Search found 426 matches

by h1a8
Fri Oct 09, 2020 6:49 pm
Forum: Computer Chess Club: General Topics
Topic: CCRL flawed testing : SF12 above SF12 8CPU
Replies: 59
Views: 5803

Re: CCRL flawed testing : SF12 above SF12 8CPU

Both got a very different mix of opponents. Both don't have that much games so small sample size doesn't help, but : Stockfish 12 64-bit 3666 +22 −22 89.3% −325.7 21.0% Stockfish 12 64-bit 8CPU 3658 +28 −28 63.4% −75.1 70.7% Elo transitivity flat out doesn't work, and we can get absurd results like...
by h1a8
Mon Sep 14, 2020 3:12 am
Forum: Computer Chess Club: General Topics
Topic: Fat Fritz appears to be a good sparring partner for SF 12 NNUE
Replies: 8
Views: 1635

Re: Fat Fritz appears to be a good sparring partner for SF 12 NNUE

How do you know even update fat fritz engine and its net?
by h1a8
Sun Sep 13, 2020 5:56 am
Forum: Computer Chess Club: General Topics
Topic: Grandmasters and Engines Couldn't Solve This Puzzle.
Replies: 16
Views: 2386

Re: Grandmasters and Engines Couldn't Solve This Puzzle.

With a single thread on my three years old phone, Crystal 3.0 needs only a few seconds to play Nf6+: [18] >=1.78 Nf6+ d:18/36 1:Nf6+ t:15.21 n:3543k nps:232k tb:662 h:3 After 2 min: [24] 5.74 Nf6+ Kg7 Nh5+ Kg6 Bc2+ Kxh5 d8Q Kg4 Qf6 Kxg3 Qe5+ Kg2 Qxe3 c4+ Kd5 Bc5 Qxc3 Bf2 Qxc4 Nc5 Bf5 Kg1 Qg4+ Kf1 K...
by h1a8
Wed Aug 26, 2020 3:37 pm
Forum: Computer Chess Club: General Topics
Topic: Android compile of latest Stockfish
Replies: 14
Views: 3278

Re: Android compile of latest Stockfish

Yes it works. Thanks a lot. I'm getting around 35% of Stockfish 11 nps.
by h1a8
Tue Aug 25, 2020 11:11 pm
Forum: Computer Chess Club: General Topics
Topic: Android compile of latest Stockfish
Replies: 14
Views: 3278

Re: Android compile of latest Stockfish

does the nnue even work? engine is terminating when I try to install it with nnue enabled. when nnue is disabled then engine installs and works. Before you start the engine with NNUE on, write the full path to the network in the EvalFile uci option, i.e. /sdcard/DroidFish/uci/eval/mynet.bin (or .nn...
by h1a8
Mon Aug 24, 2020 6:09 pm
Forum: Computer Chess Club: General Topics
Topic: Fun fact: SF NN net beats syzygy files easily!
Replies: 9
Views: 2156

Re: Fun fact: SF NN net beats syzygy files easily!

Can you please post results of these tests including the test sets themselves?
by h1a8
Mon Aug 24, 2020 9:28 am
Forum: Computer Chess Club: General Topics
Topic: Android compile of latest Stockfish
Replies: 14
Views: 3278

Re: Android compile of latest Stockfish

does the nnue even work? engine is terminating when I try to install it with nnue enabled. when nnue is disabled then engine installs and works.
by h1a8
Sat Aug 22, 2020 1:51 pm
Forum: Computer Chess Club: General Topics
Topic: Is there anymore unsolved positions at all!?
Replies: 40
Views: 6629

Re: Is there anymore unsolved positions at all!?

Sorry for not reading carefully. So you don't doubt the correctness of the study. Your assertion is that finishing correctly from here is not easy for humans. 4q2n/2KN1p1k/1B1PpPp1/4P1P1/8/8/6P1/8 b - - 0 35 Yes. But I now know how to win after going over all the different variations where black ha...
by h1a8
Sat Aug 22, 2020 3:12 am
Forum: Computer Chess Club: General Topics
Topic: Is there anymore unsolved positions at all!?
Replies: 40
Views: 6629

Re: Is there anymore unsolved positions at all!?

4q1kn/2Bp1p2/1N1PpPp1/1P2P1P1/2N3b1/6p1/1P4Pb/7K w - - 0 1 If the best move is all that is needed, then there seems to be a "lot" of best moves. I took "best" to mean "key". Have you followed perejaslav's link to the discussion of this study? I assume many engines will find the initial key sequence...
by h1a8
Fri Aug 21, 2020 9:04 pm
Forum: Computer Chess Club: General Topics
Topic: Is there anymore unsolved positions at all!?
Replies: 40
Views: 6629

Re: Is there anymore unsolved positions at all!?

4q1kn/2Bp1p2/1N1PpPp1/1P2P1P1/2N3b1/6p1/1P4Pb/7K w - - 0 1 If the best move is all that is needed, then there seems to be a "lot" of best moves. I took "best" to mean "key". Have you followed perejaslav's link to the discussion of this study? I assume many engines will find the initial key sequence...