Seventy-sixth KGS Computer Go Tournament

Sunday October 2nd, 2011

These results also appear on an official KGS page.


format15-round Swiss
board size19×19
time14 minutes plus 10/30s


The first round started at 08:00 UTC.

Result table

Zen19 pachi PueGo stv AyaMC Orego PNUGo oakfo
W13R B17R W111R W12R B110R W114R W16R B112R W115R B18R W19R B14R W113R W15R B11F 15114114Winner
2pachi2 B03R W07R B011R
W16R B19R W115R B12R W113R B11R W110R W15R B112R B145.5 W114R W18R 1211974
3PueGo B02R W010R B014R B06R W09R B015R
B13R W14R B07R B11R W111R B181.5 W112½ B15R W113R 812334
4stv B06R W012R B015R W02R B013R W03R
B05R W111R B114R B17R W18R B1121.5 W192.5 B14R W110R 811934
5AyaMC W08R B09R W01R B010R B04R W17R W15R B011R W014R
W13R B115R W12R B1139.5 B16R W112R 811234
6Orego12 W04R B013R B05R W012R W01R B011R W07R B08R B03R W015R
W16½ B1101.5 W12F B19R W114R 51108
7PNUGo B05R W045.5 B014R W081.5 B012½ W0121.5 B092.5 B02R W0139.5 B06½ W010
B13F W17R B111R W115R 4970
8oakfoam W01F B08R W05R B013R W04R B010R W06R B012R B02F W09R B014R W03F B07R W011R B015R

In the table above,
   0 is a loss
   1 is a win
   J is jigo
   left superscript is the player's colour
   right superscript is the round in which the game was played
   a subscript shows how the result was determined:
      R for resignation
      T for time
      F for forfeit
      a number for the points difference after counting.
All the 0s, 1s and Js are links to the game record.

Eight players initially entered. Then gomorra3 withdrew, as it has a bug which sometimes causes it to hang and refuse to make any more moves. I persuaded its programmer, Lars Schäfers, to re-enter it, as this would make the numbers even, which is better than having byes.

Shortly before play started, Petr Baudiš decided to enter PueGo (his build of Fuego, not the latest version), so I removed gomorra3 again.


In round 1, oakfoam was missing. I learned later that its programmer, Francois van Niekerk, was having connection problems. Oakfoam eventually appeared on KGS, and played from round 4 onwards.

PueGo vs Zen19
Moves 43-49

In round 2, PueGo appeared to do well against Zen19, in the sequence shown to the right – my impression was that this killed the white group. But there is a ko involved, and the white group eventually lived despite Black making the next two moves in this corner, so I am probably wrong.
    Also in round 2, stv chased pachi2's stones in a ladder, not noticing that pachi2 had a ladder-breaker. This was disastrous for stv, and it lost the game.

In round 6, PueGo chased pachi2's stones in a ladder, not noticing that pachi2 had a ladder-breaker. This was disastrous for PueGo, and it lost the game.

In round 8, pachi2 chased oakfoam's stones in a simple unbroken ladder, killing them when it reached the edge of the board. This was disastrous for oakfoam, and it lost the game.

In round 12, stv chased Zen19's stones in a ladder, not noticing that Zen19 had a ladder-breaker. This was disastrous for stv, and it lost the game.

Zen19 vs PueGo
Move 27

In round 14, PueGo played move 27 as shown to the left, in its game against Zen19. I found this surprising: pulling out its cutting stones, in sente, looks better.

Annual points

Players receive points for the 2011 Annual KGS Bot Championship as follows:


Details of processor numbers, power, etc.

Aya, running on 6 cores of an i980X 3.3GHz
Oakfoam, running on a quad-core Intel Xeon E5520 with hyperthreading (so 8 threads).
Orego, probably running on one of the five nodes of a custom Linux cluster built by PSSC Labs: the node has two AMD Six Core Dual Opteron 2427 2.2 GHz (12 cores total), 8 GB RAM, Centos Linux.
pachi, running on 64 platforms, each x86 64 bits, 32 GB ram, using 22 cores of each, giving about 1500 playouts/s/core at the beginning of a 19x19 game.
Fuego,version 0.4.1.SVN, running on an 8-core i7 920
GNU Go, unspecified platform
Steenvreter, running on an Intel Core2, Q6600 (4 cores)
Zen, running on a Mac Pro 8 core, Xeon 2.26GHz