen een spelerformulier, waarop ze voor
ieder spel van deze fase moeten aangeven welke beslissingen ze achtereenvolgens genomen hebben.
Voor ieder spel is een rij gereserveerd waarin de spelers de achtereenvolgens genomen beslissingen
noteren (deze beslissingen zijn: fold, call, raise, check en bet). Dit dient goed uitgelegd te worden.
Aan het einde van de eerste spelfase vermelden de spelers op hun formulier hoeveel fiches ze op dat
moment in bezit hebben, daarna worden de formulieren ingeleverd. Voor de tweede en de derde
spelronde ontvangen alle spelers een nieuw formulier en ze worden geacht hiermee op dezelfde
manier als in de eerste spelronde mee om te gaan.
Iedere tafel kent een spelleider (die zelf niet meespeelt). Deze deelt de kaarten (en mag in dit spel
dus nooit schudden of couperen). Hij ziet er op toe dat de spellen in de juiste volgorde worden
gespeeld. Verder controleert de spelleider dat de spelers hun formulieren invullen. Hij krijgt (ter
controle) een tafellijst: deze bevat 36 rijen voor elk spel. In iedere rij noteert hij ter controle de
riverkaart en de hoogte van de bereikte pot. Na afloop zijn er dus drie tafellijsten van de tafels A,B en
C beschikbaar. De spelleiders aan de drie tafels zorgen ervoor dat het spelen synchroon verloopt.
De spelers beginnen iedere spelronde met hetzelfde aantal fiches op tafel. Hiermee worden
psychologische effecten die een afhankelijkheid van de eerste fase en de tweede fase kunnen geven
uitgesloten. Tijdens het spelen mag geen commentaar uitgewisseld worden over te nemen of
genomen beslissingen.
123
3.7 Tafelindeling
Tabel 5 geeft de tafelindeling voor iedere ronde weer. Deze indeling is tot stand gekomen door
middel van loting. De ervaren speler en de fictieve speler zitten iedere ronde op positie 6 (aan
verschillende tafels), zodat hun resultaten kunnen worden vergeleken. Ze krijgen immers precies
dezelfde kaarten. De beginners die op positie 6 zitten, zijn in dit experiment extra belangrijk. Ook hun
acties en resultaten worden vergeleken met die van de ervaren en fictieve speler. De belangrijke
beginners zijn speler 1 in ronde 1, speler 7 in ronde 2 en speler 10 in ronde 3.
Dealers Ronde 1 Ronde 2 Ronde 3
Tafel A Dealer 1 Dealer 1 Dealer 1
Tafel B Dealer 2 Dealer 2 Dealer 3
Tafel C Dealer 3 Dealer 3 Dealer 2
Spelers Ronde 1 Ronde 2 Ronde 3
A1 5 5 15
A2 16 2 17
A3 14 3 5
A4 18 17 13
A5 8 14 11
A6 6 7 12
B1 2 19 2
B2 15 1 19
B3 19 9 8
B4 9 8 18
B5 13 15 7
B6 12 6 10
C1 11 16 1
C2 4 4 16
C3 3 10 4
C4 7 11 9
C5 10 13 14
124
C6 1 12 6
Tabel 5: Tafelindeling
3.8 Uitvoering
De invulformulieren die in dit experiment zijn gebruikt, bleken minder geschikt voor beginnende
spelers. Zodra een speler meedere acties in een fase deed (bijvoorbeeld eerst “check” en na een
raise van een speler achter hem “fold”), ontstonden er vaak problemen met het invullen. Meestal
werd uitsluitend de eerste actie genoteerd, slechts enkele spelers hebben meerdere acties per fase
ingevuld. Ook de dealers zijn niet overal zorgvuldig geweest. De grootte van de pot aan het einde van
ieder spel is vaak incorrect genoteerd, soms werd zelfs volstaan met een uitdrukking als “veel”, wat
niet erg handig is voor de analyse van de spellen. Het nagaan van wat er daadwerkelijk in ieder spel
gebeurd is, was nochtans mogelijk, maar vergde zodoende behoorlijk wat werk.
Voor een volgend experiment zijn de volgende zaken aan te bevelen:
- De invulformulieren dienen duidelijk te zijn voor spelers die nooit poker hebben gespeeld. - De dealers hebben de belangrijke taak om nauwkeurig to te zien dat ook de registratie
volgens de regels verloopt. - Het spelen van een proefronde, die niet meetelt voor het uiteindelijke resultaat is aan te
bevelen. Zo kan de ervaren speler wennen aan het spel van de beginners, en krijgen de
beginners handigheid in het correct noteren van hun acties.
Na afloop van het experiment heeft iedere speler zijn fiches ingeruild voor geld. Uit de op de
deelnemerslijsten vermelde fiches is een resultatenlijst gemaakt. Met iedere deelnemer
correspondeert een rij van drie getallen die zijn hoeveelheid fiches vermeldt aan het eind van iedere
fase. Deze lijst is in de computer ingevoerd en die heeft hieruit een schatting van het
behendigheidsniveau berekent. Vervolgens is de echte opzet van het experiment uitgelegd en is (met
de nodige reserve) de schatting bekendgemaakt.
4. Analyse experiment
4.1 Spelverlopen
De bijlage Spelverlopen bevat de acties van alle spelers in alle 36 spellen. Per spel kan hier bekeken
worden wat iedere speler in iedere fase gedaan heeft. Verder wordt weergegeven hoeveel fiches
iedere speler per spel heeft ingezet, wie de winnaar van het spel is, en wat de fichetotalen van de
spelers zijn na ieder spel. De eigen kaarten van de spelers, en ook de community cards, zijn hier terug
te vinden. Deze kunnen gebruikt worden om te bekijken hoe een typische beginner speelt: “met
welke handen verhoogt hij?”, “hoe speelt hij op de flop als hij een pair heeft gemaakt?”, etc.
125
In deze bijlage valt ook te zien hoe actief de spelers zijn. De fictieve speler speelt erg weinig handen
(hij trekt zich vaak voor de flop terug). Dit is niet vreemd aangezien deze speler al voorafgaand aan
het spel al weet of hij de beste hand gaat maken of niet. De ervaren speler weet dat het inzetten met
marginale starthanden (zoals T7s, K3o) niet winstgevend is, ook deze speler speelt relatief weinig
handen. De beginners zijn een stuk actiever, de meeste van hen zijn bereid om met zowat iedere
hand (dus ook de slechtste handen in Texas Hold’Em, zoals 72o, 83o) fiches in te zetten om te zien of
de flop aansluiting biedt.
4.2 Resultatenlijst
Tabel 6 geeft de fichetotalen van alle spelers aan het einde van elk van de drie de spelronden.
Speler Ronde 1 Ronde 2 Ronde 3
1 540 483 374
2 553 530 531
3 447 461 461
4 518 499 464
5 570 566 577
6 534 695 914
7 449 479 476
8 456 428 475
9 444 457 542
10 495 539 507
11 551 554 527
12 467 557 558
13 473 407 470
14 456 427 484
15 507 418 356
16 515 463 246
17 500 512 526
18 469 469 488
19 556 556 524
Tabel 6: Fichetotalen van de spelers per ronde
126
Hierbij dient te worden opgemerkt dat spelers 3,17 en 18 slechts in twee van de drie spelronden
hebben meegedaan. De grote winnaar in het experiment is de fictieve speler (speler 6). Hij heeft zijn
aantal fiches bijna weten te verdubbelen, de andere spelers komen hier niet bij in de buurt. In de
derde spelronde heeft hij een winst behaald van 219 fiches in slechts 12 spellen. De ervaren speler
(speler 12) heeft het experiment ook met winst afgesloten, maar komt toch niet in de buurt van de
fictieve speler. De eerste spelronde heeft hij zelfs met verlies af moeten sluiten. Zijn uiteindelijke
winst bedraagt 58 fiches. De drie beginners die op dezelfde posities als de ervaren en de fictieve
speler gespeeld hebben, hebben in de drie spelronden een winst van 38 fiches gerealiseerd. Tabel 7
geeft de winsten van de belangrijke spelers per spelronde.
Winst Belangrijke Beginner Ervaren Fictief
Ronde 1 40 -33 34
Ronde 2 30 90 161
Ronde 3 -32 1 219
Totaal 38 58 414
Tabel 7: Winsten van de belangrijke spelers per spelronde
In de uitzending van VARA-Nieuwslicht zijn de getallen gebruikt die voortkomen uit de ingevulde
resultatenlijst. Bovenstaande tabel komt voort uit de gecorrigeerde spelverlopen, en is dus niet
hetzelfde als de tabel die direct na afloop van het experiment is gemaakt. Gelukkig zijn de verschillen
klein. In het vervolg van dit verslag wordt vanzelfsprekend gerekend met de gecorrigeerde getallen.
4.3 Variantieanalyse resultaten
Met behulp van een beknopte variantieanalyse hebben we bepaald welke factoren (spelerstype,
spelronde, tafelposititie, etc.) het meest bepalend zijn voor de behaalde spelresultaten. Deze analyse
bevestigt het beeld dat de fictieve speler weliswaar een fiks hogere winst behaalt dan de andere
spelerstypes, maar dat het overall verschil tussen de ervaren speler en de beginners statistisch niet
significant is. Ook de overige factoren blijken statistisch nauwelijks van invloed te zijn. Het zeer
beperkt aantal van 36 spellen zorgt namelijk voor een hoge variantie in de resultaten.
4.4 Behendigheidsniveau
Het geschatte behendigheidsniveau van Texas Hold’Em op basis van alle rondes is 0,05. Dit komt
voort uit de volgende berekening (zie ook tabel 7):
127
Leereffect = winst ervaren speler - winst beginner = 58 - 38 = 20
Random effect = winst fictieve speler - winst ervaren speler = 414 - 58 = 356
Behendigheidsniveau = leereffect / (leereffect + random effect) = 20/(20+356) = 0,05
De waarde van 0.05 is ver verwijderd van het theoretische behendigheidsniveau van minstens 0.30.
Hier zijn in het experiment ook wat duidelijke oorzaken voor te vinden. Door onregelmatigheden in
de gespeelde spellen is de winst van de ervaren speler aan het einde van het experiment erg laag. De
winst van de fictieve speler daarentegen is extreem hoog. Voor een meer realistische schatting van
het behendigheidsniveau zullen we een specifieke correctie doorvoeren aan de hand van
geconstateerde uitschieters.
In de eerste spelronde behaalt de belangrijke beginner een grotere winst dan de ervaren speler (die
zelfs verliest maakt). Dit resultaat is voornamelijk het gevolg van spel 9. In dit spel krijgen de spelers
op tafelpositie 6 (dus de ervaren en de fictieve speler, en de belangrijke beginner) de beste hand
gedeeld, namelijk AdAc. Tabel 8 toont dit spel, aan de tafel van de beginner.
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Jc3c 6h3h AsJs JdJh 8d3s AdAc
Preflop SB BB Fold Bet
Fold Call Call Raise Raise
Call Call Call
Flop Check Bet Raise
Call Call Call
Turn Check Bet Raise
Call Call Call
River Check Check Bet
Fold Fold Raise Call
Tabel 8: Spel 9, Tafel C, Ronde 1
Speler 1 (de belangrijke beginner in de eerste spelronde), verhoogt tot twee keer toe preflop. Ook op
de flop (7s8c5d) en op de turn (Qh) is hij ervan overtuigd dat zijn hand de beste is, zelfs tegen drie
128
tegenstanders. Hij blijft verhogen, ook op de river, en zijn hand blijkt bij de showdown ook inderdaad
de beste hand te zijn. Hij verdient 68 fiches met deze hand.
De ervaren speler ontvangt ook AdAc, dit spel verloopt volgens tabel 9.
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Jc3c 6h3h AsJs JdJh 8d3s AdAc
Preflop SB BB Fold Bet
Fold Call Raise Raise Call
Call Call
Flop Bet Raise Raise
Call Call Call
Turn Bet Call Raise
Raise Call Fold Fold
River Bet
Fold
Tabel 9: Spel 9, Tafel B, Ronde 1
De acties op de preflop zijn vergelijkbaar met het spel aan tafel C, dezelfde vier spelers zien de flop.
De flop is 7s8c5d, niet bepaald een flop die speler 19 (op positie B3) gunstig gezind is. Met zoveel
raises op de preflop is het onwaarschijnlijk dat deze speler op de river de beste hand gaat hebben (hij
kan alleen nog winnen als de turn en de river allebei een schoppen brengen, dit heeft slechts een
kans van 5,49%). Dit weerhoudt hem echter niet om toch voor bet te kiezen, de volgende speler (met
JdJh, en dus wel een sterke hand) verhoogt, en de ervaren speler verhoogt nogmaals. Nog altijd met
vier spelers in het spel brengt de turn Qh. Waar de beginner zonder angst door blijft zetten, geeft de
ervaren speler de hand op. Hij denkt dat wanneer drie tegenstanders blijven verhogen, er toch op
zijn minst een speler is die zijn paar azen verslagen heeft. Als ook maar een van de tegenstanders een
van deze handen heeft: QQ, 88, 77, 55, 96, 64, dan is het voor hem correct om te passen. Tegen de
three of a kind handen kan de ervaren speler alleen nog winnen met een aas op de river (kans van
4,55%), tegen de handen die al een straight gemaakt hebben is de ervaren speler nu helemaal
kansloos (dit heet “drawing dead” in poker). Gezien de echte kaarten van de tegenstanders blijkt het
passen van de ervaren speler een dure misstap. Hij had op ieder moment de beste hand, en zou ook
in de showdown de pot gewonnen hebben. In plaats van ongeveer 75 fiches winst te maken in dit
spel, bereikt de ervaren speler een verlies van 22 fiches. Dit is dus een verschil van 97 fiches, wat een
129
enorm aantal is. Dit spelresultaat valt met recht een uitschieter te noemen. Om nu te stellen dat de
ervaren speler een grote blunder heeft begaan, gaat toch wat te ver. Zijn inschatting (dat iemand
inmiddels een betere hand dan zijn AA gemaakt heeft) is logisch, maar helaas ook incorrect
(aangezien niemand ook maar in de buurt kwam van zijn hand). Dit spel heeft wel een erg grote
invloed op het behendigsheidsniveau in de eerste spelronde. Tabel 10 geeft de gecorrigeerde
winsten en het gecorrigeerde behendigheidsniveau.
Belangrijke Beginner Ervaren Fictief Behendigheidsniveau
Ronde 1 -28 -11 10 0,45
Tabel 10: Gecorrigeerde winsten en behendigheidsniveau in spelronde 1
In de tweede spelronde zijn geen uitschieters te vinden. In de derde spelronde gebeurt wel iets
opmerkelijks aan tafel C. Waar de beginners aan de andere tafels rustig en passief spelen (ze
verhogen zelden), schakelen sommige spelers aan tafel C nu over op een hyper-aggressieve speelstijl.
Zij verhogen wanneer dit mogelijk is, ongeacht de kracht van hun hand. Dit werkt natuurlijk sterk in
het voordeel van de fictieve speler, die in deze ronde een recordwinst boekt van 219 fiches. Het
toevalseffect is in deze ronde erg hoog, omdat de ervaren speler niet de kans kreeg om potten van
meer dan 100 fiches te kunnen winnen.
Het complete overzicht van behendigheidsniveaus, gecorrigeerd voor spel 9 in spelronde 1, in alle
spelrondes is weergegeven in tabel 11.
Belangrijke Beginner Ervaren Fictief Behendigheidsniveau
Ronde 1 -28 -11 10 0,45
Ronde 2 30 90 161 0,46
Ronde 3 -32 1 219 0,13
Totaal -30 80 390 0,26
Tabel 11: Gecorrigeerde winsten en behendigheidsniveau in alle rondes
Laten we spelronde 3 weg, dan krijgen we als behendigheidsniveau 0,46.
130
4.5 Beginnersstrategie
De meeste beginners in dit experiment bleken achteraf toch geen echte beginners te zijn; velen van
hen spelen regelmatig poker met vrienden, en sommigen spelen zelfs om echte dollars op internet.
Toch is een classificatie als beginner op basis van het spelgedrag bij veel van deze spelers wel correct,
gezien de inschattingsfouten die al deze spelers met regelmaat maken.
Iedere beginner speelt natuurlijk op een andere manier. Enkele globale conclusies wat betreft
spelgedrag zijn zeer wel mogelijk, zeker daar de speelstijl van de meeste beginners erg vaak
overeenkomt. In meerdere spelsituaties is het zelfs zo, dat vrijwel alle beginners dezelfde actie
ondernemen, terwijl de ervaren speler het juist anders aanpakt.
Ieder spel begint met het delen van de kaarten, en meteen op dit punt verschilt de strategie van de
ervaren speler enorm met de strategie van de meeste beginners. De openingsrange van de ervaren
speler (dit is de set van alle handen waarmee hij iets in gaat zetten) is afhankelijk van zijn positie aan
tafel. Hij speelt minder handen vanuit de positie direct na de big blind (deze positie heet “under the
gun”). Dit is een slechte positie, omdat er telkens een beslissing moet worden gemaakt, voordat alle
andere spelers hun actie bekendmaken. Het is onverstandig om vanuit deze positie met slechte
kaarten in de pot te komen. Vanuit betere posities (de dealer positie is de beste positie, deze beslist
namelijk altijd als laatste, en hier kunnen dus de overige acties worden meegenomen in de
beslissingen) speelt de ervaren speler meer handen. Typische minimale openingshanden voor de
slechte tafelpositie zijn {55, AQs, AKo}, d.w.z. dat deze speler verhoogt met pairs hoger dan 44, met
AQ en AK suited en met AK off-suit. Met alle overige handen wordt gepast.
De beginner let niet op tafelpositie. Voor hem is iedere positie hetzelfde, en hij baseert dus zijn
preflop actie puur op de sterkte van zijn hand, en eventueel op de acties die er al hebben
plaatsgevonden. Het grootste verschil met de ervaren speler ligt in het percentage van de handen
waarmee gespeeld wordt. Waar de ervaren speler zonder moeite een hand als A5o past, kiezen
vrijwel alle beginners voor de call optie. Ook met slechtere handen wordt vaak voor call gekozen;
sommige beginners callen zelfs met iedere hand. Tabel 12 toont het preflop gedrag van de spelers in
dit experiment.
Speler Ronde 1 Ronde 2 Ronde 3 %
1 9 5 10 66,67
2 3 4 4 30,56
3 11 8 - 79,17
4 11 6 12 80,56
5 7 10 9 72,22
6 5 7 5 47,22
7 8 5 7 55,56
131
8 10 7 12 80,56
9 11 12 10 91,67
10 10 7 9 72,22
11 8 9 11 77,78
12 4 4 4 33,33
13 9 9 6 66,67
14 11 9 3 63,89
15 9 9 10 77,78
16 11 10 12 91,67
17 - 4 9 54,17
18 11 - 11 91,67
19 7 8 8 63,89
Beginner 71,57
Ervaren 47,22
Fictief 33,33
Tabel 12: Flops gezien
Het percentage flops dat gehaald wordt door de beginners ligt ver boven het percentage van de
ervaren speler. De fictieve speler ziet nog minder flops, maar dat is logisch. Hij past namelijk over het
algemeen bij de preflop als hij weet dat iemand anders op de river een betere hand heeft dan hij. Het
gemiddelde over alle beginners ligt op 71,57% van alle handen. Gaan we ervan uit dat de beginners
de beste 71,57% van alle handen spelen, dan komt dat ongeveer neer op de range {22, 92s, T6o}.
Met alle suited handen beter dan 92s, en met alle handen beter dan T6o wordt dus gespeeld. Dit is
nogal een groot verschil met de range van de ervaren speler, die zelfs in de beste positie een hand als
K6s (ruim binnen de 71,57%) zal passen.
Het lijkt aantrekkelijk om met veel handen naar de flop te gaan, het kost meestal immers slechts 2
fiches, terwijl de meeste potten wel boven de 40 fiches uitkomen. Toch is dit op lange termijn geen
winstgevende strategie. Investeren met T6o loont dan wel als de flop 987 brengt, of T66, in de
meeste andere gevallen wordt gewoon 2 fiches aan een andere speler gedoneerd. Voor 2 fiches kans
maken op een grote pot (van 40 fiches) klinkt mooi, maar als de kans om die pot te winnen erg klein
is, en er meestal nog wel meer fiches geïnvesteerd dienen te worden, is dit een verliesgevende
tactiek.
132
Wat ook typisch is voor beginners in poker in het algemeen, maar ook voor de beginners in het
experiment is de voorkeur voor call boven bet/raise. De eerste gedachte van deze spelers is dat ze
proberen zo goedkoop mogelijk naar de river te komen, om aldaar te beslissen of ze de beste hand
hebben. Op de river wordt met goede handen wel veel verhoogd, daarvoor echter nogal weinig. We
nemen als voorbeeld spel 12 uit de eerste spelronde aan tafel C.
Spel 12
Community Cards 5s4s5dTd9d
Seat C1 C2 C5
Speler 11 4 10
Hand Qd4d KdTs Jh5h
Preflop BB Call Call
Check
Flop Check Check Check
Turn Bet Call Raise
Call Call
River Check Check Bet
Raise Call Fold
Tabel 13: Spel 12, Tafel C, Ronde 1
De overige spelers zijn weggelaten. De speculatieve preflop call van speler 10 pakt deze keer goed
uit, hij maakt three of a kind al op de flop. De flop-actie is typisch voor een beginner. Speler 10 heeft
een ijzersterke hand na deze flop, het belangrijkste is nu om snel veel fiches in de pot te krijgen. Door
te checken zal dit echter niet lukken. Welk motief hij hiervoor ook heeft (hij is bang om de kracht van
zijn hand te verraden, of hij wil eerst de overige community cards zien), hier checken is onverstandig.
Op de flop heeft speler 10 maar liefst 91,81% kans om de pot te gaan winnen. Natuurlijk weet deze
speler niet wat de andere twee spelers op hand hebben, maar waarschijnlijk beseft hij wel dat zijn
hand nu het sterkste is. Als hij de andere spelers ertoe krijgt dat zij 2 fiches inzetten, dan zullen die in
91,81% van de gevallen aan het einde van het spel voor hem zijn. Dit is dus duidelijk een investering
met positieve verwachtingswaarde. Door te checken krijgt speler 11 gratis de kans om zijn hand te
verbeteren. Zijn winstkans gaat van 3,54% op de flop naar 16,67% op de turn, doordat de vierde
ruiten verschijnt. Speler 10 heeft nagelaten zijn sterke hand te beschermen door een bet te plaatsen
(wellicht was hij dan van speler 11 af geweest).
133
Een ander kenmerk in het spel van beginners is dat ze moeite hebben met passen. Dit kon al eerder
gezien worden op spel 9 toen de beginner met AA de pot won, waar de ervaren speler eieren voor
zijn geld koos en het spel op de turn verliet. Ook speler 3, die bij deze beginner aan tafel zat in dat
spel, kiest voor een opmerkelijke speelwijze, te zien in tabel 14.
Spel 9
Community Cards 7s8c5dQhQc
Speler 3 7 1
Hand AsJs JdJh AdAc
Preflop SB BB Bet
Call Raise Raise
Call Call
Flop Check Bet Raise
Call Call
Turn Check Bet Raise
Call Call
River Check Check Bet
Fold Raise Call
Tabel 14: Spel 9, Tafel C, Ronde 1
Preflop is AJs een prima hand, al neemt de kracht van deze hand snel af na een raise en een reraise
door de tegenstanders (tegen AK en 99 bijvoorbeeld zal AJ slechts in 24% van de gevallen winnend
zijn). Na de flop, die niets brengt voor speler 3 zou hij toch eigenlijk de handdoek in de ring moeten
gooien. Hij speelt echter rustig verder. De flop kost hem 4 fiches, de turn zelfs 8. Een ervaren speler
had hier eenvoudig 12 fiches kunnen besparen.
Het laatste verschil tussen de tactiek van een ervaren speler en een beginner ligt in het bluffen.
Bluffen is een essentieel onderdeel van het pokerspel, en kan bij juiste uitvoering veel winst
134
opleveren. Belangrijk is de timing van een bluf. Een bluf is over het algemeen goed in twee specifieke
situaties. Het eerste geval doet zich voor als een speler op dit moment van het spel een waardeloze
hand heeft, maar met de juiste community cards toch een extreem sterke hand kan maken. Een
speler die bijvoorbeeld 8h5h heeft, en de getoonde community cards zijn Kh9h3s6c, heeft op dit
moment geen goede hand, maar met een harten op de river maakt hij waarschijnlijk de beste hand.
Een bet op de turn is nu in principe een bluf, de speler ziet het liefst alle andere spelers passen.
Mocht dit echter niet lukken, dan heeft hij nog steeds kansen om de pot te winnen. We noemen zo’n
bluf een semi-bluf, omdat als de bluf niet slaagt er nog niets verloren is.
Het andere geschikte moment om te bluffen is wanneer er nog weinig tegenstanders over zijn (liefst
slechts een), die geen van allen veel kracht tonen. Als alle speler zowel op de flop als op de turn
checken, valt het te proberen te pot te winnen door middel van een bluf.
De ervaren spelers kiest de momenten waarop hij een bluf plaatst zorgvuldig uit. De meeste
beginners doen dit niet, zij analyseren de situatie niet grondig genoeg, en kiezen vaak een willekeurig
moment uit om ervoor te gaan. Zij zijn zich niet bewust van de verhouding tussen de slagingskans
van een bluf en de kosten. Als voorbeeld tabel 15.
Spel 3
Community Cards Td8h6s3c3h
Seat C2 C3 C4
Speler 16 4 9
Hand Kc9c Kh7d Th6h
Preflop SB BB
Raise Call Raise
Call Call
Flop Check Bet
Raise Call Raise
Call Call
Turn Check Bet
Raise Call Raise
Raise Call Call
135
River Check Bet
Raise Fold Raise
Call
Tabel 15: Spel 3, Tafel C, Ronde 3
Speler 16 heeft blijkbaar besloten dat hij flink wil gaan bluffen in dit spel. Op een flop die voor hem
buiten een straight draw (met een 7 maakt speler 16 een straight) totaal niet geholpen heeft, en met
twee tegenstanders die redelijk wat kracht tonen, besluit hij zowel op de flop als op de turn te
reraisen. Zelfs op de river, waar zijn straight draw gemist heeft, probeert hij nogmaals de pot te
stelen.
Aan de hand van alle spelverlopen van dit experiment kan een karakterisering van de
beginnersstrategie in Texas Hold’Em gemaakt worden. Deze karakterisering kan mogelijk in de
toekomst worden gebruikt om het behendigheidsniveau van Texas Hold’Em uit te rekenen met
behulp van simulaties. Een typische beginner in dit experiment speelt als volgt:
- Preflop: De beginner speelt ruim 70% van zijn handen, wat neerkomt op minimum openingshanden van {22, 92s, T6o}. Met deze handen wordt voor call gekozen, ongeacht tafelpositie en ongeacht de acties die al door andere spelers zijn genomen. Zelfs met een hand als J8o wordt een raise en een reraise gecalld. De beginner raiset zelf nooit op de preflop, zelfs niet met AA.
- Flop en turn: Zolang een beginner nog kansen ziet om de beste hand te maken, blijft hij in het spel. Dit betekent dat hij met een flush-draw (nog 1 kaart nodig om een flush te maken), met een straight-draw (nog 1 kaart nodig voor een straight, bijvoorbeeld KT op een Q95 flop) en zelfs met overcards (de eigen kaarten van de beginner zijn hoger dan alle kaarten op de flop) een check/call strategie verkiest. Hij checkt dus wanneer dit nog kan, en callt anders de geplaatste bets. Ook met handen zwakker dan top pair (top pair houdt in dat er een pair is gemaakt met de hoogste kaart op de flop/turn, zoals K9 op een KJ4 flop) wordt gekozen voor check/call, ongeacht het aantal raises dat al is geplaatst. Zelfs met 22 op een QT4 flop wordt een reraise nog gecalld door de beginner. Met handen die op de flop/turn top pair of beter hebben gemaakt, wordt wel verhoogd, maar niet altijd. De beginner is bang om zijn sterke hand te verraden, en is ook bang voor ongunstige kaarten op de turn en river. Als niemand nog heeft verhoogd, zal met deze handen in de helft van de gevallen voor bet worden gekozen, en in de andere helft voor check. Als een andere speler al een bet heeft gedaan, kiest de beginner altijd voor call en nooit voor raise.
- River: Met alle pairs die slechter zijn dan top pair, kiest de beginner op de river voor een check/call tactiek. Hij is niet van plan zijn hand op de geven, maar wil wel zo goedkoop mogelijk zien of hij is verslagen. Bij de top pair handen is een bet/call strategie van toepassing, de beginner zal nooit folden, maar ook nooit raisen met alleen top pair. Met alle handen die beter zijn dan top pair wordt bet/raise gespeeld. Er wordt verhoogd zolang dit mogelijk is.
136
- Bluffen en sandbagging: Deze elementen maken geen onderdeel uit van de strategie van een beginner.
4.6 De ervaren speler
De ervaren speler heeft niet erg goed gepresteerd in dit experiment. De hand waarin hij zijn paar
azen opgaf bleek een dure vergissing. Die hand illustreert dat het moeilijk is om te spelen tegen
echte beginners. Sowieso brengt het spelen tegen beginners een grote variantie met zich mee.
Omdat er meer spelers in meer potten zitten, worden de potten groter, en zal de ervaren speler
vaker een spel verliezen. Als hij tegen goede spelers speelt en voor de flop, die 7h5s2c luidt,
gereraised heeft, kan hij er wel van uit gaan dat zijn paar koningen nog steeds goed is. Tegen
beginners zal hij aan zo’n tafel meer winnen per spel, maar ook af en toe verliezen van een beginner
die met 7d2d de raise van de ervaren speler gecalld heeft.
5. Conclusies
Doordat er slechts 36 spellen gespeeld zijn door iedere speler in dit experiment, is de schatting van
het behendigheidsniveau statistisch gezien niet erg betrouwbaar. Toch komt het geschatte
behendigheidsniveau van Texas Hold’Em goed in de buurt van de theoretische waarde (tenminste
0,30). Op basis van dit experiment is er dus geen aanleiding om de theoretische conclusie te herzien.
De (gecorrigeerde) behendigheid in de eerste twee spelrondes bedraagt respectievelijk 0,45 en 0,46.
In de derde spelronde komt het behendigheidsniveau uit op 0,13.
De ervaren speler had vooral in het begin moeite zich aan te passen aan de beginners. Hun strategie
bleek wezenlijk anders dan de strategie van de speler waar onze ervaren speler het normaal
gesproken tegen opneemt. Uit het spelgedrag van de beginners hebben we een globale
karakterisering van hun strategie kunnen destilleren. Deze kan voor een verdere precisering van het
theoretische behendigheidsniveau en voor toekomstige experimenten als input fungeren.
137
6. Bijlage: Spelverlopen
Spelronde 1:
Tafel A
Spel 1
Dealer Dealer 1
Community Cards 4d5d3s3dQh
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Th8s Ks8h 9s2s Ah3c QcJc Ac5c
Preflop SB BB Call Call Call Call
Fold Check
1 2 2 2 2 2
Flop Check Bet Raise Call Fold
Fold Call
0 0 4 4 4 0
Turn Check Bet Fold
Fold
0 0 0 4 0 0
River 0 0 0 0 0 0
Total Invested 27 1 2 6 10 6 2
Stack Size 499 498 494 517 494 498
Spel 2
Community Cards 6h6cTc9hTd
138
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand 7s6s Ts3h 9d2h Ac4h Qs3c Ad9s
Preflop SB BB Call Fold Fold
Call Call Check
2 2 2 2 0 0
Flop Bet Fold Fold
Raise Call
4 4 0 0 0 0
Turn Bet
Raise Call
8 8 0 0 0 0
River Bet
Raise Raise
Call
12 12 0 0 0 0
Total Invested 56 26 26 2 2 0 0
Stack Size 473 528 492 515 494 498
Spel 3
Community Cards Kd2h5dQs8s
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand KsQh AcAh 3d3h 7d2d Ad3s 6d5s
Preflop SB BB Call Fold
Call Call Bet Fold Call
Call Call
139
4 4 4 2 4 0
Flop Fold Check
Check Check
0 0 0 0 0 0
Turn Check
Check Bet Call
Call
4 4 0 0 4 0
River Check
Bet Raise Fold
Call
8 8 0 0 0 0
Total Invested 46 16 16 4 2 8 0
Stack Size 503 512 488 513 486 498
Spel 4
Community Cards 4cKd4s7d6c
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Ac5s 5h2s Td9c Qc7s Kc7c Ah3s
Preflop SB BB Fold
Call Fold Call Call Check
2 0 2 2 2 0
Flop Check Bet
Fold Fold Call
0 0 0 2 2 0
Turn Bet Raise
140
Call
0 0 0 8 8 0
River Bet Call
0 0 0 4 4 0
Total Invested 36 2 0 2 16 16 0
Stack Size 501 512 486 497 506 498
Spel 5
Community Cards 3s6s3dAdKc
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Kh9s KsTd Ts4h Ac6h 7d4c 9h9c
Preflop SB BB
Fold Call Fold Call Call Check
0 2 0 2 2 2
Flop Check Check
Check Bet Call Fold
Call
0 2 0 2 2 0
Turn Check
Check Bet Call
Call
0 4 0 4 4 0
River Check
Check Bet Fold
Fold
0 0 4 4 0
141
Total Invested 30 0 8 0 12 8 2
Stack Size 501 504 486 515 498 496
Spel 6
Community Cards Ac5d7c3h8h
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand 6h2c As4s KcTd Ts4h QcTc Jc4c
Preflop SB
BB Call Call Call Call Fold
Check
2 2 2 2 2 1
Flop Check Check Check Check Check
0 0 0 0 0 0
Turn Check Check Check Check Check
0 0 0 0 0 0
River Check Bet Fold Fold Fold
Fold
0 4 0 0 0 0
Total Invested 15 2 6 2 2 2 1
Stack Size 499 513 484 513 496 495
Spel 7
Community Cards 6d3sKsAd3d
142
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand 9s6s KhTc Qh8d Ac5d 7h7s 3c2d
Preflop SB BB Call Call Call Call
Call Check
2 2 2 2 2 2
Flop Check Check Check Check Check Check
0 0 0 0 0 0
Turn Check Check Check Check Check Bet
Fold Call Fold Fold Fold
0 4 0 0 0 4
River Check Bet
Call
0 4 0 0 0 4
Total Invested 28 2 10 2 2 2 10
Stack Size 497 503 482 511 494 513
Spel 8
Community Cards 5cQhKd4c9c
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Qc3h 7s6s 9s4s KhTs KcTc Jh5d
Preflop SB BB Call Call Fold
Fold Call Check
0 2 2 2 2 0
Flop Check Check Check Check
143
0 0 0 0 0 0
Turn Check Check Bet Call
Fold Fold
0 0 0 4 4 0
River Bet Raise
Call
0 0 0 8 8 0
Total Invested 32 0 2 2 14 14 0
Stack Size 497 501 480 497 512 513
Spel 9
Community Cards 7s8c5dQhQc
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Jc3c 6h3h AsJs JdJh 8d3s AdAc
Preflop SB BB Fold Bet
Call Call Call Call
4 4 4 4 0 4
Flop Check Check Bet
Fold Fold Fold Raise Call
0 0 0 4 0 4
Turn Check Check
0 0 0 0 0 0
River Check Bet
Call
0 0 0 4 0 4
144
Total Invested 36 4 4 4 12 0 12
Stack Size 493 497 476 485 512 537
Spel 10
Community Cards 4c6dJc9s3c
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Jd5h 8h4s Qc8c 7s5c As4d Kd8s
Preflop SB BB Fold
Fold Call Call Call Check
0 2 2 2 2 0
Flop Check Check
Check Bet Raise Call
Fold Call
0 0 4 4 4 0
Turn Check Check
Check
0 0 0 0 0 0
River Check Check
Bet Call Call
0 0 4 4 4 0
Total Invested 32 0 2 10 10 10 0
Stack Size 493 495 498 475 502 537
Spel 11
Community Cards Jh3d2d9d5d
145
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand 4c3s As4d 5h5c 9s2c Ah8c Qc5s
Preflop SB BB
Fold Call Call Call Call Check
0 2 2 2 2 2
Flop Check Check
Bet Call Fold Call Fold
0 2 2 0 2 0
Turn Bet
Call Call
0 4 4 0 4 0
River Bet
Call Fold
0 4 0 0 4 0
Total Invested 36 0 12 8 2 12 2
Stack Size 493 519 490 473 490 535
Spel 12
Community Cards 5s4s5dTd9d
Seat A1 A2 A3 A4 A5 A6
Speler 5 16 14 18 8 6
Hand Qd4d KdTs Jc5c Jd8c Jh5h 6c3h
Preflop SB
BB Call Bet Call Call Fold
Call Call
4 4 4 4 4 1
146
Flop Check Check Bet Fold Call
Raise Fold Raise Call
Call
6 0 6 0 6 0
Turn Check Bet Raise
Call Raise Call
Call
12 0 12 0 12 0
River Check Bet Call
Raise Raise Call
Call
12 0 12 0 12 0
Total Invested 111 34 4 34 4 34 1
Stack Size 570 515 456 469 456 534
Tafel B
Spel 1
Dealer Dealer 2
Community Cards 4d5d3s3dQh
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Th8s Ks8h 9s2s Ah3c QcJc Ac5c
Preflop SB BB Fold Call Bet Fold
Call Call Call
4 4 0 4 4 0
Flop Check Check Bet Call
Fold Call
0 2 0 2 2 0
147
Turn Bet Raise Fold
Call
0 8 0 8 0 0
River Check Check
0 0 0 0 0 0
Total Invested 38 4 14 0 14 6 0
Stack Size 496 486 500 524 494 500
Spel 2
Community Cards 6h6cTc9hTd
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand 7s6s Ts3h 9d2h Ac4h Qs3c Ad9s
Preflop SB BB Call Fold Bet
Fold Fold Fold Call
0 1 2 4 0 4
Flop Check Bet
Call
0 0 0 2 0 2
Turn Check Bet
Fold
0 0 0 0 0 4
River
0 0 0 0 0 0
Total Invested 19 0 1 2 6 0 10
Stack Size 496 485 498 518 494 509
148
Spel 3
Community Cards Kd2h5dQs8s
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand KsQh AcAh 3d3h 7d2d Ad3s 6d5s
Preflop SB BB Fold Fold
Call Bet Call Call
Call
4 4 4 4 0 0
Flop Check Bet
Call Raise Call Raise
Call Call Call
6 6 6 6 0 0
Turn Check Bet
Raise Call Call Raise
Call Call Call
12 12 12 12 0 0
River Check Check
Check Check
0 0 0 0 0 0
Total Invested 88 22 22 22 22 0 0
Stack Size 562 463 476 496 494 509
Spel 4
Community Cards 4cKd4s7d6c
149
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Ac5s 5h2s Td9c Qc7s Kc7c Ah3s
Preflop SB BB Fold
Fold Fold Fold Call Check
0 0 0 2 2 0
Flop Check Bet
Fold
0 0 0 0 2 0
Turn
0 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 6 0 0 0 2 4 0
Stack Size 562 463 476 494 496 509
Spel 5
Community Cards 3s6s3dAdKc
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Kh9s KsTd Ts4h Ac6h 7d4c 9h9c
Preflop SB BB
Call Call Fold Bet Call Raise
Fold Call Call Call
0 6 0 6 6 6
Flop Check Bet
Fold Call Fold
150
0 0 0 2 0 2
Turn Bet
Call
0 0 0 4 0 4
River Bet
Call
0 0 0 4 0 4
Total Invested 46 2 6 0 16 6 16
Stack Size 560 457 476 524 490 493
Spel 6
Community Cards Ac5d7c3h8h
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand 6h2c As4s KcTd Ts4h QcTc Jc4c
Preflop SB
BB Call Call Call Bet Fold
Fold Call Call Call
2 4 4 4 4 1
Flop Check Check Bet Call
Call Call
0 2 2 2 2 0
Turn Bet Call Call Call
151
0 4 4 4 4 0
River Check Check Bet Fold
Call Fold
0 4 0 4 0 0
Total Invested 51 2 14 10 14 10 1
Stack Size 558 494 466 510 480 492
Spel 7
Community Cards 6d3sKsAd3d
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand 9s6s KhTc Qh8d Ac5d 7h7s 3c2d
Preflop SB BB Fold Call Bet Fold
Fold Call Call
1 4 0 4 4 0
Flop Check Check Bet
Call Fold
0 2 0 0 2 0
Turn Check Bet
Raise Call
0 8 0 0 8 0
River Check Bet
Call
0 4 0 0 4 0
Total Invested 41 1 18 0 4 18 0
Stack Size 557 517 466 506 462 492
Spel 8
152
Community Cards 5cQhKd4c9c
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Qc3h 7s6s 9s4s KhTs KcTc Jh5d
Preflop SB BB Call Call Fold
Call Call Check
2 2 2 2 2 0
Flop Check Check Check Bet
Fold Fold Fold Call
0 0 0 2 2 0
Turn Check Bet
Call
0 0 0 4 4 0
River Check Bet
Call
0 0 0 4 4 0
Total Invested 30 2 2 2 12 12 0
Stack Size 555 515 464 494 480 492
Spel 9
Community Cards 7s8c5dQhQc
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Jc3c 6h3h AsJs JdJh 8d3s AdAc
Preflop SB BB Fold Bet
Fold Call Raise Raise Call
Call Call
153
0 8 8 8 0 8
Flop Bet Raise Raise
Call Call Call
0 6 6 6 0 6
Turn Bet Call Raise
Raise Call Fold Fold
0 12 12 4 0 8
River Bet
Fold
0 0 4 0 0 0
Total Invested 96 0 26 30 18 0 22
Stack Size 555 489 530 476 480 470
Spel 10
Community Cards 4c6dJc9s3c
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Jd5h 8h4s Qc8c 7s5c As4d Kd8s
Preflop SB BB Fold
Fold Fold Call Call Check
0 0 2 2 2 0
Flop Check Check
Bet Call Fold
0 0 2 2 0 0
Turn Bet
Raise Call
0 0 8 8 0 0
154
River Bet
Raise Raise
Raise Call
0 0 16 16 0 0
Total Invested 58 0 0 28 28 2 0
Stack Size 555 489 560 448 478 470
Spel 11
Community Cards Jh3d2d9d5d
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand 4c3s As4d 5h5c 9s2c Ah8c Qc5s
Preflop SB BB
Fold Call Call Fold Call Check
0 2 2 0 2 2
Flop Check Check
Check Check
0 0 0 0 0 0
Turn Check Check
Bet Call Call Fold
0 4 4 0 4 0
River Bet
Call Call
0 4 4 0 4 0
Total Invested 32 0 10 10 0 10 2
Stack Size 555 511 550 448 468 468
155
Spel 12
Community Cards 5s4s5dTd9d
Seat B1 B2 B3 B4 B5 B6
Speler 2 15 19 9 13 12
Hand Qd4d KdTs Jc5c Jd8c Jh5h 6c3h
Preflop SB
BB Call Call Bet Call Fold
Fold Call Call
2 4 4 4 4 1
Flop Check Check Check Bet
Fold Call Fold
0 0 2 0 2 0
Turn Check Bet
Raise Raise
Call
0 0 12 0 12 0
River Check Bet
Call
0 0 4 0 4 0
Total Invested 55 2 4 22 4 22 1
Stack Size 553 507 556 444 473 467
Tafel C
Spel 1
Dealer Dealer 3
Community Cards 4d5d3s3dQh
156
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Th8s Ks8h 9s2s Ah3c QcJc Ac5c
Preflop SB BB Call Call Call Call
Call Check
2 2 2 2 2 2
Flop Check Check Check Bet Call Call
Fold Fold Fold
0 0 0 2 2 2
Turn Bet Fold Raise
Raise Call
0 0 0 8 0 8
River Bet Fold
0 0 0 4 0 0
Total Invested 38 2 2 2 16 4 12
Stack Size 498 498 498 522 496 488
Spel 2
Community Cards 6h6cTc9hTd
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand 7s6s Ts3h 9d2h Ac4h Qs3c Ad9s
Preflop SB BB Call Call Call
Call Call Check
2 2 2 2 2 2
Flop Bet Fold Fold Fold Fold
Call
157
2 2 0 0 0 0
Turn Check
Check
0 0 0 0 0 0
River Bet
Raise Call
8 8 0 0 0 0
Total Invested 32 12 12 2 2 2 2
Stack Size 486 518 496 520 494 486
Spel 3
Community Cards Kd2h5dQs8s
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand KsQh AcAh 3d3h 7d2d Ad3s 6d5s
Preflop SB BB Call Fold
Call Call Call Check
2 2 2 2 2 0
Flop Check Check Check
Check Bet Call Fold Fold
Call
2 2 2 0 0 0
Turn Check
Bet Raise Call
Raise Call Call
12 12 12 0 0 0
158
River Check
Bet Raise Fold
Raise Call
12 12 0 0 0 0
Total Invested 76 28 28 16 2 2 0
Stack Size 534 490 480 518 492 486
Spel 4
Community Cards 4cKd4s7d6c
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Ac5s 5h2s Td9c Qc7s Kc7c Ah3s
Preflop SB BB Call
Call Call Call Call Call
2 2 2 2 2 2
Flop Check Check Check
Check Check Check
0 0 0 0 0 0
Turn Bet Call Fold
Fold Fold Fold
0 0 0 4 4 0
River Bet Call
0 0 0 4 4 0
Total Invested 28 2 2 2 10 10 2
Stack Size 532 488 478 508 510 484
Spel 5
Community Cards 3s6s3dAdKc
159
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Kh9s KsTd Ts4h Ac6h 7d4c 9h9c
Preflop SB BB
Call Call Call Call Fold Check
2 2 2 2 1 2
Flop Bet
Fold Fold Fold Call
0 0 0 2 0 2
Turn Bet
Call
0 0 0 4 0 4
River Bet
Fold
0 0 0 0 0 4
Total Invested 27 2 2 2 8 1 12
Stack Size 530 486 476 500 509 499
Spel 6
Community Cards Ac5d7c3h8h
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand 6h2c As4s KcTd Ts4h QcTc Jc4c
160
Preflop SB
BB Call Call Fold Call Call
Call
2 2 2 0 2 2
Flop Bet
Fold Raise Fold Raise Call
Call
0 6 0 0 6 6
Turn Check
Check Check
0 0 0 0 0 0
River Check
Bet Call Fold
0 4 0 0 4 0
Total Invested 36 2 12 2 0 12 8
Stack Size 528 510 474 500 497 491
Spel 7
Community Cards 6d3sKsAd3d
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand 9s6s KhTc Qh8d Ac5d 7h7s 3c2d
Preflop SB BB Call Bet Call Fold
Call Call Call
4 4 4 4 4 0
Flop Check Bet Fold Fold Call
Call
2 2 0 0 2 0
161
Turn Check Check Check
0 0 0 0 0 0
River Bet Call Fold
4 4 0 0 0 0
Total Invested 34 10 10 4 4 6 0
Stack Size 518 534 470 496 491 491
Spel 8
Community Cards 5cQhKd4c9c
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Qc3h 7s6s 9s4s KhTs KcTc Jh5d
Preflop SB BB Call Call Call
Fold Call Check
0 2 2 2 2 2
Flop Check Check Bet Raise
Fold Fold Call
0 0 0 4 4 0
Turn Bet Call
0 0 0 4 4 0
River Bet Raise
Call
0 0 0 8 8 0
Total Invested 42 0 2 2 18 18 2
Stack Size 518 532 468 478 515 489
162
Spel 9
Community Cards 7s8c5dQhQc
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Jc3c 6h3h AsJs JdJh 8d3s AdAc
Preflop SB BB Fold Bet
Fold Call Call Raise Raise
Call Call Call
0 8 8 8 0 8
Flop Check Bet Raise
Call Call Call
0 4 4 4 0 4
Turn Check Bet Raise
Call Call Call
0 8 8 8 0 8
River Check Check Bet
Fold Fold Raise Call
0 0 0 8 0 8
Total Invested 96 0 20 20 28 0 28
Stack Size 518 512 448 450 515 557
Spel 10
Community Cards 4c6dJc9s3c
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Jd5h 8h4s Qc8c 7s5c As4d Kd8s
163
Preflop SB BB Bet
Fold Fold Call Fold Check
0 0 2 1 2 2
Flop Check Bet
Call Fold
0 0 2 0 0 2
Turn Bet
Call
0 0 4 0 0 4
River Bet
Raise Fold
0 0 8 0 0 4
Total Invested 31 0 0 16 1 2 12
Stack Size 518 512 463 449 513 545
Spel 11
Community Cards Jh3d2d9d5d
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand 4c3s As4d 5h5c 9s2c Ah8c Qc5s
Preflop SB BB
Fold Call Call Fold Call Bet
Call Call Call
0 4 4 0 4 4
Flop Check Check
Check Check
164
0 0 0 0 0 0
Turn Check Check
Bet Call Fold Fold
0 4 4 0 0 0
River Bet Raise
Call
0 8 8 0 0 0
Total Invested 40 0 16 16 0 4 4
Stack Size 518 536 447 449 509 541
Spel 12
Community Cards 5s4s5dTd9d
Seat C1 C2 C3 C4 C5 C6
Speler 11 4 3 7 10 1
Hand Qd4d KdTs Jc5c Jd8c Jh5h 6c3h
Preflop SB
BB Call Fold Fold Call Fold
Check
2 2 0 0 2 1
Flop Check Check Check
0 0 0 0 0 0
Turn Bet Call Raise
Call Call
8 8 0 0 8 0
River Check Check Bet
165
Raise Call Fold
8 8 0 0 4 0
Total Invested 51 18 18 0 0 14 1
Stack Size 551 518 447 449 495 540
Spelronde 2:
Tafel A
Spel 1
Dealer Dealer 1
Community Cards TcKsJhAs8h
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Stack Size (Start) 570 553 447 500 456 449
Hand 4h2s Th6c Qs9s 7d2h AhJd 6d2c
Preflop SB BB Call Fold Call Fold
Call Call
2 2 2 0 2 0
Flop Check Check Check Check
0 0 0 0 0 0
Turn Check Check Bet Call
Fold Fold
0 0 4 0 4 0
River Bet Call
0 0 4 0 4 0
166
Total Invested 24 2 2 10 0 10 0
Stack Size 568 551 461 500 446 449
Spel 2
Community Cards Ac8hAd3s2h
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand 4c4s 9s3d Tc6d Qs2s 5s2d 7h7d
Preflop SB BB Fold Call Call
Call Fold Call
2 1 2 0 2 2
Flop Check Check Bet
Raise Call Fold Fold
4 0 4 0 0 2
Turn Check
Bet Fold
4 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 23 10 1 6 0 2 4
Stack Size 581 550 455 500 444 445
Spel 3
Community Cards JsTd9hJc6s
Seat A1 A2 A3 A4 A5 A6
167
Speler 5 2 3 17 14 7
Hand Ac4s 5c4d 3c2s 9s7s Kd3d KsQd
Preflop SB BB Call Call
Call Fold Fold Check
2 0 1 2 2 2
Flop Check Check Bet
Call Call Call
2 0 0 2 2 2
Turn Check Check Bet
Fold Call Call
0 0 0 4 4 4
River Check Check Bet
Fold Fold
0 0 0 0 0 4
Total Invested 33 4 0 1 8 8 12
Stack Size 577 550 454 492 436 466
Spel 4
Community Cards JdJhKs4cJs
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand Tc3h Th8c Ts8d 8s5s Ad7c Jc8h
Preflop SB BB Fold
Fold Fold Call Fold Check
168
0 0 2 1 2 0
Flop Bet
Fold
0 0 0 0 2 0
Turn
0 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 7 0 0 2 1 4 0
Stack Size 577 550 452 491 439 466
Spel 5
Community Cards Jc7d8h2dJd
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand Th5h 6c2h Ah3d Qh3h Ts4c AsKs
Preflop SB BB
Call Fold Call Fold Fold Bet
Call Call
4 0 4 0 1 4
Flop Bet
Raise Fold Call
4 0 0 0 0 4
Turn Bet
Raise Raise
169
Fold
8 0 0 0 0 12
River
0 0 0 0 0 0
Total Invested 41 16 0 4 0 1 20
Stack Size 561 550 448 491 438 487
Spel 6
Community Cards 9sQh6cQc6h
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand AhTh 5h3s 8h2h 7c2d Td7d 7h4s
Preflop SB
BB Fold Fold Fold Call Fold
Call
2 0 0 0 2 1
Flop Check Bet
Fold
0 0 0 0 2 0
Turn
0 0 0 0 0 0
River
0 0 0 0 0 0
170
Total Invested 7 2 0 0 0 4 1
Stack Size 559 550 448 491 441 486
Spel 7
Community Cards 2sQsQd6hQc
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand 9s4c Kh8d AsAc 7h3s 8c6c 7c4h
Preflop SB BB Call Fold Call Fold
Fold Bet Raise Call
Call
1 6 6 0 6 0
Flop Check Check
0 0 0 0 0 0
Turn Check Bet
Call
0 4 4 0 0 0
River Check Bet
Call
0 4 4 0 0 0
Total Invested 35 1 14 14 0 6 0
Stack Size 558 536 469 491 435 486
Spel 8
Community Cards Kh7hQc5sJh
Seat A1 A2 A3 A4 A5 A6
171
Speler 5 2 3 17 14 7
Hand 8c8d Ah2d Ts4s Kd5c 4h4c Kc4d
Preflop SB BB Fold Call Fold
Call Bet Call Call
Call
4 4 4 0 4 0
Flop Bet Call Fold
Raise Call Call
4 4 4 0 0 0
Turn Check Check
Bet Call Fold
4 4 0 0 0 0
River Check
Check
0 0 0 0 0 0
Total Invested 36 12 12 8 0 4 0
Stack Size 582 524 461 491 431 486
Spel 9
Community Cards 7c9d3h4s3s
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand Qh5c 8s6d 7s2h Ts7h Td6h Ah6c
Preflop SB BB Fold Call
Call Fold Fold Check
2 0 1 2 0 2
Flop Bet Fold
172
Call
2 0 0 2 0 0
Turn Check
Bet Call
4 0 0 4 0 0
River Check
Bet Call
4 0 0 4 0 0
Total Invested 27 12 0 1 12 0 2
Stack Size 570 524 460 506 431 484
Spel 10
Community Cards 9sKs7h4sQh
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand Td8c Jc3d 6d4d Ah7d Jh9h Qc8d
Preflop SB BB Fold
Call Fold Call Call Check
2 0 2 2 2 0
Flop Check Check
Check Check
0 0 0 0 0 0
Turn Bet Fold
Fold Call
0 0 4 4 0 0
173
River Bet
Call
0 0 4 4 0 0
Total Invested 24 2 0 10 10 2 0
Stack Size 568 524 450 520 429 484
Spel 11
Community Cards 4s5s5hAhQh
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand Js4c 7d6s 6d5d Th2c Td3s 9s8c
Preflop SB BB
Fold Call Fold Fold Call Call
0 2 0 0 2 2
Flop Check Bet
Raise Fold Fold
0 4 0 0 0 2
Turn
0 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 12 0 6 0 0 2 4
Stack Size 568 530 450 520 427 480
174
Spel 12
Community Cards AhQcKh8dAs
Seat A1 A2 A3 A4 A5 A6
Speler 5 2 3 17 14 7
Hand Js4d Jd2s 5s5c 3d2h Tc9c 8c4h
Preflop SB
BB Fold Call Call Fold Fold
Check
2 0 2 2 0 1
Flop Check Check Bet
Fold Call
0 0 2 2 0 0
Turn Check Bet
Call
0 0 4 4 0 0
River Check Check
0 0 0 0 0 0
Total Invested 19 2 0 8 8 0 1
Stack Size 566 530 461 512 427 479
Tafel B
Spel 1
Dealer Dealer 2
Community Cards TcKsJhAs8h
175
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Stack Size (Start) 556 540 444 456 507 534
Hand 4h2s Th6c Qs9s 7d2h AhJd 6d2c
Preflop SB BB Bet Fold Raise Fold
Fold Call Call
1 6 6 0 6 0
Flop Check Bet Call
Fold
0 0 2 0 2 0
Turn Check Check
0 0 0 0 0 0
River Bet Call
0 0 4 0 4 0
Total Invested 31 1 6 12 0 12 0
Stack Size 555 534 463 456 495 534
Spel 2
Community Cards Ac8hAd3s2h
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand 4c4s 9s3d Tc6d Qs2s 5s2d 7h7d
Preflop SB BB Call Fold Bet
Call Fold Call Call
4 1 4 4 0 4
176
Flop Check Check Check
Check
0 0 0 0 0 0
Turn Check Check Bet
Call Fold Fold
4 0 0 0 0 4
River Bet
Call
4 0 0 0 0 4
Total Invested 33 12 1 4 4 0 12
Stack Size 543 533 459 452 495 555
Spel 3
Community Cards JsTd9hJc6s
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand Ac4s 5c4d 3c2s 9s7s Kd3d KsQd
Preflop SB BB Call Bet
Call Fold Call Call Call
4 0 4 4 4 4
Flop Check Check Check Bet
Call Fold Call Fold
2 0 0 2 0 2
Turn Check Bet
Fold Call
177
0 0 0 4 0 4
River Check Bet
Call
0 0 0 4 0 4
Total Invested 42 6 0 4 14 4 14
Stack Size 537 533 455 438 491 583
Spel 4
Community Cards JdJhKs4cJs
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand Tc3h Th8c Ts8d 8s5s Ad7c Jc8h
Preflop SB BB Bet
Fold Fold Call Call Call
0 0 4 4 4 4
Flop Check Check Bet
Call Call Fold
0 0 2 2 0 2
Turn Check Bet
Fold Fold
0 0 0 0 0 4
River
0 0 0 0 0 0
Total Invested 26 0 0 6 6 4 10
Stack Size 537 533 449 432 487 599
178
Spel 5
Community Cards Jc7d8h2dJd
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand Th5h 6c2h Ah3d Qh3h Ts4c AsKs
Preflop SB BB
Call Fold Call Fold Call Bet
Call Call Call
4 0 4 0 4 4
Flop Check Bet
Fold Call Call
0 0 2 0 2 2
Turn Check Bet
Fold Raise Call
0 0 0 0 8 8
River Check Bet
Fold
0 0 0 0 0 4
Total Invested 42 4 0 6 0 14 18
Stack Size 533 533 443 432 473 623
Spel 6
Community Cards 9sQh6cQc6h
179
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand AhTh 5h3s 8h2h 7c2d Td7d 7h4s
Preflop SB
BB Fold Call Fold Call Bet
Call Call Call
4 0 4 0 4 4
Flop Check
Check Check Check
0 0 0 0 0 0
Turn Check
Bet Fold Fold Fold
4 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 20 8 0 4 0 4 4
Stack Size 545 533 439 432 469 619
Spel 7
Community Cards 2sQsQd6hQc
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand 9s4c Kh8d AsAc 7h3s 8c6c 7c4h
Preflop SB BB Bet Fold Call Fold
Fold Raise Call Call
180
1 6 6 0 6 0
Flop Bet Raise Fold
Call
0 4 4 0 0 0
Turn Check Bet
Fold
0 0 4 0 0 0
River
0 0 0 0 0 0
Total Invested 31 1 10 14 0 6 0
Stack Size 544 523 456 432 463 619
Spel 8
Community Cards Kh7hQc5sJh
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand 8c8d Ah2d Ts4s Kd5c 4h4c Kc4d
Preflop SB BB Call Bet Fold
Call Call Call Call
4 4 4 4 4 0
Flop Bet Fold Fold Raise
Call Call
4 4 0 0 4 0
Turn Check Check
Bet Fold Call
4 0 0 0 4 0
181
River Check
Check
0 0 0 0 0 0
Total Invested 40 12 8 4 4 12 0
Stack Size 572 515 452 428 451 619
Spel 9
Community Cards 7c9d3h4s3s
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand Qh5c 8s6d 7s2h Ts7h Td6h Ah6c
Preflop SB BB Fold Fold
Call Bet Call Call
Call
4 4 4 4 0 0
Flop Check Check
Check Bet Call Call
Fold
0 2 2 2 0 0
Turn Check Check
Bet Call Call
0 4 4 4 0 0
River Check Bet
Fold Fold
0 0 0 4 0 0
Total Invested 38 4 10 10 14 0 0
Stack Size 568 505 442 452 451 619
182
Spel 10
Community Cards 9sKs7h4sQh
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand Td8c Jc3d 6d4d Ah7d Jh9h Qc8d
Preflop SB BB Bet
Fold Fold Raise Call Call Call
0 0 6 6 6 6
Flop Bet Call Raise
Call Call Call
0 0 4 4 4 4
Turn Bet Raise Raise
Fold Call Call
0 0 0 12 12 12
River Check Check Bet
Fold Call
0 0 0 0 4 4
Total Invested 84 0 0 10 22 26 26
Stack Size 568 505 432 430 425 677
Spel 11
Community Cards 4s5s5hAhQh
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
183
Hand Js4c 7d6s 6d5d Th2c Td3s 9s8c
Preflop SB BB
Call Bet Raise Fold Fold Fold
Call Call
6 6 6 0 1 2
Flop Check Bet Raise
Call Call
4 4 4 0 0 0
Turn Check Bet Raise
Fold Call
0 8 8 0 0 0
River Bet Call
0 4 4 0 0 0
Total Invested 57 10 22 22 0 1 2
Stack Size 558 483 467 430 424 675
Spel 12
Community Cards AhQcKh8dAs
Seat B1 B2 B3 B4 B5 B6
Speler 19 1 9 8 15 6
Hand Js4d Jd2s 5s5c 3d2h Tc9c 8c4h
Preflop SB
BB Fold Call Call Call Call
Check
2 0 2 2 2 2
Flop Check
Check Check Check Check
184
0 0 0 0 0 0
Turn Bet
Fold Call Fold Call
0 0 4 0 4 4
River Bet
Call Fold
0 0 4 0 0 4
Total Invested 30 2 0 10 2 6 10
Stack Size 556 483 457 428 418 695
Tafel C
Spel 1
Dealer Dealer 3
Community Cards TcKsJhAs8h
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Stack Size (Start) 515 518 495 551 473 467
Hand 4h2s Th6c Qs9s 7d2h AhJd 6d2c
Preflop SB BB Call Fold Bet Fold
Fold Call Call
1 4 4 0 4 0
Flop Check Check Bet
Call Call
0 2 2 0 2 0
Turn Check Bet Raise
185
Fold Call
0 0 8 0 8 0
River Check Bet
Raise Call
0 0 8 0 8 0
Total Invested 51 1 6 22 0 22 0
Stack Size 514 512 524 551 451 467
Spel 2
Community Cards Ac8hAd3s2h
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand 4c4s 9s3d Tc6d Qs2s 5s2d 7h7d
Preflop SB BB Call Fold Bet
Call Fold Fold Call
4 1 2 4 0 4
Flop Check Bet
Call Fold
2 0 0 0 0 2
Turn Check
Bet Call
4 0 0 0 0 4
River Check
Bet Call
8 0 0 0 0 8
Total Invested 43 18 1 2 4 0 18
Stack Size 496 511 522 547 451 492
186
Spel 3
Community Cards JsTd9hJc6s
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Ac4s 5c4d 3c2s 9s7s Kd3d KsQd
Preflop SB BB Call Call
Call Bet Fold Call Call Call
Call
4 4 1 4 4 4
Flop Check Check Bet
Raise Call Call Call Call
4 4 0 4 4 4
Check Check Bet
Turn Fold Fold Call Call
0 0 0 4 4 4
River Check Check Bet
Call Fold
0 0 0 4 0 4
Total Invested 65 8 8 1 16 16 16
Stack Size 488 503 521 531 435 541
Spel 4
Community Cards JdJhKs4cJs
187
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Tc3h Th8c Ts8d 8s5s Ad7c Jc8h
Preflop SB BB Fold
Call Call Fold Call Call
2 2 0 2 2 0
Flop Check Check
Check Check
0 0 0 0 0 0
Turn Check Check
Check Check
0 0 0 0 0 0
River Bet Fold
Fold Call
0 4 0 4 0 0
Total Invested 16 2 6 0 6 2 0
Stack Size 486 513 521 525 433 541
Spel 5
Community Cards Jc7d8h2dJd
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Th5h 6c2h Ah3d Qh3h Ts4c AsKs
Preflop SB BB
Call Fold Call Call Call Bet
Call Call Call Call
4 0 4 4 4 4
Flop Check Check
188
Check Check Check
0 0 0 0 0 0
Turn Check Bet
Call Fold Fold Fold
4 0 0 0 0 4
River Check
Check
0 0 0 0 0 0
Total Invested 28 8 0 4 4 4 8
Stack Size 478 513 517 521 429 561
Spel 6
Community Cards 9sQh6cQc6h
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand AhTh 5h3s 8h2h 7c2d Td7d 7h4s
Preflop SB
BB Fold Fold Fold Call Fold
Bet Call
4 0 0 0 4 1
Flop Bet Call
2 0 0 0 2 0
Turn Bet Call
189
4 0 0 0 4 0
River Bet Fold
4 0 0 0 0 0
Total Invested 25 14 0 0 0 10 1
Stack Size 489 513 517 521 419 560
Spel 7
Community Cards 2sQsQd6hQc
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand 9s4c Kh8d AsAc 7h3s 8c6c 7c4h
Preflop SB BB Bet Fold Fold Fold
Call Call
4 4 4 0 0 0
Flop Check Check Bet
Fold Call
0 2 2 0 0 0
Turn Check Check
0 0 0 0 0 0
River Check Bet
Call
0 4 4 0 0 0
Total Invested 24 4 10 10 0 0 0
Stack Size 485 503 531 521 419 560
190
Spel 8
Community Cards Kh7hQc5sJh
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand 8c8d Ah2d Ts4s Kd5c 4h4c Kc4d
Preflop SB BB Call Call Fold
Call Call Check
2 2 2 2 2 0
Flop Check Check Bet Fold
Fold Fold Fold
0 0 0 2 0 0
Turn
0 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 12 2 2 2 4 2 0
Stack Size 483 501 529 529 417 560
Spel 9
Community Cards 7c9d3h4s3s
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Qh5c 8s6d 7s2h Ts7h Td6h Ah6c
191
Preflop SB BB Fold Fold
Call Fold Fold Bet
Call
4 0 1 4 0 0
Flop Bet
Fold
0 0 0 2 0 0
Turn
0 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 11 4 0 1 6 0 0
Stack Size 479 501 528 534 417 560
Spel 10
Community Cards 9sKs7h4sQh
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Td8c Jc3d 6d4d Ah7d Jh9h Qc8d
Preflop SB BB Fold
Call Fold Call Bet Call
Call Call
4 0 4 4 4 0
Flop Check Check
Bet Fold Call Call
192
2 0 0 2 2 0
Turn Check Check
Bet Call Fold
4 0 0 4 0 0
River Check
Bet Call
4 0 0 4 0 0
Total Invested 38 14 0 4 14 6 0
Stack Size 465 501 524 558 411 560
Spel 11
Community Cards 4s5s5hAhQh
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Js4c 7d6s 6d5d Th2c Td3s 9s8c
Preflop SB BB
Fold Call Call Call Call Check
0 2 2 2 2 2
Flop Check Check
Check Check Check
0 0 0 0 0 0
Turn Check Check
Check Check Check
0 0 0 0 0 0
River Check Check
193
Check Check Check
0 0 0 0 0 0
Total Invested 10 0 2 2 2 2 2
Stack Size 465 499 532 556 409 558
Spel 12
Community Cards AhQcKh8dAs
Seat C1 C2 C3 C4 C5 C6
Speler 16 4 10 11 13 12
Hand Js4d Jd2s 5s5c 3d2h Tc9c 8c4h
Preflop SB
BB Fold Call Call Call Fold
Check
2 0 2 2 2 1
Flop Check Check Check Check
0 0 0 0 0 0
Turn Check Check Check Check
0 0 0 0 0 0
River Check Check Check Check
0 0 0 0 0 0
Total Invested 9 2 0 2 2 2 1
Stack Size 463 499 539 554 407 557
194
Spelronde 3:
Tafel A
Spel 1
Dealer Dealer 1
Community Cards QhQs7sKhKs
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Stack Size (Start) 418 512 566 407 554 557
Hand JhTh 6c3h Qd4d 6h2c 7c5d Qc9d
Preflop SB BB Call Fold Call Fold
Call Check
2 2 2 0 2 0
Flop Check Check Check Check
0 0 0 0 0 0
Turn Check Check Bet Call
Fold Fold
0 0 4 0 4 0
River Bet Call
0 0 4 0 4 0
Total Invested 24 2 2 10 0 10 0
Stack Size 416 510 580 407 544 557
Spel 2
Community Cards Td3s3h6sKs
195
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand QcTc 4c2s 8c7h 4h3c 4s2h Ah3d
Preflop SB BB Call Fold Fold
Bet Fold Call Call
4 1 4 4 0 0
Flop Check Check
Bet Fold Call
2 0 0 2 0 0
Turn Bet
Raise Call
8 0 0 8 0 0
River Bet
Call
4 0 0 4 0 0
Total Invested 41 18 1 4 18 0 0
Stack Size 398 509 576 430 544 557
Spel 3
Community Cards Td8h6s3c3h
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand Qh5h Kc9c Kh7d Th6h Tc9s Qc7s
196
Preflop SB BB Call Fold
Call Call Call Check
2 2 2 2 2 0
Flop Check Bet Call
Fold Raise Fold Raise Call
Call
0 6 0 6 6 0
Turn Bet Call
Call
0 4 0 4 4 0
Bet Call
River Raise Call Call
0 8 0 8 8 0
Total Invested 64 2 20 2 20 20 0
Stack Size 396 489 574 474 524 557
Spel 4
Community Cards Ks7cQhJcTd
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand Js8c AhTc Ad4c AcKh Kc3d Th2d
Preflop SB BB Fold
Call Call Call Bet Call
Call Call Call
4 4 4 4 4 0
197
Flop Bet Call
Fold Raise Call Call Call
0 4 4 4 4 0
Turn Bet Call
Raise Call Call Call
0 8 8 8 8 0
River Bet Fold
Raise Raise Raise
Call Call
0 16 16 16 0 0
Total Invested 116 4 32 32 32 16 0
Stack Size 392 496 580 481 508 557
Spel 5
Community Cards 8dAs4dAh9h
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand 4h2d 5c2c KcQh 7s3d 6h5d AcJs
Preflop SB BB
Fold Call Call Fold Call Bet
Call Call Call
0 4 4 0 4 4
Flop Check Bet
Call Call Call
0 2 2 0 2 2
198
Turn Check Bet
Call Call Call
0 4 4 0 4 4
River Check Bet
Fold Call Fold
0 0 4 0 0 4
Total Invested 48 0 10 14 0 10 14
Stack Size 392 486 566 481 498 591
Spel 6
Community Cards 2c9sTh5s4h
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand QhJh As8d 8h7h 3h2d 8s4c Tc4d
Preflop SB
BB Call Call Fold Call Fold
Bet Call Call Call
4 4 4 0 4 1
Flop Check Check Check Check
0 0 0 0 0 0
Turn Bet Fold Call Fold
199
4 0 4 0 0 0
River Bet Call
4 0 4 0 0 0
Total Invested 33 12 4 12 0 4 1
Stack Size 413 482 554 481 494 590
Spel 7
Community Cards Kd7dThQc4h
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand 6h2c Ah8c Js6c 5h4s Jh3d 8s3h
Preflop SB BB Fold Call Call Fold
Call Check
2 2 0 2 2 0
Flop Check Check Check Check
0 0 0 0 0 0
Turn Check Check Check Check
0 0 0 0 0 0
River Check Check Check Bet
Fold Fold Fold
0 0 0 0 4 0
Total Invested 12 2 2 0 2 6 0
Stack Size 411 480 554 479 500 590
200
Spel 8
Community Cards 7h5d9d6h6d
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand Jc9s KhKs 8s2d 7s5c Jh4s QhQd
Preflop SB BB Fold Call Bet
Raise Call Fold Fold Call Raise
Call Call Call
8 8 2 0 8 8
Flop Bet Fold Call
Raise Raise Call
Call
12 12 0 0 0 12
Turn Bet Fold
Call
4 4 0 0 0 0
River Bet
Call
4 4 0 0 0 0
Total Invested 86 28 28 2 0 8 20
Stack Size 383 538 552 479 492 570
Spel 9
Community Cards 2cKdQd4d6s
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
201
Hand 4s4h 7s3c 9h8c Th3s 8s5h Kc6c
Preflop SB BB Call Fold
Call Fold Fold Check
2 0 1 2 2 0
Flop Check Check
Check
0 0 0 0 0 0
Turn Check Check
Bet Fold Fold
4 0 0 0 0 0
River
0 0 0 0 0 0
Total Invested 11 6 0 1 2 2 0
Stack Size 388 538 551 477 490 570
Spel 10
Community Cards 6s2sKs7hJh
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand 8h6h Ad7d 7c5s 4s2h AcAh 5d2d
Preflop SB BB Fold
Call Call Call Fold Bet
Call Call Call
4 4 4 1 4 0
202
Flop Bet
Raise Fold Call Raise
Call Call
6 0 6 0 6 0
Turn Bet
Raise Call Call
8 0 8 0 8 0
River Bet
Call Fold
4 0 0 0 4 0
Total Invested 67 22 4 18 1 22 0
Stack Size 366 534 533 476 535 570
Spel 11
Community Cards Jd3d7s8cKd
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand 5h3c 6c5s AcKc 9s3s 4c2s Kh2c
Preflop SB BB
Call Call Call Fold Call Check
2 2 2 0 2 2
Flop Check Check
Check Check Bet Fold Fold
Call Call
2 2 2 0 0 0
203
Turn Bet Call Call
4 4 4 0 0 0
River Check Check Bet
Fold Fold
0 0 4 0 0 0
Total Invested 32 8 8 12 0 2 2
Stack Size 358 526 553 476 533 568
Spel 12
Community Cards Ac4hAd6h2c
Seat A1 A2 A3 A4 A5 A6
Speler 15 17 5 13 11 12
Hand 8d3s 4s3c KhKd 9s7c 6c2d QcQs
Preflop SB
BB Fold Bet Call Call Raise
Fold Raise Call Call Call
2 0 6 6 6 6
Flop Check
Check Check Check
0 0 0 0 0 0
Turn Check
Check Check Check
0 0 0 0 0 0
River Check
Bet Fold Fold Call
204
0 0 4 0 0 4
Total Invested 34 2 0 10 6 6 10
Stack Size 356 526 577 470 527 558
Tafel B
Spel 1
Dealer Dealer 3
Community Cards QhQs7sKhKs
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Stack Size (Start) 530 556 428 469 479 539
Hand JhTh 6c3h Qd4d 6h2c 7c5d Qc9d
Preflop SB BB Call Bet Fold Call
Call Call Call
4 4 4 4 0 4
Flop Check Check Bet Fold Call
Fold Fold
0 0 2 0 0 2
Turn Bet Call
0 0 4 0 0 4
River Bet Call
0 0 4 0 0 4
Total Invested 40 4 4 14 4 0 14
Stack Size 526 552 434 465 479 545
205
Spel 2
Community Cards Td3s3h6sKs
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand QcTc 4c2s 8c7h 4h3c 4s2h Ah3d
Preflop SB BB Call Fold Call
Fold Fold Check
0 1 2 2 0 2
Flop Check Bet Fold
Call
0 0 2 2 0 0
Turn Check Bet
Call
0 0 4 4 0 0
River Check Bet
Call
0 0 4 4 0 0
Total Invested 27 0 1 12 12 0 2
Stack Size 526 551 422 480 479 543
Spel 3
Community Cards Td8h6s3c3h
Seat B1 B2 B3 B4 B5 B6
206
Speler 2 19 8 18 7 10
Hand Qh5h Kc9c Kh7d Th6h Tc9s Qc7s
Preflop SB BB Call Call
Fold Call Call Check
0 2 2 2 2 2
Flop Check Check Bet Call
Call Call Call
0 2 2 2 2 2
Check Bet Call Call
Turn Call Fold
0 4 0 4 4 4
River Bet Call Fold
Fold
0 0 0 4 4 0
Total Invested 44 0 8 4 12 12 8
Stack Size 526 543 418 512 467 535
Spel 4
Community Cards Ks7cQhJcTd
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand Js8c AhTc Ad4c AcKh Kc3d Th2d
Preflop SB BB Fold
Fold Call Call Call Check
207
0 2 2 2 2 0
Flop Bet Raise
Fold Call Raise Raise
Call Call
0 0 8 8 8 0
Turn Bet Call
Raise Call Call
0 0 8 8 8 0
River Bet Fold
Raise Raise
Raise Call
0 0 32 32 0 0
Total Invested 120 0 2 50 50 18 0
Stack Size 526 541 428 522 449 535
Spel 5
Community Cards 8dAs4dAh9h
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand 4h2d 5c2c KcQh 7s3d 6h5d AcJs
Preflop SB BB
Fold Fold Call Fold Call Call
0 0 2 0 2 2
Flop Bet Call
Call
208
0 0 2 0 2 2
Turn Bet Call
Call
0 0 4 0 4 4
River Check Check
Bet Fold Call
0 0 4 0 0 4
Total Invested 32 0 0 12 0 8 12
Stack Size 526 541 416 522 441 555
Spel 6
Community Cards 2c9sTh5s4h
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand QhJh As8d 8h7h 3h2d 8s4c Tc4d
Preflop SB
BB Call Call Call Fold Call
Check
2 2 2 2 0 2
Flop Check
Check Check Bet Fold Call
Call Call
2 2 2 0 0 2
209
Turn Check
Check Check Bet Fold
Call Call
4 4 4 0 0 0
River Check Check Bet
Fold Fold
0 0 4 0 0 0
Total Invested 34 8 8 12 2 0 4
Stack Size 518 533 438 520 441 551
Spel 7
Community Cards Kd7dThQc4h
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand 6h2c Ah8c Js6c 5h4s Jh3d 8s3h
Preflop SB BB Call Call Call Fold
Fold Check
1 2 2 2 2 0
Flop Bet Call Fold Fold
0 2 2 0 0 0
Turn Bet Call
0 4 4 0 0 0
River Bet Call
0 4 4 0 0 0
Total Invested 29 1 12 12 2 2 0
210
Stack Size 517 550 426 518 439 551
Spel 8
Community Cards 7h5d9d6h6d
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand Jc9s KhKs 8s2d 7s5c Jh4s QhQd
Preflop SB BB Call Call Call
Fold Call Check
0 2 2 2 2 2
Flop Check Bet Call Fold Call
Call
0 2 2 2 0 2
Turn Check Bet Raise Fold
Fold Raise Call
0 0 8 8 0 0
River Bet Call
0 0 4 4 0 0
Total Invested 42 0 4 16 16 2 4
Stack Size 517 546 452 502 437 547
Spel 9
Community Cards 2cKdQd4d6s
211
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand 4s4h 7s3c 9h8c Th3s 8s5h Kc6c
Preflop SB BB Fold Call
Call Fold Call Check
2 0 2 2 0 2
Flop Check Check Check
Bet Fold Fold Call
2 0 0 0 0 2
Turn Check
Bet Call
4 0 0 0 0 4
River Check
Bet Call
4 0 0 0 0 4
Total Invested 28 12 0 2 2 0 12
Stack Size 533 546 450 500 437 535
Spel 10
Community Cards 6s2sKs7hJh
Seat
Speler 2 19 8 18 7 10
Hand 8h6h Ad7d 7c5s 4s2h AcAh 5d2d
Preflop SB BB Fold
Fold Call Call Call Check
212
0 2 2 2 2 0
Flop Check Bet
Call Call Fold
0 2 2 0 2 0
Bet
Turn Call Raise Raise
Call Call
0 12 12 0 12 0
River Bet
Call Call
0 4 4 0 4 0
Total Invested 62 0 20 20 2 20 0
Stack Size 533 526 430 498 479 535
Spel 11
Community Cards Jd3d7s8cKd
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand 5h3c 6c5s AcKc 9s3s 4c2s Kh2c
Preflop SB BB
Fold Fold Call Bet Fold Call
Call
0 0 4 4 1 4
Flop Check
Bet Call Call
213
0 0 4 4 0 4
Turn Check
Check Check
0 0 0 0 0 0
River Check
Bet Fold Call
0 0 4 0 0 4
Total Invested 33 0 0 12 8 1 12
Stack Size 533 526 451 490 478 523
Spel 12
Community Cards Ac4hAd6h2c
Seat B1 B2 B3 B4 B5 B6
Speler 2 19 8 18 7 10
Hand 8d3s 4s3c KhKd 9s7c 6c2d QcQs
Preflop SB
BB Call Call Call Call Call
Check
2 2 2 2 2 2
Flop Check
Check Check Bet Fold Fold Call
Fold Fold
0 0 2 0 0 2
Turn Check
Bet Raise
Call
0 0 8 0 0 8
214
River Check
Bet Call
0 0 4 0 0 4
Total Invested 40 2 2 16 2 2 16
Stack Size 531 524 475 488 476 507
Tafel C
Spel 1
Dealer Dealer 2
Community Cards QhQs7sKhKs
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Stack Size (Start) 483 463 499 457 427 695
Hand JhTh 6c3h Qd4d 6h2c 7c5d Qc9d
Preflop SB BB Call Fold Fold Bet
Raise Call Call Call
6 6 6 0 0 6
Flop Check Check Bet Raise
Fold Fold Raise Call
0 0 6 0 0 6
Turn Check Bet
Raise Call
0 0 12 0 0 12
River Bet Raise
Call
0 0 12 0 0 12
Total Invested 84 6 6 36 0 0 36
215
Stack Size 477 457 505 457 427 701
Spel 2
Community Cards Td3s3h6sKs
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand QcTc 4c2s 8c7h 4h3c 4s2h Ah3d
Preflop SB BB Call Fold Bet
Raise Call Call Call Raise
Call Call Call Call
8 8 8 8 0 8
Flop Check Check Bet Call
Raise Fold Fold Raise Raise
Call Call
8 0 0 8 0 8
Turn Bet Raise
Raise Raise Call
Call
16 0 0 16 0 16
River Bet Raise
Fold Raise Raise
Call
0 0 0 16 0 16
Total Invested 144 32 8 8 48 0 48
Stack Size 445 449 497 409 427 797
Spel 3
Community Cards Td8h6s3c3h
216
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand Qh5h Kc9c Kh7d Th6h Tc9s Qc7s
Preflop SB BB Fold Fold
Bet Raise Call Raise
Call Call Call
6 6 6 6 0 0
Flop Check Bet
Fold Raise Call Raise
Call Call
0 6 6 6 0 0
Turn Check Bet
Raise Call Raise
Raise Call Call
0 16 16 16 0 0
River Check Bet
Raise Fold Raise
Call
0 12 0 12 0 0
Total Invested 114 6 40 28 40 0 0
Stack Size 439 409 469 483 427 797
Spel 4
Community Cards Ks7cQhJcTd
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand Js8c AhTc Ad4c AcKh Kc3d Th2d
217
Preflop SB BB Fold
Bet Raise Call Raise Call
Call Call Call
8 8 8 8 8 0
Flop Bet Call
Fold Raise Call Raise Call
Call Call
0 6 6 6 6 0
Turn Bet Call
Raise Call Raise Call
Raise Call Call Call
River 0 16 16 16 16 0
Bet Fold
Raise Raise Raise
Call Call Call
0 16 16 16 0 0
Total Invested 176 8 46 46 46 30 0
Stack Size 431 422 481 496 397 797
Spel 5
Community Cards 8dAs4dAh9h
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand 4h2d 5c2c KcQh 7s3d 6h5d AcJs
Preflop SB BB
Fold Bet Call Fold Call Raise
Call Call Call
0 6 6 0 6 6
218
Flop Check Bet
Raise Fold Fold Raise
Raise Call
0 8 0 0 0 8
Turn Bet
Raise Raise
Call
0 12 0 0 0 12
River Bet
Fold
0 0 0 0 0 4
Total Invested 68 0 26 6 0 6 30
Stack Size 431 396 475 496 391 835
Spel 6
Community Cards 2c9sTh5s4h
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand QhJh As8d 8h7h 3h2d 8s4c Tc4d
Preflop SB
BB Call Call Call Fold Bet
Raise Call Call Call Raise
Call Call Call Call
8 8 8 8 0 8
Flop Bet
Raise Call Call Call Raise
219
Raise Call Call Call Call
8 8 8 8 0 8
Turn Bet
Raise Fold Call Fold Raise
Call Call
12 0 12 0 0 12
River Bet
Fold Fold
0 0 0 0 0 4
Total Invested 120 28 16 28 16 0 32
Stack Size 403 380 447 480 391 923
Spel 7
Community Cards Kd7dThQc4h
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand 6h2c Ah8c Js6c 5h4s Jh3d 8s3h
Preflop SB BB Call Call Fold Fold
Fold Check
1 2 2 2 0 0
Flop Check Check Check
0 0 0 0 0 0
Turn Check Check Check
0 0 0 0 0 0
River Check Check Check
220
0 0 0 0 0 0
Total Invested 7 1 2 2 2 0 0
Stack Size 402 378 452 478 391 923
Spel 8
Community Cards 7h5d9d6h6d
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand Jc9s KhKs 8s2d 7s5c Jh4s QhQd
Preflop SB BB Bet Fold Fold
Raise Raise Call Call
8 8 8 8 0 0
Flop Bet Call Raise
Raise Raise Call Call
Call
8 8 8 8 0 0
Turn Bet Call Raise
Raise Raise Call Call
Call
16 16 16 16 0 0
River Check Check Bet
Raise Call Call
8 8 8 8 0 0
Total Invested 160 40 40 40 40 0 0
Stack Size 362 338 412 598 391 923
Spel 9
221
Community Cards 2cKdQd4d6s
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand 4s4h 7s3c 9h8c Th3s 8s5h Kc6c
Preflop SB BB Fold Call
Bet Call Call Call Call
4 4 4 4 0 4
Flop Check Bet Call
Raise Fold Call Raise Fold
Call Call
6 0 6 6 0 2
Turn Check Bet
Raise Call Raise
Raise Call Call
16 0 16 16 0 0
River Check Bet
Raise Fold Fold
8 0 0 4 0 0
Total Invested 100 34 4 26 30 0 6
Stack Size 428 334 386 568 391 917
Spel 10
Community Cards 6s2sKs7hJh
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand 8h6h Ad7d 7c5s 4s2h AcAh 5d2d
222
Preflop SB BB Fold
Bet Raise Call Call Raise
Call Call
8 8 8 8 8 0
Flop Check Bet
Fold Raise Call Call Raise
Call Call Call
0 6 6 6 6 0
Turn Check Bet
Raise Call Fold Raise
Raise Call Call
0 16 16 0 16 0
River Bet
Raise Fold Raise
Call
0 12 0 0 12 0
Total Invested 136 8 42 30 14 42 0
Stack Size 420 292 356 554 485 917
Spel 11
Community Cards Jd3d7s8cKd
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand 5h3c 6c5s AcKc 9s3s 4c2s Kh2c
Preflop SB BB
Bet Call Call Call Fold Fold
4 4 4 4 1 2
223
Flop Bet Call Call Fold
2 2 2 0 0 0
Turn Bet Call Call
4 4 4 0 0 0
River Check Check Bet
Fold Fold
0 0 4 0 0 0
Total Invested 41 10 10 14 4 1 2
Stack Size 410 282 383 550 484 915
Spel 12
Community Cards Ac4hAd6h2c
Seat C1 C2 C3 C4 C5 C6
Speler 1 16 4 9 14 6
Hand 8d3s 4s3c KhKd 9s7c 6c2d QcQs
Preflop SB
BB Bet Call Call Fold Fold
Raise Raise Call Call
8 8 8 8 0 1
Flop Bet Raise Call Fold
Raise Raise Call
8 8 8 0 0 0
Turn Bet Raise Call
Raise Raise Call
224
16 16 16 0 0 0
River Bet Call Call
4 4 4 0 0 0
Total Invested 117 36 36 36 8 0 1
Stack Size 374 246 464 542 484 914
225
7. Bijlage: Resultaten per positie
Ronde 1
Tafel
Positie A B C
1 70 53 51
2 15 7 18
3 -44 56 -53
4 -31 -56 -51
5 -44 -27 -5
6 34 -33 40
Ronde 2
Tafel
Positie A B C
1 -4 0 -52
2 -23 -57 -19
3 14 13 44
4 12 -28 3
5 -29 -89 -66
6 30 161 90
Ronde 3
Tafel
Positie A B C
1 -62 1 -109
226
2 14 -32 -217
3 11 47 -35
4 63 19 85
5 -27 -3 57
6 1 -32 219