Upload
msphix
View
226
Download
0
Embed Size (px)
Citation preview
8/20/2019 Enclave First Issue
1/38
The EnclaveA PREMIUM ENGLISH LANGUAGE GO MAGAZINE
바둑
8/20/2019 Enclave First Issue
2/38
As the first year of The Enclave draws to a close, I find my-self looking back at some fond memories. The Enclave hasturned out to be everything I wanted it to be, with some minorsetbacks , and some very unexpected developments here and
there. When the room started, there was just me, and sometimesmy brother for moral support. In a few days, we’d managed tospread word among our friends, and it grew. Within just a fewdays we had doubled our numbers to four! Then, as we gainedmembers, we started to gain volunteers. Without the help
of our volunteers, we may never have le� the proverbial ground. When it becameclear that we would grow larger, I began looking for stronger teachers. The ones weeventually chose were Shygost, an excellent teacher, and Baousai, also an excellentteacher. With the advent of lessons from Shygost and Baousai, the room almostliterally exploded. Within weeks, the Enclave was the single largest social room onKGS outside of Computer Go. As the numbers grew, so did my ambition. We had some wonderful programsgoing for a while, and the room gained recognition from some of the beer knownKGS users, including the admins, whom I was constantly pestering with questions.Alas, a real life change le� me with the hard decision to leave the Enclave for the realworld for a while. It took a few months, but I’ve now made my return. And I’m so happy I have. Through the Enclave, I’ve met people I never wouldhave, and made some friends that have changed the way I look at the world, politics,
the internet, and life as a whole. I’ve passed on learning to people I may never get toknow, and I’ve even helped people become beer players than I am. When I first founded the Enclave, I thought I might enjoy doing it for a fewmonths, then I could let it go and it would slowly erode and I would be forgoenand go back to my niche in KGS and be happy. But instead of eroding, the roomkeeps going, instead of being just a room where people come to learn, the Enclavehas turned into a kind of extended family for me, where I don’t just help people withGo, but I get to know them as people and we become friends. While my purpose in founding the Enclave hasn’t changed, the way I do it andmy priorities for it have. I’ve learned that Go is not everything, even if it is an obses-
sion. As someone said just yesterday, ‘I enjoy the game while I make friends.’ Remem- ber, though Go may be a deep road, it is an empty one without friends. And that’s why you, and all of your fellow Enclavers hold such a special placein my heart. I may never meet you outside of the computer screen, but your wordsand your actions have touched lives thousands of miles away. Thank you all for everything you have done. Here’s to another great year, formany years to come.
~Eric Dunham
A Letter from Our Founder
8/20/2019 Enclave First Issue
3/38
2 A Letter From Our Founder
4 Beginner Lessons — The Object is to Take Control
8 Beginner Life and Death
9 Amateur Life and Death
12 From Heights Above—Alexander Dinerchtein, 3p
17 Go: A Matter of Life and Death
22 Insights
28 A History of Go—Sunjang Special
31 A Picture Worth a Thousand Moves
30 Marketplace
Index
Features
Man vs. Machin
World 9x9 Go
Championships p
Articles
5 Go in a Land of Ice and Snow
10 Future History
18 Man vs. Machine
37 2007 Volunteers and Contributors
Go in a Land of IceSnow:
The Norwegian
Championships
4 Beginner Lessons
8 Beginner Life and Death
9 Amateur Life and Death
12 From Heights Above — Alexander Dinerchtein, 3p
17 A Matter of Life and Death
22 Insights
32 A History of Go
35 A Picture Worth a Thousand Moves
36 Marketplace
From Heights Abo
Alexander Dinerch
Russian 3p p. 1
8/20/2019 Enclave First Issue
4/38
Beginner Lessons
Within the first few pages of his book, Go - A Complete Introduction to the
Game, Cho Chikun, 9p, makes the point: “The object of Go is to take control of terri-
tory. At the end of the game, the side which controls more territory wins the game.” Many beginners do not truly understand this concept. Raised on a steady diet
of Checkers and Sorry, they believe that the object of all board games is to capture
their opponent’s pieces. However, in Go, captur-
ing stones that do not need to be captured is like-
ly to be a wasted move, lead to bad positions and
even cause dead groups.
Diagram 3 shows a typical case. Ifile upload
n the last three moves, white has completely sur-
rounded his opponent’s stones. At move 4, black
has played a tenuki elsewhere, realizing his
stones are dead. Many a beginner would follow
up with A and kill black’s stones.
However, since black cannot hope to save
his surrounded stones, this move is worthless in
the end. Many a beginner is excited to see his op-
ponent’s stones removed from the board. But if he
wants to win the game, he is best served by play-ing elsewhere and surrounding territory rather
than killing stones that his opponent cannot save.
To put this in harmony with Cho Chikun’s quote, white is already in control
here. He does not need to prove the point. If he does, he is not only wasting a valu-
able move, but he is also stealing a point from himself by moving inside of his own
territory.
The beginner must not focus on capturing stones, as his instinct will almost
invariably lead him to do, but realize that it is territory, not stones, that win Go. His
restraint will lead to greater control of the board. Such restraint can lead to the dif-
ference between a major win or a minor loss.
The Object is to Take Control
Be careful what you think while playing Go
The board is a mirror of the mind of the players as the moments pass. When a master studies the
record of a game he can tell at what point greed overtook the pupil, when he became tired, when
he fell into stupidity, and when the maid came by with tea.
by Lamar Bishop
4
8/20/2019 Enclave First Issue
5/38
Far north of where most of us call home, the Norwegians
are putting out record numbers of Go players for their Go
championship.
8/20/2019 Enclave First Issue
6/38
It was time again for
the yearly Norwegian
Go Championship. So I
packed my bags and got
ready to go.
There were about 35
participants this year,which was a new record!
The event took place at
the Informatics build-
ing at the University
of Oslo, with gobans,
clocks and coffee for ev-
eryone, so all was set for
a successful event. At the last championship there were around 30 participants, which was
also a record at the time. Go in Norway is growing, which is great! Most of the
growth seems to be comong from the clubs in Bergen, Stavanger and Trondheim.
There hasn’t been that much activity surrounding the Go clubs in these cities be-
fore, so this is really good news. It’s exciting to see that clubs are starting to form
in every self-respecting city in Norway. Even though Go in has been around in
Oslo for about three decades, clubs in other cities are still relatively new.
There were volunteers scurrying around seing up the event, carrying go- bans, opening doors, bringing the coffeemaker and many other things that I had
no time to keep track of. New this year were the gobans; each had an exquisitely
hand drawn number, which made it easier to pair up at the start of each game.
The Championships started out at 10 am on Saturday. Due to a long trip,
the players from Stavanger and Bergen were slightly delayed, so we started out a
bit behind schedule, but everybody was in good sprits, still.
A laptop was used to keep track of the score and pair players. A�er each
game was played, the players entered a win or a loss on the laptop. When all thegames were played, the so�ware calculated which players should play next.
The first day had one round, then lunch, another round, dinner and then
the final round at the end of the day. At lunchtime most people
just relaxed and showed funny Go-challenges to each other. At dinner time, most
of us went two subway stations away to eat at some restaurant. The tradition is
to eat Indian food, but this year the restaurant we had in mind was full, so some
of us ended up at a Chinese restaurant instead. Since a lot of us were involved in
Every move must be carefully considered when you are among the best players in your
country
6
8/20/2019 Enclave First Issue
7/38
mathematics, and computers or who liked XKCD, many a conversation circled
around those topics that night.
The second day we started a few minutes past 10. It’s hard to make 35
people meet up at the exact same time, but the atmosphere was relaxed, so there
weren’t any nervous breakdowns. It helped that people behaved very responsi-
bly and helpfully throughout the championship.
At the end the topplayers took turns play-
ing for the grand prize,
which was being sent to
championships in Japan
and Korea. For us mere
mortals there were also
chinese Go books that
we had a chance to win, but this was considered
a bonus, since it’s not a
hard rule that they are
there every year.
The winner of the
Norwegian Champi-
onship this year was
Morten Ofstad. Congratulations, Morten! For anyone interested, the resultsof the championship can be found at: hp://norway.european-go.org/norway/
nm2008.txt
Thanks to all the participants and everyone involved, I hope to see you all
again next year!
Story by Alexander Rødseth
Alexander is a 27 year old Norwegian Go player, currently at-
tending school at the Norwegian University of Science & Tech-
nology.
He lives with his family and plays Go on KGS while discuss-
ing Linux with anybody interested.
Opening moves at the Norwegian Go Championship
Photos by Andreas Oppebøen
7
8/20/2019 Enclave First Issue
8/38
Life and death is a vital element of improving one’s game. Every problem requires
the player to kill or save a group of stones. In each problem there is a vital point.
The vital point is a brilliant play that will either save or kill the group. Take a look at
these problems and see if you can solve them!
Beginner Life and Death
White to kill
White to kill White to kill
White to kill
8
White to kill
White to get the best possible result
8/20/2019 Enclave First Issue
9/38
8/20/2019 Enclave First Issue
10/38
Future History
From November 16, 2007, when the Enclave was founded, it has been striving to up
the ante and deliver improved service completely free of cost. We have certainly had
our ups and downs, with some very high ups and very low downs. But overall, the
Enclave has put forth its contribution to Go players worldwide, with a modest de-
gree of success.
And to keep that contribution going, we would like to propose a few new projects,
and gauge the overall interest. Please review the proposed programs if you have time
and let us know if you’d like to join or help make them a reality! The Enclave is de-
pendent upon volunteers just like you who are willing to set aside some time to pro-
mote Go worldwide.
Mentor Program:
The mentor program was an ongoing Enclave project in late 2007 through early 2008.
It enjoyed a modest degree of success, and helped quite a few beginners learn to play
Go beer.
This id the first program we would like to rebuild, as it is the most effective way to
reach a variety of players. Any former members would be welcome to rejoin, keeping
their previous partners if they like, or geing different ones if necessary.
The long-term goal of the mentor program will be to establish a constant cycle of
learning, as previous students may mentor new players in exchange for what they
themselves were taught.
Those with comments or wishing to join as mentors or students, may write to:
Newsle�er:
This is the first edition of The Enclave, the Enclave’s self-titled e-magazine. The
magazine has been a long-term goal of ours, only now being realized a�er months ofaempting to start it.
The magazine will provide go problems of all kinds, interviews of professionals and
strong amateurs, reviews of strong games, both current and old, articles on go for
beginners as well as amateurs, and more.
The magazine will be published on an ‘as-ready’ basis, with the goal being a bi-
monthly production schedule. There will never be a cost for this publication.
Anyone with the will and the patience is welcome to write an applicable article for
A review of projects and future goals by Eric Dunham
10
8/20/2019 Enclave First Issue
11/38
this publication; however, all articles submied are subject to editing and/or rejec-
tion. This is so we may continue the high standard which we strive to provide to the
Go world.
The long term goal of this project will be to provide a permanent, renewing source of
teaching for Go players to learn from.
Anyone interested in submiing go problems, articles, interviews, quotations or any-thing else may write to: newsle�[email protected]
Lesson Records:
Lesson records have been a long-term goal of the Enclave since it was founded, but
there has never been a way for them to be realized. At the moment, we are still trying
to find a way to make video records of our lessons that are both high quality and eas-
ily accessible to our members.
When we have the ability, this program will be one of the first we implement.
The goal of lesson records is to provide an easily accessible way to access all past
information the Enclave has provided, so that new members may learn immediately
from old lessons and current members may remember material more easily.
Anybody who can provide video, and is willing to record lessons is welcome to con-
tact us at: [email protected]
Advertisement:
The Enclave is a not-for-profit organization; all money donated to the Enclave be-comes property of its members, no individual may use it for anything other than
Enclave projects. In the past, the Enclave has made ends meet through generous do-
nations from many people worldwide, but this has become harder as those generous
individuals have fallen upon hard times.
Therefore, the Enclave proposes using advertisment as a way to help make ends meet
once again. Any money made through advertisement will become property of the
Enclave’s members, it will not be pocketed by anyone at any point.
The long term goal of advertising is to make the Enclave self-sustainable and to help
it improve general quality.
Anyone interested in advertising may contact: [email protected]
If anyone has any suggestions for new projects, or if anyone is willing to volunteer to
help with current projects, such as teaching, please contact: administration@kgsencl
ave.com
11
8/20/2019 Enclave First Issue
12/38
From Heights Abv
12
Alexander Dinerchtein, 3p
Alexander Dinerchtein is a Russian born professional Go player. He is also one of themost-sought a�er Go professionals in the English speaking world. A�er extensive studying
in Korea, he aained professional status, and a�erward returned to Russia to spread the love
of Go throughout his homeland.
In recent years, he has founded Go4Go, a website with reviews of professional games.
He has also founded Goama, a free newsleer that includes Go news, interviews with pro-
fessional players, and game reviews. He also represented Russia in the World Mind Sports
Games 2008.
We managed to catch him right a�er the WMSG and ask him a few questions.
8/20/2019 Enclave First Issue
13/38
The Enclave: First of all, let’s start at the beginning. It is widely known that at a
young age you competed in both Chess and Go, but as you matured, you leaned
more and more towards Go. Why did you make the switch?
Alexander: My father, who introduced me to both games, is a strong chess player,
but he is only 10-kyu in Go. It was easier to beat him in Go, so I decided to pay more
aention to Go. Soon I was able to beat my father and was very happy about it.
E: A�er you began taking more interest in Go, you studied in your hometown of
Kazan, correct? On your website you state that at the time, Kazan was the center for
Russian Go, and you were surrounded by many stronger players. Has this changed
to any large degree?
A: Kazan is still the strongest Go playing city. First of all, Go is doing well here be-
cause of our local Sports Commiee. They budget good money for the best players
and teachers our city can afford.
For example, our Sports Commiee sent seven players from Kazan to the World
Mind Sports Games (WMSG) in Beijing, their support totaling about 10,000 Euros.
We have 2 professionals (me and Svetlana) and 2 players of almost-pro-strength (Ilya
Shikshin and Andrey Kulkov). I am sure that we can beat any European city easily
on 5-10 boards.
E: By the way, how was the WMSG in Beijing?
A: Terrible event. They didn’t even allow us to watch the games played by other
players! For the first time in my Go life! Game records were not available, you could
not watch games yourself, you couldn’t find any tournament tables in normal for-
mat… It was organized by real amateurs that had never played Go themselves.
E: I guess we’ll skip questions about WMSG in the future. In 1996, you and Svetlana
Shikshina were invited to study in Korea under Chun Pungcho, 8-dan pro. What
were your first impressions of Korean Go? Was it extremely different than European
Go? Or was it merely, as you say on your website, at a higher level than European
Go?
A: I was about 5-dan in 1996-1997. Here in Russia I was the youngest player in most
of the important tournaments – the Russian Championship, for example. When I
went to Korea I was surprised to find seven to eight year old Korean kids who were
not only not weaker but sometimes even stronger than me. Take Kim Chiseok, 4p for
example (He played for the Korean team in the last WMSG). He was 7 years old
13
8/20/2019 Enclave First Issue
14/38
while we were staying at one Go club. He had about a 50-50 score with me and Svet-
lana. We were probably stronger at strategy, but his reading ability was much beer
than ours. At the age of 7 he was able to solve Go problems beer than 80 percent of
the Korean pros!
Another impression: My teachers spoke really badly about Go books I had studied in
Russia. I was impressed by Kajiwara, 9p’s ideas, but they were laughing at him!
E: Did you study Go very differently in Korea? I have heard from many people that
Asian Go players are more insistent on form and style, whereas Western Go players
are more about ‘the idea’ of the game. Would you say this is true, or did you have a
different impression?
A: I know of only one big difference. In Korea they don’t use computer databases at
all. I don’t know why, but they keep all variations in their heads. The main reason
why Koreans are stronger is because they study Go much more seriously – some top
professionals never visited any schools or universities.
The average level of a Korean Go lover is six to eight kyu. There are plenty of such
players around the world. But the Koreans are more aggressive and they are usually
beer at reading than at strategy. That’s all. No secrets or special weapons here.
14
E: How long did you study in Korea?
A: I stayed in Korea for five years in four different clubs.
E: Who were some of your teachers? Who would you choose as your best teacher?
A: I worked with over 20 pro teachers and over ten amateur teachers. Who was the
best teacher? It’s hard for me to decide just one person. All of them were very kind to
me.
E: Having participated in Go events in both Europe and Asia, how large would you
“...Koreans are more aggressive andthey are usually be�er at reading...
That’s all.No secrets or special weapons here.”
8/20/2019 Enclave First Issue
15/38
say the skill gap between the two continents is?
A: If we played a 100 board match, the score between Asia and Europe would be 100-
0 or maybe 99-1, if we were lucky.
E: In 2002, you became the first honorary Russian Go professional. How did this
come about?
A: It was the special decision of the Hankuk Kiwon. Chun Pungcho, 8-dan, suggest-
ed the idea of giving us 1-dan and most of the other professionals agreed with him.
We were lucky, actually.
15
“If we played a 100 board match,
the score between Asia and Europewould be
100-0.”E: Would you say you are an average Go professional in Korea, or are you beer or
worse than others?
A: In my last Go club for Korean inseis I was ranked around twel�h or thriteenth
place, but suddenly I became a 1-dan. I am prey sure that nowadays there are sev-
eral hundred Korean kids who are stronger than me, but only few of them will ever
have a chance to become pro. There are 240 professionals in Korea. I am in the bot-
tom 40, I believe, but I’m not the last one. I played maybe 15-20 official pro games
before I returned to Russia and I won about 30% of my games.
E: In 2008, you and Svetlana Shikshina were both promoted to 3p by the Korean
Baduk Association. First of all, congratulations! However, some speculated that this
was a move to aract aention to European Go, rather than to reflect upon your skill
level. What would you say to this? Would you tend to agree, or do you believe your
rank was earned?
A: I feel that my Go has improved since I came back. I also feel that if I stayed in Ko-
rea, I could reach 3-dan myself without any help, but what is the difference? Korean
1-dan players are not weaker than most 9-dans. Dan level is not important at all in
8/20/2019 Enclave First Issue
16/38
Korea. For them, rating is more important, but here in Europe high dan Go players.
get more respect, so I am glad to become 3d.
E: In recent years, you have made a lot of steps to help advance Go in Europe and
the US, such as founding Go4go.net, thus giving average players access to cheap
reviews of professional games, and founding Goama, the e-magazine giving profes-
sional interviews and comments of exciting games. What inspired you to do thesethings, and who has helped you along the way? Has it been a good investment of
your time, overall?
A: I learned Korean, so now I have to use it if I don’t want to forget the language.
Most of the Goama articles are translated from Korean Go sites. Now you know one
of the reasons I keep it up.
E: You also offer Go lessons, easily available over the internet. What has been the
most satisfying thing about teaching Go?
A: I think that today I have the largest internet Go school in Europe. I can make a
living from giving Go lessons and I enjoy them a lot. I learn some things from my
students, too, and they make it so that I have to study a lot myself, so I can explain
modern paerns to them.
E: Being a teacher, may I ask you what you would suggest a beginning student first
study when they learn to play Go?
A: If you want to become a strong player, you must start with reading. Lots of Life
and Death problems every day! It’s not necessary to study strategy at all during the
first few years of playing. If you start with reading, you will easily learn strategy
with playing experience. But if you start from strategy you will always have reading
problems in your games.
16
Alexander Dinerchtein
Alexander started playing Go and Chess as a child, but began
focusing his efforts on Go as he grew older.
He achieved honorary Korean 1p status in 2002, and in 2008
became the first Russian 3p. He currently live in Russia, where
he spends his time promoting Go.
8/20/2019 Enclave First Issue
17/38
Some scribes are better than others.
Here’s an example of a not particu-
larly accurate one:
apotheosis [-]: Battousai: just be-
cause you have a two space exten-sion, does not mean you are alive
apotheosis [-]: Battousai: lots of
weak players think it’s true
apotheosis [-]: Battousai: they do
this dumb stuff, then wonder why
they died
apotheosis [-]: Battousai: well, it’s
cause they were dumb
G o: A M a tter of Li f e a n d Dea th
Have a funny Go-related story that you would like to share? A favorite Go joke?
Another lame Life and Death pun that we could pretend we didn’t base this sec-
tion’s name off of? Maybe you think you can outdo our bad sense of humor?
If you do, send your joke to us, and if we like it, we’ll stick it in the next issue!
Send all jokes to: [email protected]
Please be tasteful, all jokes can and will be viewed by children.
17
Because Go players are the only ones who learn by dying
OrangeKyo [13k]: you know... the
winner of a ko bale only gets like1 more pt than the loser right?
We can all take comfort in this tidbit that Shygost
points out:
You get strong when you learn to
hate weakness.
apotheosis [-]: Battousai:
SE says w is winning by 4.5,so bow on knees, okay?
classic: You know you’ve been play-
ing go too much recently when you are
reading a book and you are trying to trap
the words with imaginary stones
We all need help winning once in a while. Maybe we could take a lesson in Sublimi-
nal Messaging:
Lynx: In the middle of a game, casually li� up a sheet of paper with “YOU ARE
LOSING” on it in bright pink...
Lynx: then pretend as if nothing is happening, and put it down a few minutes later..
Muley: no, I think you should say I am going to the your bar group would will you
die like a drink too?
Quietp [17k]: if he’s called TheOne,
why does he have TWO accounts?
8/20/2019 Enclave First Issue
18/38
In1992,
TD Gammon,
a computer program,
played at the World Cup
of Backgammon, and achieved an
even record with the best players in the
world. Since then, humans have never been able
to overcome backgammon programs. In 1997, Gary
Kasparov, then-world chess champion, was dethroned by
Deep Blue, an IBM supercomputer. That day, chess ceased to
be a stronghold of human intelligence. And on other fronts,
humans are losing ground to machines. Most board games
are already being won by computers. Humans can still win in
Scrabble, poker, bridge, and others. But computers are already
starting to win those games. The only game where computers re-
ally fall short is Go.
That’s because Go is an incredibly complex game, hundreds andhundreds of times more complex than chess. In chess, there are 1050
possible moves, in Go 2.08x10170 possible moves on a 19x19 goban. On
a chess board, each turn presents just a dozen different viable moves,
but in Go, this can jump up easily to a few hundred different moves.
Where chess programs can read some 500,000 moves in a second, Go
programs can run only 50 moves in a second. To put it in perspective, a
supercomputer as powerful as Deep Blue could calculate some 200 million
chess moves in a second, but it would take a year and a half to do the exact
same thing in a game of Go.
MAN vs. MACHINEby Lamar Bishop
8/20/2019 Enclave First Issue
19/38
1992—TD Gammon achieves an even
score against top backgammon players at
the World Cup of Backgammon
Therefore Go presents a much greater challenge to computers and their pro-
grammers. To even the playing field, computers and humans face each other on
much smaller 9x9 boards. The smaller boardsize cuts down on the number of pos-
sible moves, greatly decreasing the workload for the computer. On a 9x9 goban, Go
programs have already beaten players as strong as 5p in official matches.
Around the world, supercomputers are starting to win against strong amateur
players. Soon, even Go will no longer be a game dominated by humans. As process-ing power increases, and ever more intelligent algorithms are brought to the fore,
computers will begin to win even in this last board game.
In Tainan, Taiwan, at the National University of Tainan (NUTN), the bale was
brought to the front in September during the World 9x9 Computer Go Champion-
ships. On September 27, MoGo, a Go program, running on the Dutch National Super-
computer, known as Huygens, faced off against Zhou Junxun, 9p. The event began
on the 25th, with MoGo playing against two amateurs, Professor Dong [5d], and Mr.
Luoh [6d]. MoGo beat Dong in every game, but lost every game to Mr. Luoh.
That same day the forum heard from many speakers, including Dr. Olivier Tey-
taud, one of MoGo’s developers. This was followed on the 27th by the main arac-
tion—MoGo facing off against Junxun in three rounds.
In Conference Room B309, in Chengzheng Hall, the movers and shakers of the
World 9x9 Computer Go Championships are geing ready for the highlight. The first
two games will be MoGo versus Junxun on a 9x9 board. As the time draws near for
the first game, the participants sit and prepare for the game. In front of the computer
is Dr. Olivier Teytaud, one of MoGo’s designers, who will read the moves off of thecomputer and place them on a physical board. Siing beside him is Professor Tsai,
who will be the witness for the game. Across from them sits Zhou Junxun, his face
drawn into a mask of concentration as he prepares for the game. Across the room,
photographers watch and snap pictures as the game begins.
1997—IBM’s Deep Blue beats chess world
champion Garry Kasparov in a traditional
match
August, 2008—Dutch supercomputer Huy-
gens beats Kim Myungwan, 8p on a 19x19
board with a nine stone handicap
September, 2008—Huygens plays aga
Zhou Junxung in a three round match
the World 9x9 Go Championships
19
Photo by Brian Allen
8/20/2019 Enclave First Issue
20/38
Zhou Junxun makes a move during the second round
Junxun has a lot to think about. MoGo
has been playing decently against strong
amateur players, not well enough to threat-
en Junxun, but in those games MoGo has
been using no more than half the power
Huygens will bring to bear in this match. In
these rounds, MoGo will be using the fullpower of 800 processor cores. Just a single
one of these processors is powerful enough
to beat the average Go player.
The tension in the room is a palpable
thing for the players. MoGo will be moving
first in this game. MoGo makes its first move at tengen. It is well known that tengen
is the best possible move on a 9x9 board, so this is not a surprise. Junxun places a
stone, Teytaud copies it on the computer screen, and the game begins.
As the game goes back and forth between the liquid screen and the wooden goban,
it becomes apparent that MoGo is winning. At move 11, Junxun stops and examines
the game closely. His turn stretches on for a long time; over five minutes have passed
since his last move.
Watchers over the internet begin to speculate that Junxun has already lost this
match. Nobody expects Junxun to be able to win against such a powerful computer
on such a small board.
However, Junxun continues the game, in spite of what people were expecting.
At first his move looks desperate, almost unecessary, like the last-ditch effort of alosing team. But then, as it plays out, suddenly it becomes clear that Junxun knew
exactly what he was doing. Within another few moves, onlookers have changed their
mind about the game. Everyone agrees the game looks fairly even.
Suddenly, a�er just eleven moves, the game doesn’t just look even, it looks like
Junxun has won. By move 36, even MoGo agrees, and resigns. A�er the match, Junx-
un comments, “In the first 9x9 game with MoGo, I was shocked to find a�er eleven moves
that I had already lost the game. I was really very shocked. It took me five minutes to calm
down and start designing a hamete that the computer could not see and could not reverse.”
On the second game, Junxun takes black and makes the first move. In this sec-
ond game, he is much more confident and almost immediately he begins to take the
lead. Within 25 moves, the game is decided. MoGo takes another 15 moves and then
resigns. Junxun says, “I took black in the second game, and it was a relatively easy game.”
A�er a short break, the matches continued with the third game, a full-sized
19x19 match. Because of the enormously larger workload it will be facing, Mogo
takes a seven stone handicap. But unfortunately for MoGo and its developers, seven
stones is not enough. Though Mogo has been able to beat many strong players with
20
Photo provided by the National University of Tainan, Taiwan (hp:// go.nutn.edu.tw/)
8/20/2019 Enclave First Issue
21/38
eight and nine stones, it seems that is not yet ready to move down to seven stones.
Almost immediately MoGo makes a costly mistake and loses a corner. Mogo
cannot immediately see that it has lost the corner and ends up wasting precious
moves trying to save it. As it struggles to make the dead group live again, Junxun
strengthens his hold on the corner.
The onlookers are not too surprised. Some comment that MoGo could have
kept the corner, but nobody says much else. On a 19x19 board, this is expected. A�er the first mistake, Junxun takes each corner of the goban, one by one,
stripping MoGo of its lead. Less than halfway into the game, onlookers are proclaim-
ing the game over. And as MoGo makes mistake a�er mistake, they are proved right.
Junxun later says that he figured he had won the game by move 30. Their is
an almost nonchalant air about the way that the internet watchers dismiss MoGo as
unimportant. There has not yet been a computer program that can challenge a pro-
fessional Go player on a 19x19 board, and this one is the same as all the others.
What many of them are forgeing is that just a few years ago, Go programs
could barely compete with the average club Go player, much less a professional.
But, with the advent of Monte Carlo algorithms, that has changed completely. Where
computers used to barely manage an 8 kyu level of play, they now are able to aver-
age low amateur dan play.
As Dr. Jaap van den Herik of Maastricht University said, “This is remarkable,
since around 2000 it was generally believed that the game of Go was safe to any aack by a
computer program.”
Mr. Junxun may have won the match today, and he may win in the years to
come, but soon it will be much more difficult. Dr. Herik predicts, “...before 2020 a com- puter program will defeat the best human Go
player on a 19x19 Go board in a regular match
under normal tournament conditions.”
For now, Go players laugh at the pros-
pect. But the next few years may tell a very
different story.
MoGo TiTan vs. Zhou Junxung
Seven handicap. Result: Junxung by resign
MoGo TiTan vs. Zhou Junxung
Result: Junxung by resign
MoGo TiTan vs. Zhou Junxun
Result: Junxung by resign
21
Many thanks to Dr. Chang-Shing of the National Uni-versity of Tainan for his help in writing this article!
8/20/2019 Enclave First Issue
22/38
Insights
Koram [7d KGS] vs. Aguilar [7d KGS] contributed by Joshua Allen
Game Kifu
This is a review of a game played
on October 16, 2008 between Koram and
Aguilar, both KGS 7 dans. White won the
game by resignation.
Our reviewer for this issue is Joshua
Allen, Baousai on KGS, 5d AGA.
B3 @ q4 B37 @ q11 B151 @ p10 B197 @ r3
B203 @ q11 B209 @ q11 B229 @ r3 B235 @
r3 B241 @ r3 B245 @ g11 B247 @ r3 W76 @
p11 W186 @ q3 W200 @ p11 W206 @ p11
W224 @ q3 W232 @ q3 W238 @ q3 W244 @
q3
Interesting approach. Some of us might take a
look at this low approach, glance across the board
at the 4-4 stones, and say to ourselves ‘this feels
wrong’. If you were aempting to play this gamewith a more influence oriented strategy in mind,
that would be correct.
We can see in Diagram 3 that if white moves at A, we see a common varia-
tion. I recommend, if you don’t already know this variation, to memorize
it, especially since its so short. It gains black a base without being denied
access to the center. It’s one way to sele nicely if you’re unsure of what to
do.
White now has a few choices. Some com-
mon ones would be A and B, as shown in
Diagram 2. Let’s take a closer look at the
results of these moves.
Diagram 1
Diagram 2
Diagram 3
Game Notes:
22
8/20/2019 Enclave First Issue
23/38
The move at B is also possible, however, as you’re
glancing at the sequence to come, ask yourself,
why would it be inappropriate here?
If you answered, ‘Because white lacks a stone at
K16 in order to effectively use the newly gained
wall he has created, you would be quite correct!
If we look at Diagram 4, we can see an example ofthe result.
Diagram 4
In the game, the players choose standard joseki.
Both white and black get some territory, a tiny bit
of influence, while keeping some lingering aji (the
cut at f14 and the aachment at f17) for later.
Diagram 5
A�er finishing up the top for now, black chooses
an aggressive approach towards the boom, as
shown in Diagram 6.
If he had chosen a more
peaceful variation at B, as
shown in Diagram 7, both
players would be able to
develop frameworks/moyo.
However, this is a variation
that black has decided to reject.
Diagram 6
Diagram 7
23
8/20/2019 Enclave First Issue
24/38
Modern (at least to me) joseki.
To some, the aachment at e8 looks strange. Why
go through all that trouble if you’re just going to
turn around and dive into the 3-3? Surely your
moves are wasted at that point?
If that is your reasoning, I must say its sound.
However, you might notice that white is now
flat on the boom. Sure, the 2 marked stones are
deep in enemy territory (and I use the term ter-
ritory here loosely), but they aren’t off the board
yet! Now the question is raised, how does one
deal with all that aji lingering from them?
White answers most simply by burying them evendeeper, Diagram 9.
Diagram 8
Diagram 9
This is a very telling move. White could have
dove into the corner and gained himself some ter-
ritory.
Looking at the board, we see that white is com-pletely without a corner at the moment. So why
doesnt he?
Diagram 10
24
8/20/2019 Enclave First Issue
25/38
A variation like Diagram 11 is why.
If white had chosen this, for example, black gets
influence and doesnt really have many other con-
siderations besides reducing the le�
Diagram 11
Now a loose plan is formed.
Black has two weak groups that white can aack,
as shown in Diagram 12.
Diagram 12
Seeing the double threat, black chooses to aid the
original group, abandoning the second.
Diagram 13
25
8/20/2019 Enclave First Issue
26/38
8/20/2019 Enclave First Issue
27/38
Here we see the value of the ko. But, did white
lose out here? Lets examine more closely.
White had influence towards the center of the
board and still does.
Black had weak groups toward the center whichcould be taken advantage of and those groups are
still there.
Black gains points but loses development along
the top.Diagram 17
Some might wonder, why not respond to white’s
aachment? The reason is simple; he would have
2 weak groups.
As it stands, he still needs to be wary of k17 being
cut off, because of the aachment that lingers at
f17.
Diagram 18
This is the best result that white could hope for.
Once again he has 2 weak groups to assault.
Diagram 19
27
8/20/2019 Enclave First Issue
28/38
Vital.
O�entimes while we are playing our game we like
to play instinctively. For some of us that would
mean responding to p12 immediately.
To do so, however, would mean that black wouldget to connect his two weak groups and not have
to worry about living, just reducing and taking
profit.
I cannot stress enough just how many points
would be thrown away by allowing that.Diagram 20
Again, do we respond?
This is the point in our fighting that we must force
ourselves to stop and constantly assess what is go-
ing on in the game.
Here, it seems that one of black’s groups is on its
way to being alive. White’s group on the right is in
no immediate danger.
Diagram 21
My feeling here is that black cannot live without
a miracle and thus the game is already over.
Diagram 22
28
8/20/2019 Enclave First Issue
29/38
Black now agrees with me and changes ideas.
Instead of uselessly trying to live, perhaps he can
take enough on the boom to compensate.
Diagram 23
This is a nice defense.
Diagram 24
This is a shape point.
Remember: don’t get greedy when you are trying
to live!
Diagram 25
29
Diagram 26 is an example of greed
when trying to take the corner.
Diagram 26
8/20/2019 Enclave First Issue
30/38
If we follow that greedy path the whole way
through, we get to here.
Now what?
This time, white has two different groups that
need to live.
Diagram 27
Diagram 28
This is an odd move. White doesnt have to worry
about life anymore.
Diagram 29
I think perhaps this is the variation black wanted.
This would have been great for him. However, as
we’ll see, that’s not what happened.
30
8/20/2019 Enclave First Issue
31/38
Diagram 30
Looking at white’s move in Diagram 30, we notice
immediately that this has to die for black to hope
to win.
Diagram 31
So, as we see in Diagram 31, black totally ignores
it.
Very interesting.
A bit further down the road, and we are back to a
ko.
Diagram 32
Diagram 33
Black could take this, but then white would live
and he knows that he cannot allow that.
A few moves
later, and game
over.
31
Screenshots used by permission of William Shubert
Special thanks toWilliam Shubert,creator of KGS, for permission to usescreenshots.
8/20/2019 Enclave First Issue
32/38
A Hstry G:
Go is an ancient game. In fact, it’s the oldest game known to man thatis still played as it was in ancient times. It’s been known by many differentnames, been played by royalty in various countries, and decided the result ofreal-life bales.
Along with all of that, Go has aquired quite a few variations in rules andstructure. In this issue we’ll be focusing on one of these variants: Sunjang. Sunjang Baduk (Hangul: 순장 바둑) is an old Korean variant of Go, orBaduk (바둑) as it is known in Korea. Sunjang enjoyed a Golden Age in Ko-rea from the 16th to mid-20th centuries, duringwhich it was almost the exclusive variant of Goplayed in Korea. There is evidence that Koreans
also played ‘open board’ games during this timeas well, but Sunjang was a much more popularversion. During the early 20th century, modernGo spread to Korea from Japan. Korean play-ers accepted the new rules as an improvementover old rules, and for a while played both ver-sions alongside each other, but as time went on,Sunjang became less and less popular as players
began to think of the Japanese rules as the prod-uct of a ‘superior ’ modern culture. In 1945, Cho Namchul (조남철), the ‘Pio-neer and Father’ of Korean Go, returned to Korea from studying Go in Japan,and immediately went about founding the first private Go school in Korea, theHansung Kiwon (한성 기원). In his new school, Cho taught the Japanese ver-sion of the game, which cemented Japanese rules as the accepted version inKorea. The main difference between Sunjang and modern Go is in the very
Cho Namchul, ‘Pioneer and Father’ of Korean
Go, helped solidify the end of Sunjang Baduk
32
8/20/2019 Enclave First Issue
33/38
beginning of the game, before any players have made a move. Both playersset eight stones on prescribed points of the
board. This is called poseok (布石), a word nowequivalent to fuseki. A�er the stones have beenplaced for poseok, the move virtually pre-scribed for black is cheonweon (천원)—tengen.
Though the black player is not required to playat tengen, it is widely considered to be the bestmove he can play. Diagram 1 shows the poseokand first move of Sunjang. Because of the placement of the stones,once the players have finished poseok, the
game leads immediately intoa fierce jungban (중반)—chu-
ban, or middlegame. The vio-lent middlegame is impossibleto avoid, every single movemade builds tension with thepre-placed stones.
In Diagram 2 we can seean example of this from the‘Last Game’ between Ch’aeKeuk-mun (채 극문) and NoSach’o (노 사초). Within just afew moves, both players musttake the defensive, while si-multaneously keeping up anoffensive so that their oppo-nent does not gain an advantage. Note that none of the groups appear to be alive right away, and that majorthreats are ignored for bigger ones. In just a few moves, this game has turnedinto the passive player’s nightmare. As the game progresses, it gets even more ferocious and difficult. Weakgroups are abandoned in favor of large points.
While neither player is willing to lose a large group, they are also unwill-ing to give up even larger areas to defend them.
Diagram 1: Poseok and black’s first move atcheonweon
Diagram 2: Sunjang openings cannot help but be violent
33
8/20/2019 Enclave First Issue
34/38
The scoring method
is a cross between modern Japanese scoring, where nostone has any point value,and modern Chinese scor-ing, where all stones areworth a point. Thoughstones are not worth points,placing stones in your ter-ritory will not take awaypoints.
Sunjang is a great op-tion for players who wouldlike to strengthen their chu-
ban, or an exciting gamewithout the option of beingcalm or, let’s face it, boring.
Such playing is typical of Sunjang games, it is in the nature of the game.From the moment that the poseok has been laid, the board is set for a fight.
The second way that Sunjang differs from modern Go is in scoring. Dia-gram 3 shows the end position of the ‘Last Game.’ In modern Go, dead stonesare removed and territory is then counted up. If the ‘Last Game’ were scored
like this, we would see a position like Diagram 4. But in Sunjang, all unecessary stones are removed from the territory be-fore scoring. If the stone doesn’t hold territory in, it is removed before scoring.An example of this can be found in Diagram 5.
B45 at n1
B47 at q2
B157 at s15
B173 at q2
B177 at l3
B185 at q2
B191 at q2
B197 at q2
B203 at q2
B225 at s10
B231 at s10
B233 at t4
B241 at s10W60 at p2
W182 at p2
W188 at p2
W194 at p2
W200 at p2
W228 at r10
W238 at r10
W244 at r10
34
Diagram 3 Diagram 4 Diagram 5
Ch’ae Keuk-mun (채 극문) and No Sach’o (노 사초)
Result: No Sach’o (노 사초) +.5
8/20/2019 Enclave First Issue
35/38
A Picture Worth a Thousand Moves
A Graphic Celebration of Go
This issue’s photo was submied by Vance Dunham.
Do you have a photo that captures your love of Go? A drawing that expresses the depth of thegame? Artwork that you think should be shared? If so, send it to us at: [email protected]!
The best piece of art recieved by the next issue will be featured here!A few quick requirements:Please make certain the picture is high-resolution. It’s not that we don’t like small pictures, they justaren’t magazine friendly.All pictures need to be family friendly, nothing you wouldn’t want your children or grandmotherto see.And finally, make sure that it’s a picture that the whole world needs to see!
Those interested in impressing others with their intelligence play chess. Those who would sele for
being chic play backgammon. Those who wish to become individuals of quality take up Go.
What your game says about you
35
8/20/2019 Enclave First Issue
36/38
Te Marketplace
Lessons
Offers online teaching on both KGSand IGS for a small, negotiable fee.Baousai has been teaching for five
years, and is the Enclave’s premierteacher.Highly recommended by Enclavestaff and students!‘Baousai is our oldest teacher, and Ihave never been displeased with him.His sharp tongue and fun sense of hu-mor lend a memorability to his lessons
not paralleled by any other teacher Iknow. He is also very affordable andextremely reasonable in dealing withstudents.’ —Eric Dunham, Ownerand Founder, the KGS Enclave
Contact:[email protected]
for pricing, questions, comments,or to set up a lesson.
Offers online teaching on KGS fora small fee.GeorgeW has been teaching for
two years, and highly recom-mended!He is also fluent in Chinese, foranyone wishing to study with aChinese speaker‘GeorgeW has been volunteering les-sons for the Enclave for the past fewmonths and has done an excellent job
every time. I guarantee, personallythat he is worth his weight in gold,not to mention the paltry price heasks.’ —Eric Dunham, Owner andFounder, the KGS Enclave
Contact: [email protected]
with any questions or comments,or to set up a lesson.
GeorgeW, Chinese 2dBattousai, AGA 5d
Would you like to advertise here? The Enclave will post a free classified ad for anyone, including businesses. We
can advertise lessons, Go equipment, wanted ads, or anything else that reasonably fits the spot!
If you’re interested in advertising anything Go-related, please send an e-mail to:
If you have a problem with, or a question about any of the ads here, please contact:
36
8/20/2019 Enclave First Issue
37/38
2007 Volunteers and ContributorsIn our first year, we have had some wonderful volunteers, and some dedicated
contributors. It is thanks to the contributions of people like these that the Enclave
has been able to continue providing its services to the Go community, so we would
like to publicly thank them for their help.
Enclave Administration: $383.86
Nathan Wall $50.00
Magnus Jensen $50.00
Felix Ufer: $38.38
Shishinn Sun: $20.00
Chedo�: $20.00
Uberness: $19.87
Jonathan Nyquist: $10.00
David Boss Jr.: $5.00
Monetary Contributors
Volunteer TeachersGeorgeW [4d]
TCHATTE [2d]
Krayle [1d]
acid [1k]hikarujr [1k]
Iink ~[1k]
Meepy ~[1k]
backpack [2k]
xioshe [2k]
Dice [3k]
Special ThanksVance Dunham
Alexander Rødseth
Nqua Xiong
Hu of KGS
NoKno�
Zachary Dunham
Cheddot
Joshua Allen
37
8/20/2019 Enclave First Issue
38/38