Unofficial World Championships Of Computer Programs Rapid 2019

[Krzysztof Grzelak]

Thanks again everyone for participating in the tournament. I must also apologize to Fabien Letouzey for a mistake in Moby Dam - Scan. I will apply for the next two draughts. Now I will be gradually updating my website. Now I will be gradually updating my website. All games appear in pdn format. Also the "Games" site will come after some time. I'll write it like this - see you in the next live tournament on new computer equipment.

[Krzysztof Grzelak]

Okay, I will repeat the game one more time so that it is fairly.

You are happy now ????????, you played again and the game finished by a draw.I think that you organize this tournament with the intention to support certains programs. I prefere stopping to assist to your tournament. Do you think that if all the games won by all the programs was played again, the win will be evident. A win it's a win, a lose it's a lose , a draw it's a draw. Good continuation. I am very disappointed !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Sidiki - everything has been fixed. The game was won by Scan.

[clp]

What is the end position of that game?

I don't have it; the game has been replaced.

To avoid this problem: have the arbitration performed by an external endgame database, and show an optimal win/loss path. The game could also be broken off in positions with 8 pieces or less instead of the fairly random 65-move rule.

[BertTuyt]

I would like to thank Krzysztof Grzelak for organizing the match.
I followed it every day, and really enjoyed it!

Bert

[BertTuyt]

Regarding the statement of Jan-Jaap

In human competitions we also have "tournament players" and "match players". In a tournament winning is important, in a match not losing is important. Any thoughts on this?

I guess this is also valid for Computer programs, and the paradox is that the program with perfect knowledge will never win a tournament.

Bert

[Rein Halbersma]

I would like to thank Krzysztof Grzelak for organizing the match.
I followed it every day, and really enjoyed it!

Bert

Hi Bert,

Does this mean that you have made peace with Krzysztof and are "back" in business here?

[BertTuyt]

Rein, it means that I appreciated the huge amount of work Krzysztof Grzelak did, and I viewed the games every day.
And I thought it was fair from my side to share that with the community.

Bert

[Rein Halbersma]

Rein, it means that I appreciated the huge amount of work Krzysztof Grzelak did, and I viewed the games every day.
And I thought it was fair from my side to share that with the community.

Bert

good to hear!

[Krzysztof Grzelak]

I am inviting to download the file with all games of championships. File available on the website http://www.uwccpi2019.draughtsprograms.eu/end.html

Krzysztof.

[Krzysztof Grzelak]

To avoid this problem: have the arbitration performed by an external endgame database, and show an optimal win/loss path. The game could also be broken off in positions with 8 pieces or less instead of the fairly random 65-move rule.

I think the solution is simpler. Play 5 minutes for 100 or 120 moves so programs can finish the game to the end.

[Sidiki]

Regarding the statement of Jan-Jaap

In human competitions we also have "tournament players" and "match players". In a tournament winning is important, in a match not losing is important. Any thoughts on this?

I guess this is also valid for Computer programs, and the paradox is that the program with perfect knowledge will never win a tournament.

Bert

Hi Bert,

You have right, it's a very real paradox to see that programs with perfect knowledge don't "ALWAYS" win tournaments. I thing that it's also due to the playing style of program also due to the setting. A more agressive program with a perfect knowledge will win more games against weakers programs than another with a setting that making it play opened and simplified games.

[clp]

To avoid this problem: have the arbitration performed by an external endgame database, and show an optimal win/loss path. The game could also be broken off in positions with 8 pieces or less instead of the fairly random 65-move rule.

I think the solution is simpler. Play 5 minutes for 100 or 120 moves so programs can finish the game to the end.

If there is no decision after 5 minutes, 100 or 120 moves then the game results in a draw I assume?

[Krzysztof Grzelak]

Sorry to ask Bert if you have a new version of the program Damage.

[BertTuyt]

Krzysztof, i did not find the time last years to work on Damage or Computer Draughts. But at least your tournament motivates me to restart, but I have no concrete plans on short term.
Anyway I might react now and then on posts in this forum.

Bert

[Krzysztof Grzelak]

We sincerely thank you for your answer Bert. It is a pity that the version I have cannot be played with the processor AMD Ryzen Threadripper 1950X. The tournament would be even more interesting because I have 128 GB of RAM.