2nd Crazyhouse Computer Championships 2017

[Ferdy]

I plan of running this champ this coming 12-27 to 12-29. You may send your new engine or updated engine to me or show me the link to download.

Last year's Champion was Stockfish Zh
http://ccva.challonge.com/1st_ccva_comp_champ_2016

2017 format:
Phase 1: Candidates tour
Phase 2: Champ tour

All engines except Stockfish zh will play in the candidates. The winner in the candidates will challenge Stockfish zh to a 12 game match at TC 15m + 30s.

The candidates will be an 8-RR tour at TC 5m+3s

Each engine is only allowed 1 core, no book, and no learning. I will be using cutechess gui without adjudications and will use pgn start opening set at random, each opening will be played twice.

Possible Candidates' players:
Imortal
TJchess
Sunsetter
CrazyWa
Sjaak II
Sjeng
Pulsar
NebiyuAlien
KKFChess
TSCP zh
PyChess

[IanO]

Send a special invitation to DeepMind, they might surprise us all!

[jhellis3]

It would be nice. I imagine beating SF 95-5 in Zh might shut some people up. Then again hopeless trolls are just that.

[Fabian Fichter]

It will be interesting to see whether CrazyWa can challenge the top 3 seeds of the candidates. Do you have a new version of Imortal for the tournament?

For the final, I would suggest to use the last release https://github.com/niklasf/Stockfish/re ... net-071017 of Stockfish that is used for about two months now on lichess, so it should be stable. I will check with Daniel Dugovic whether he agrees.

Send a special invitation to DeepMind, they might surprise us all!

AlphaZero Crazyhouse would be awesome. It would probably beat Stockfish by quite a margin considering the amount of effort that has been put into development of chess and shogi engines compared to crazyhouse.

[Vinvin]

Send a special invitation to DeepMind, they might surprise us all!

Yeah, that would be great to get some test games at vert high level !!
ZH opening books still so weak now !

[hgm]

It will be interesting to see whether CrazyWa can challenge the top 3 seeds of the candidates.

It won't. I have not done any work on it whatsoever, the past year. And if I would find some time to work on it, it would be used for implementing Kyoto Shogi, not for improving Crazyhouse.

[Daniel Shawul]

AlphaZero Crazyhouse would be awesome. It would probably beat Stockfish by quite a margin considering the amount of effort that has been put into development of chess and shogi engines compared to crazyhouse.

NebiyuAlien can use MCTS for crazyhouse in this tournament. All I need to beat Stockfish is alpha's awesome NN for move selection and evaluation...

[Ferdy]

Do you have a new version of Imortal for the tournament?

I have.

For the final, I would suggest to use the last release https://github.com/niklasf/Stockfish/re ... net-071017 of Stockfish that is used for about two months now on lichess, so it should be stable. I will check with Daniel Dugovic whether he agrees.

That is good stable engine is important to avoid stopping the running tour unnecessarily.

[Ferdy]

Send a special invitation to DeepMind, they might surprise us all!

they already tried shogi, but maybe zh later.

[Toadofsky]

Best of luck to all participants! Thanks again to Ferdinand Mosca for running these excellent events!
(Fabian did check with me and I agree with that release. Presumably we will use setting "Move Overhead=1000" which I believe Stockfish used in TCEC.)

Incidentally, Lichess is releasing a broadcast (beta) site where viewers can discuss live tournament games! Were it feasible (for the live PGN to be hosted on any web server so Lichess could poll & broadcast the games) I (~2000) and probably others would be glad to livestream commentary outside of working hours.

It will be interesting to see whether CrazyWa can challenge the top 3 seeds of the candidates. Do you have a new version of Imortal for the tournament?

For the final, I would suggest to use the last release https://github.com/niklasf/Stockfish/re ... net-071017 of Stockfish that is used for about two months now on lichess, so it should be stable. I will check with Daniel Dugovic whether he agrees.