Tournament results:

Home page

The 13th tournament (April the 16th, 2012)

The results of the 13th tournament Gomocup 2012

The second tournament AI vs. Human (21st June 2011)

The second tournament AI vs. Human took place on 21st June 2011 in Prague in Shangri-la tee shop. The winners of Gomocup 2011 has been confronted against Humans, against the best Czech gomoku players. This time the result AI:Humans was 5:3. More in czech language is here.

12th tournament:(4th-7th April 2011)

The results of the 12th tournament Gomocup 2011

11th tournament:(24th April 2010)

The results of the 11th tournament Gomocup 2010

10th tournament:(4th April 2009)

The results of the 10th tournament Gomocup 2009

9th tournament:(19th April 2008)

The results of the 9th tournament Gomocup 2008

8th tournament:(27th April 2007)

The eighth tournament took place at The University of West Bohemia in PC lab UI 303 as the last time. There were 11 computers Intel Pentium 4 3.4GHz HT, 1GB RAM in the lab. This year was very rich in new programs, Tito and Gmotor (used to be G_motor) showed off with a new version. There were four brand new programs: Jude by Wooming (China), Pecuchet by Vilem Walter (Czech Republic), demo of Onix by István Virag - Janos Wágner (Hungary) and at the last moment created NaBaMoKu from Jiri Isa (Czech Republic).

There were programs written in interpreted languages C# and Java for the first time.

There were 23 programs from 5 countries in competition. According to last year's results, every following trio was divided into three groups by lots:

• 1.-3. (Goro, Tito, HGarden)
• 4.-6. (Pela, swine, Kanec)
• 7.-9. (BestMove, Noesis, QMentat)

The rest was draw from remaining programs. Finally there were three basic groups of 8,8,7 programs.

• 1. Tito, Kanec, Qmentat, Trunka, Gmotor, Benjamin, Pisq7, Puškvorec
• 2. HGarden, Pela, Noesis, Jude, Hewer, Onix, Leonardo, Curnpis
• 3. BestMove, Swine, Goro, Brutefor, AI1000, NaBaMoku, Pecuchet

Openings number 1, 3 and 5 were decided by lots.

There were several technical problems at the beginning. Proceeding basic groups were aborted without reason, so I restarted the tournament to ensure everything would be ok. Unfortunately, due to high networking security, online broadcast didn't work. After that, everything worked fine. 11 full loaded multithreading computers (21 clients and server) heated up the room so much that a permanent draught wasn't enough to cool the room down.

The basic groups were finished in 45 minutes at the second attempt. It was finished so fast that I hardly caught to set up all computers. There were only three buggy programs: memory consuming old Hewer that I couldn't set up to work properly, several times Onix exceed memory limit too. And I saw Jude to make an error too. The results are in following tabs:

1st group:

2nd group:

3rd group:

The AI that got through to the final were Tito, Kanec, Gmotor and after long period Trunkat from the first group, HGarden, Onix, Pela, Noesis from the second group and Swine, Goro, Pecuchet and Bestmove from the last group. Swine surprised by defeating Goro and seemed to be the champion. The time limits for the finals were extended to 400 seconds for a turn and 2200 seconds for entire match.

Tito and HGarden exactly knew when to defend and when to attack, they had an ace up their sleeve. Both of programs sometimes surprised by risky and successful attack from poor looking position. New Gmotor behaved a little like hewer, let himself to be encircled. Pela, HGarden and Onix, the champions in encircling, took advantage of his weakness. Onix even managed to totally encircle and choke Goro one time. Swine showed that it could use the space and not only stick stones together like others.

I didn't catch to watch first opening, because I was busy setting up the computers, the tiros tournament and the Fast-game tournament. The second opening leant to attack-defend switching battle with building up positions. The third opening was very interesting. AIs was building attack independently with quiet watching opponent. As soon as one of them started to defend it lost the game or win in the following attack. In the one of the last and interesting games Swine defeated Pela using 6 seconds per game whereas Pela needed 21 minutes. This allowed swine to get the point and to outrun Goro.

Final results:

Unequivocal champion of tournament was new Tito that had only one draw with Goro and lost only 7 games of 64. The second was HGarden with one draw and one lost and the third was swine. In the final, Kanec made one error and Onix seven.

The most time consuming brain was Pela with 55s/move and 1290/game. The quickest brain was Trunkat (278/move) followed by swine (1600/move). The most memory consuming (except Hewers 643MB in basic group) was Onix 360MB, the least memory consuming was Tito 1.32MB.

In the end there was a fast game tournament again which was won by Tito as well. This year's novelty was tournament of new programs with 400 seconds for a turn and 2200 seconds for entire match. The best tiro was Onix.

Quick tournament:

Tiros (newbies) tournament:

You can download:
Recorded results of the tournament
Openings file for the P. Lastovicka's Piskvorky manager

Match againtst human champions (11.11. 2006)

Someday in the summer 2006, we were contacted by Bretislav Krcek of piskvorky.net, who proposed a match of the best brains against best Czech human players of their league. The match was set to November in a pub in Prague with following conditions:

• best three programs will match three humans of piskvorky top ten
• the best program will match the best player, the second will fight the second and for the third likewise, each match consisting of two duels
• the time limit for match will be half an hour for each program and each human
• we will use the equal opening shown on this picture

We played on three notebooks, two of them having speed of 1.6 Ghz while Goro, who didn't need all of his time, played on a slower computer. The third best brain, HGarden, fought Tomas 'Teovan' Nemec. The match was quite equal, each duel was won by the player who started resulting in a draw. Second best brain, Tito matched Honza 'Gadael' Stradal. Of the three participating programs, Tito adapted the best way for the longer time limit and so he managed to defeat Gadael in both duels. The top match was between Goro and Pavel 'Kedlub' Laube. Goro didn't exploit all the time available and he fell to Kedlub in both matches.

Overall, there was one victory, one defeat and one draw game for the humans. The results were influenced by the fact that brains aren't adapted for the higher time limits while even the time limit of half an hour was considered insufficient by some human players. Some players would also prefer a quiter place than a pub. You can download some photos and matches and watch the matches using the P. Lastovicka's Piskvorky manager. In the downloads, you'll find the up-to-date versions of HGarden and Goro that participated.

Seventh tournament:(28th April 2006)

The seventh tournament took place at The University of West Bohemia in PC lab UI 303. There were 11 computers Intel Pentium 4 3.4GHz HT, 1GB RAM in the lab. Drawing lots to decide basic groups was similiar as the last year. Considerable change was playing from openings (watch the image). One opening was chosen by Ales Rybka (gomoku player from www.piskvorky.cz known as Kaprar) , two others by me.

On Thursday evening I had all my stuff prepared and I have found out, that my old program Beta was not able to deal with openings. Because of confusion and hurry before the last tournament I had chaos in sources and versions of old version of Beta. So I decided to try to complete (in any working form) unfinished, entirely new version of my Beta. I had finished at half past five AM, I have been sleeping for three hours and went up to the lab. In the lab I have figured out, that my new Beta is compiled in a bad way and doesn't work, so I was forced to eliminate it. It took me one hour to set all computers up and draw lots to decide the basic groups.

Programs was sorted according to the last tournament rank. Instead of removed Beta, there was one program (Tito) drawn lots from the group of not qualified and new programs to replaced it. Then one member of each qualification group was drawn lots from every trinity. The rest was drawn lots from the group of new and not qualified programs.

The basic groups was finished very fast (approximately in one hour). New Kanec was really a surprise, because it won the first basic group and defeated even swine, the former champion. Tito was present to be very good as well. The results are in the next three tables.

Then I estimated the duration of the final group and I have chosen the time limits 300s per move and 1200s per match. Final group was really suspensful. HTML interface have been automatically reloading and the placings have been continuously sorting. At first, Kanec's chances was very good, then Tito and HGarden were catching it up. This three programs have been alternating in the the top of table. Then Goro started to fight through from 4th position. A moment before the end, Tito was still the leader and there were several interesting duels which determined the winner. In the meantime, Pela and Swine overtook Kanec with big effort. The best program in using time limits was Pela 818s per match, the worst QMentat 17s per match. It is interesting that Goro won even if it needed only 33s per match. It is obvious that Goro has still big reserves. There were no errors and no timeouts in the finals.

This tournament brings 4 new versions and one entirely new program. In contrast to the last tournament, there was much more time to think and programs had a lot of time to show off. Futhermore, the influence of presence or absence of the open book was decreased because of the openings. I checked the usage of CPU randomly (via Task manager) in final group. I didn't catch any program to use more then 50% of CPU, so propably no program took advantage of two processor units.

In the end, I launched a fast game (5s per move, 30s per match), the result are in the table below. Pela stood up for the last victory.

Some programs didn't respect time limits in the fast game, especially bestmove and brutefor.

Sixth tournament:(16th April 2005)

The sixth tournament had slightly modified rules. One author may not submit more than one brain. The brains were allowed to maintain some data on hdd between the matches, which they could use to learn something from their opponents. Because the number of competing brains has become very high - 18, we had to start using qualification groups. There were three groups, 6 players each. The three best programs would continue to the finals. Both the groups and the finals were played with system each-with-each for 10 matches in the Gomotur manager. Other conditions were same as before. You can dowload all the replays of the 6th tournament for Gomotur.

The programs were divided in the way that they were sorted according to the last tournament rank with the new ones placed at the bottom in alphabetical order. Then one member of each qualification group was chosen randomly from every trinity (groups marked by different colors):

1. Swine - Jiri Fontan
2. Goro2005 - Viktor Barykin
3. Bestmove - Wen XiangDong
4. Beta(nastupce Hewera) - Tomas Kubes
5. Trunkat - Jiri Trunkat
6. Noesis - Mark Mammel
7. Puskvorec - Patrik Bezdicek
8. Qmentat - Vojta Havranek
9. Pisq7 - Martin Petricek
10. Leonardo - Milan Matlak
11. Benjamin - Benjamin Vejnar
12. AI1000 - Jan Macura
13. Curnpis - Jan Čurn
    ***Brains absent in last official tournament (alphabetical order):***
14. Brutefor - Vaclav Synacek
15. G_motor - Roman Vancura
16. HGarden - Han Bing Qing
17. Kanec - Jaroslav Babak
18. Pela - Petr Lastovicka

The qualification was played on Tuesday the 12th April on three different computers at home by three different people - Jirka Fontan, Vojta Havranek and Martin Petricek. The groups were assigned in a way, that nobody played the group where his own program was. Thank to last feature of Gomotur, we could be automatically uploading the process of qualification at web by ftp where anyone could watch it. The ftp server had some problems with simultaneous upload from three sources. Fixing that caused us a delay.

First group - Jirka Fontan

PC configuration:
AMD XP 1600+, 750MB RAM, Windows 2000

AI1000:Brutefor 5:5
Five times the same boring matches. Both programs won their beginning. AI1000 was totally crushed in the other matches.

Brutefor: Noesis 3:7
Noesis lost some important points because of too daring opening in this match, which later show fatal for him in the qualification.
Brutefor lost all other matches 0:10. It took him very long time to think, even in clear situations. The last match of Brutefor with QMentat was interesting, because they occupied two separate parts of the board. This match ended by Brutefor's timeout.

Noesis : Bestmove 1:9
Noesis managed to win only the first match because it forced Bestmove to timeout.

Noesis : HGarden 0:10
HGarden didn't give Noesis a chance.

Noesis : QMentat 5:5
This was the most exciting match, because the chances were almost equal and it was to decide, who continues to the finals. QMentat needed at least a draw. Noesis won all his beginning with no problem. QMentat's beginning turned into long matches all of which were finally won by QMentat, who placed 3rd in the group and made it to the finals.

In the match between HGarden and BestMove, both brains won easily their beginnings resulting in a 5:5 draw. None of them gave a chance to QMentat. HGarden was also the one to think the longest time, 7.23 in average.

Second group - Vojta Havranek

PC configuration:
Sempron 2500+, 512 MB DDR400 dual channel, Windows XP SP2

According to last tournament results, new Goro was estimated to be a champion of the second group. He really won the group using his typical offensive style where both brains seem to build their lines on separated parts of the board. There were two newbyes in the group also - pbrains G_Motor by Roman Vancura and Kanec by Jaroslav Babak. Even though the former had problems with match timeouts and the latter was corrected to be usable short time before the qualification, they were very ambitious. I was curious for the match Trunkat vs Kanec which was to decide, who gets to the finals. Kanec won this match with no problem. Trunkat, Pisq and Leonardo didn't make it to the finals. Btw: Leonardo presented a funny error - whenever Goro started at 10, 9 Leonardo tried to play at the same square and lost.

Third group - Martin Petricek

PC configuration:
Athlon64 3000+, 1 GB DDR400 dual channel, Windows XP SP2

After initial problems with beta (some missing libraries) third group started at 13:48 Quite interesting was beta and pela match ... these two formed almost a circle at one moment :o) Pela was thinking the longest time from all of the brains. Puškvorec timeouted once (match timeout), the results were nicely separated - 15,12,9,6,3,0

3rd group	Swine.exe	Beta.exe	Pela.exe	Puskvorec.exe	Benjamin.exe	CURNPIS.EXE	rank	wins ratio
Swine.exe	-	7:3	8:2	9:1	10:0	7:3	1	15(41:9)
Beta.exe	3:7	-	6:4	10:0	10:0	10:0	2	12(39:11)
Pela.exe	2:8	4:6	-	10:0	10:0	10:0	3	9(36:14)
Puskvorec.exe	1:9	0:10	0:10	-	9:1	8:2	4	6(18:32)
Benjamin.exe	0:10	0:10	0:10	1:9	-	7:0	5	3(8:39)
CURNPIS.EXE	3:7	0:10	0:10	2:8	0:7	-	6	0(5:42)

Finals

The finals were open for public, in the PC lab of the Prague University. There was three of us, me - Vojta Havranek, Jirka Fontan and Tomas Kubes. We used 4 PCs with speed near to 2GHz for a distribued tournament. The process was being broadcast on the web, this time in two different ways. A HTML table with single match results was being automatically uploaded. Single replay files were being uploaded also, which should have allowed others to watch single matches with Gomotur. This feature unfortunately failed.

The mathches were quite equal this time, compared to those in qualification. QMentat, who had already problems with qualification, gained a draw with Kanec only and placed last. The only matches he won were those where he forced his opponnent to timeout. Kanec, who had problems with 30 seconds for one turn, suprisingly defeated Pela and placed 8th. HGarden, which seemed very dangerous, threw an exception a few times. T. Kubes's new brain Beta, that he sent at last moment proved to be very good. The ultimate winner was noone but program Goro by Viktor Barykin, who defeated even the second Swine by Jirka Fontan and third Pela by Petr Lastovicka.

30 seconds tournament

Finally we set up an unofficial quick tournament with time limit 30 seconds for a whole match. Brains Bestmove, G_motor and HGarden, who had no chance of obeying the limit were removed. We added Krysa and Dolores instead. The winner was Pela before Goro and Beta. You may download the 30 seconds tournament replay here.

Unofficial tournament:(13th April 2004)

      In the 5th tournament the manager didn't work well and it was sending the new pbrains bad info about the remaining time for match. This may handicap the pbrains Bestmove and Hewer and that's why Jirka Fontan ran another tournament with a new, fixed manager. It ran on celeron 2GHz with Win XP. The brains matched each other 10 times, so the results were more accurate. First three positions didn't change, while the next placing changed dramatically. You can download replay of the unofficial tournament 2004(320KB) for Gomotur manager.

5th tournament:(3rd April 2004)

      Tournament 2004 took place on Saturday 3rd April in PC lab of Charles University in Prague, the same place as in 2003. We used celerons 1700MHz. The only competitors present was Jirka Fontan and me - Vojta Havranek. This was the first tournament where we used the new protocol. Martin Petricek, who tried to make a manager for both protocols, didn't have enough time to debug the program, which later caused several problems. There was plenty of new brains competing. This was the first time we had foreign competitors: Goro by Victor Barykin from Russia, and Bestmove by Wen XiangDong and Noesis by Mark Mammel, both from USA. There was also new Svine2, Hewer2, Qmentat1.5 and Pisq7. Bestmove and Hewer2 used the new protocol. We also changed the time limits. For the first time, there was a limit for match - 3 minutes, and the turn limit was prolonged to 30 seconds.

      Because we did't have enough time, we decided to play just 4 matches for one duel. As we feared, there was a lot of bugs in the new manager. It didn't check the consumed time properly and it had big problems when two pbrains(brains for new protocol) matched each other. So we decided to let play the brains for old protocol agains each other using the old manager first. When I was putting down the results home, I discovered, that even the old manager has an error and it has unfortunately overwritten some of the replay files. Then, we played the duels of pbrains with them in the new manager. Finally, we tried to play the match of Bestmove with Hewer2, but the manager used to freeze after every single match. We had to play the four matches one by one. While watching the replay, the new manager shows false time data sometimes. The correct information is in the .rec files(see Readme.First!.txt).

      The tournament was quite spectacular, and it seemed that the victory would take Svine2 or Goro. AI1000 sometimes made incorrect moves when the board was nearly filled and the game seemed to be a draw. Similar mistake was sometimes done by Benjamin, who is quite a defender, so the mistake was more apparent. 30 seconds for match was no problem, Bestmove was the only one, who used it fully. The match limit was sometimes a problem for the older brains, that weren't prepared for it(Trunkat, Puskvorec). The pbrains Bestmove and Hewer2 losed some matches because of match timeout too(Bestmove with Benjamin was a draw because of two timeouts.). The Noesis brain was very strong if it started the match, but if opponent started, Noesis's first move was too away from the first sign, that he often losed. Finally the tournament was won by Svine2, who defeated Goro and made draw with Bestmove. And so Jirka Fontan defended the title for once, even against very strong brains from abroad.

You can download replay of the 5th tournament(200KB) for windows Gomotur program.

The 4th tournament:(5th April 2003)

You can download the replay of the 4th tournament(285KB)

The 3rd tournament:(8th April 2002)

You can download the replay of the 3rd tournament(100KB)

The 2nd tournament:(24th June 2001)

You can download the replay of the 2nd tournament(53KB) for program Makrosoft Gomoku

The 1st tournament(7th October 2000)

You can download the replay of the 1st tournament(23KB) for program Makrosoft Gomoku

Notes

The number in the cell in the table means the ratio of the number of wins of player in row/number of wins of player in column. There are tree points for win and one point for draw. The score (wins/losses ratio) is determining if the number of points match. If there are more AIs with the same number of points and the same score, the match between determines.

Programs in competition can be downloaded here.

Visits from: 18. 11. 2000:

Webmaster: Vojta Havránek & Tomáš Kubeš; email: gomocupgmail.com