Sixty-fourth KGS Computer Go Tournament

Sunday October 10th 2010

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


format20-round Swiss
board size19×19
time14 minutes plus 25/60s


The first round started at 08:00 UTC.

Result table


Fourteen players registered, and apart from AyaMC all played in every round. AyaMC was present shortly before the tournament started, but lost its connection before round 1, and reappeared in time to play in round 10.


In round 2 Zen19 played ManyFaces2 for the first time (of four) SGF. ManyFaces2 won.

WeakBot50k vs pachi
Position after move 426 (pass by W).

In round 4 the game between WeakBot50k and pachi SGF ended in the position shown to the right. Pachi, as black, has clearly won; in fact it is 243½ points ahead. However, at this point pachi resigned.

PueGo vs Zen19
Position after move 245.

In round 5 PueGo and Zen19 played an interesting game SGF. After move 200 a semeai developed involving three kos (the diagram to the left shows the position after move 245). The subsequent play was confusing, but I felt that Zen19 played the triple ko more competently than PueGo did. Anyway, Zen19 eventually won.
      Also in round 5, something happened to CzechBot, and it did not join its game with the absent AyaMC. As CzechBot was black, it lost the game on time, and AyaMC scored its first win.

In round 6 CzechBot was not in the "Computer Go" room where the tournament was being held, but still joined (and won) its game with IdiotBot. I had not realised that this was possible, I thought that players had to be in a room to play their games there.

In round 9, as in round 4, pachi resigned in a won position with all groups settled and all dame filled, this time against SimpleBot.

In round 10 AyaMC reapeared in time to play, and beat, IdiotBot. ManyFaces2 played, and won, its second game against Zen19 SGF

After ten rounds, ManyFaces2 had won all its games. Zen19 had won all except those against ManyFaces2.

pachi vs ManyFaces2
Moves 70‒75.

In round 13 pachi chased ManyFaces2's stones in a ladder which didn't work SGF, see diagram to the right. With move 76 it tenukied, but it played out the rest of the ladder later.

After 14 rounds, ManyFaces2 had won all its games, Zen19 had won all except those against ManyFaces2, and no other program had more than eight wins.

Zen19 vs ManyFaces2
Move 41.

In round 15 Zen19 and ManyFaces2 played their third game of the tournament SGF. ManyFaces2 played move 41 as shown to the left. This surprised me ‒ a human 10-kyu can see that White's obvious response makes this move poor for Black. Zen19 won the game.

In round 16 CzechBot disconnected after move 7, and lost on time to WeakBot50k. WeakBot50k did remarkably well in this tournament, at one point it was in third place out of 14 on the results table. This was not achieved by good play, but by reliably making legal moves within the time limit, a dull but valuable skill. Break19 also disappeared and lost on time, to IdiotBot, in this round.

In round 17 CzechBot joined its game with break19, but did not make a move, and timed out.

In round 19 ManyFaces2 and Zen19 played their fourth game of the tournament SGF. ManyFaces2 won. PueGo obtained a won game against MCark, but lost it on time while MCark was making meaningless moves.

pachi vs ManyFaces2
After move 39.

In round 20 ManyFaces2 and pachi played a game that surprised me. SGF. After move 99, the position was as shown to the left: ManyFaces2 has two safe groups in the centre of the board, and a majority claim to all four corners, while pachi has strings of stones with several cutting points and no territory. However pachi went on to win, giving ManyFaces its second loss of the tournament.


Details of processor numbers, power, etc.

Aya, running on 8 cores of a Xeon 5355 2.6GHz
break, probably running on a single processor Intel(R) Celeron(R), 1.7Ghz
MoGo, running on double-core AMD Athlon(tm) 64 X2 Dual Core Processor 4800+ (2.5GHz).
running on Linux, 2GiB RAM, Intel(R) Celeron(R) M CPU 530 @ 1.73GHz, system shared with WeakBot50k
Many Faces of Go, running on a 12-core Xeon.
MC_ark, running on a Core2Quad 2.4GHz
Orego, running on one of the five nodes of a custom Linux cluster build by PSSC Labs:
pachi, running on a 40-thread cluster
Fuego, running on two threads on a Core2Duo E7200
GNU Go, unspecified platform
running on one processor of a 4GiB RAM, AMD Athlon(tm) 64 X2 Dual Core Processor 4000+, shared with WeakBot50k
valkyria, running on a i7-860 4 Core processor at 2.80 Ghz.
running on Linux, 2GiB RAM, Intel(R) Celeron(R) M CPU 530 @ 1.73GHz, system shared with IdiotBot
Zen, running on a Mac Pro 8 core, Xeon 2.26GHz