TalkChess.com

by Werewolf » Tue Jun 05, 2018 11:10 am The issue is whether scaling of MCTS has inherent properties - like alpha beta did for ages before Lazy - where we can predict with confidence what the efficiency will be on N threads and how hard it gets as N rises. If it turns out that at 12 it begins to dr...

New run with a Kaissa that has been updated with the latest Stockfish sources of 24 hours ago, there is not so much change in scores but somehow now Stockfish/Kaissa is doing much more iterations, each with a lower increase in score. I have no idea why exactly that is. The net effect is not so much ...

I seem to remember some modified version of Crafty that was even better at detecting these kind of fortresses etc. but it was some time ago probably in the old forum before the move to phpBB, I don't have the author's name at hand. Something to do with with a "Modified Crafy solves this position in...

I seem to remember some modified version of Crafty that was even better at detecting these kind of fortresses etc. but it was some time ago probably in the old forum before the move to phpBB, I don't have the author's name at hand. Something to do with with a "Modified Crafy solves this position in ...

Ok, this is the last one :) , special because it got over 24 pawns in the end, for the first time after a long regrouping of PV. 2b1q3/p7/1p1p2kb/nPpN3p/P1P1P2P/6P1/5R1K/5Q2 w - - Engine: Kaissa IV (512 MB) by T. Romstad, M. Costalba, J. Kiiski, G. Linscott 58 6716:25+24.20 1.e5 Kg7 2.Rf6 Nxc4 3.Qxc...

Final update: 2b1q3/p7/1p1p2kb/nPpN3p/P1P1P2P/6P1/5R1K/5Q2 w - - Engine: Kaissa IV (search_041, one thread, i7 6700, 512 MB) by T. Romstad, M. Costalba, J. Kiiski, G. Linscott 55 3098:35+21.84 1.e5 Kg7 2.Rf6 Nxc4 3.Qxc4 Qxe5 4.Qc2 Qxf6 5.Nxf6 Kxf6 6.Qc3+ Ke7 7.Qh8 Bd2 8.Qxc8 Bb4 9.Qb7+ Ke6 10.Qe4+ K...

Hi jon and Mike, as far as I can see both moves are interchangeable in what order they are played; transposing 2b1q3/p7/1p1p2kb/nPpN3p/P1P1P2P/6P1/5R1K/5Q2 w - - Engine: Kaissa IV search 041 (one thread, i7 6700, 512 MB) by T. Romstad, M. Costalba, J. Kiiski, G. Linscott 48 59:01 +12.42 1.Rf6+ Kg7 ...

Hi jon and Mike, as far as I can see both moves are interchangeable in what order they are played; transposing 2b1q3/p7/1p1p2kb/nPpN3p/P1P1P2P/6P1/5R1K/5Q2 w - - Engine: Kaissa IV search 041 (one thread, i7 6700, 512 MB) by T. Romstad, M. Costalba, J. Kiiski, G. Linscott 48 59:01 +12.42 1.Rf6+ Kg7 2...

Hi Dann! 8/1k4b1/1p1p1qbp/p1pPp1p1/P1N1PpP1/1PBP1P2/7R/3K2R1 w - - Doesn't seem to be much difference as for the order of the two moves: 8/1k4b1/1p1p1qbp/p1pPp1p1/P1N1PpP1/1PBP1P2/7R/3K2R1 w - - 0 1 Analysis by CorChess 2.6 211218 64 POPCNT: 1. +- (5.60): 1.Rb2 Bf7 2.b4 cxb4 3.Bxb4 Bxd5 4.Bxa5 Bxc4...

The strength difference seems too large. No contempt lowering could compensate. For small difference I have said that I think now Stockfish is tuned with contempt so much against itself I would just leave it on always unless it bothers you but it is anybody's guess what actually happens until you tr...