Sixtieth KGS Computer Go Tournament

Sunday June 13th 2010

These results also appear on an official KGS page which links to the records of all the games.

Rules

format10-round Swiss
board size19×19
rulesChinese
komi
time14 minutes plus 25/30s

Times

The first round started at 08:00 UTC.

Result table


placenamewinsSOSSoDOS
1stZen1995244
2ndCzechBot84936
3rdstv75029
4thManyFaces174825
5thpachi65325
6thAyaMC55625
7thPNUGo54411
8thOrego44810
9thhcBot3498
10thSimpleBot3438
11thIdiotBot2588
12thWeakBot50k1503

Thirteen players registered. HcBot is new to these events: it is a build of Fuego, modified by Michal Dvoracek.

Normally, when an odd number of players is registered, I remove IdiotBot (with its operator's approval), so as to make the number of players even and avoid byes. But one of the players registered, break19, was not present shortly before play started, and has been unreliable in the past, so I left all thirteen in the draw, intending to remove IdiotBot if break19 appeared, otherwise to remove break19.

Results

In round 1, there was an exciting game, with a lot of fighting, between Manyfaces1 and Zen19 SGF. This may have been the best game of the tournament – as both players are stronger than me, I do not feel able to comment on it, but I recommend looking at it. Zen19 won. Jonathan Chetwynd (a British 3-dan player) writes

A very exciting game, I especially enjoyed and appreciated the significant number of tenuki played.

Both players appear weak on liberties and life and death, at least from this game file.

e.g. black 137, clearly there is a liberty race, and had this move been played at G2 black would lead 4/3.

Also in round 1, SimpleBot played very slowly against WeakBot50k. This was because it was on its 9×9 settings. It became short of time early in the game and began to play very badly, and lost. Its operator Aloril changed its settings before the next round.
       Break19 had not appeared by the end of round 1, so I removed it from the draw.

In its round 2 game with CzechBot SGF, ManyFaces1 played out a losing ladder. Normally it reads simple ladders correctly, I do not know what was wrong here. CzechBot won the game.

Also in round 2, AyaBot obtained a solidly won game against IdiotBot, and lost it on time while unnecessarily filling its own territory. This gave IdiotBot its second win of the tournament, and put it briefly in the lead (hcBot has a similar loss on time to IdiotBot in round 1).

In round 6, Zen19 lost its only game of the tournament, to CzechBot SGF.

AyaMC vs SimpleBot
Central part of board only.
Moves 530-545. Black passsed throughout.

In its round 9 game with SimpleBot SGF, Aya again obtained a totally won game and then lost it on time. The final 16 moves are shown to the right.
       AyaMC was choosing its moves fast, but it was unable to make the required 25 moves in 30 seconds because of lag between Japan and the U.S. It always ends its games by converting all its terrritory to one-point eyes, though there is no need for this. It lost several games in this tournament because of lag while it was filling its own territory.
       However, my reason for showing these moves from its game is my aesthetic feeling that if it wants to convert its territory to one-point eyes, it ought to do it more efficiently. In the position shown, it has just captured a 16-stone group. It could have converted these 16 points of territory to nine one-point eyes in seven moves; and if it had, and had then passed, it would have won this game. Instead, it used 8 moves so as to still leave an eight-point connected territory, and then ran out of time.


Details of processor numbers, power, etc.

AyaMC
Aya, running on i7 980X 3.3GHz 6cores
CzechBot
MoGo, running on double-core AMD Athlon(tm) 64 X2 Dual Core Processor 4800+ (2.5GHz).
hcBot
FueGo, running on Core i7 980X (6 cores at 4GHz), 12 threads
IdiotBot
running on Linux, 2GiB RAM, Intel(R) Celeron(R) M CPU 530 @ 1.73GHz
ManyFaces1
Many Faces of Go, running on four cores of a four Core2 Quad 2.3 GHz.
Orego
Orego, unspecified but probably running on 2 x 3 GHz Dual-Core Intel Xeon
pachi
pachi, running on 36 threads mixed i7/core2 on 7 machines
PNUGo
GNU Go, unspecified platform
SimpleBot
running on one processor of a 4GiB RAM, AMD Athlon(tm) 64 X2 Dual Core Processor 4000+, shared with WeakBot50k
stv
Steenvreter, running on an Intel Core2, Q6600 (4 cores)
WeakBot50k
running on Linux, 2GiB RAM, Intel(R) Celeron(R) M CPU 530 @ 1.73GHz
Zen19
running on a Mac Pro 8 core, Xeon 2.26GHz.