TalkChess.com

Hi Larry, why MCTS search in Komodo is non-determinicstic despite using 1 thread only? I see some randomness during the search. Each run gives slightly different output. For example Leela's search is deterministic on single thread. Here is Komodo 12.3 on 1 CPU in starting position: First run: 0.02:...

Hi Larry, why MCTS search in Komodo is non-determinicstic despite using 1 thread only? I see some randomness during the search. Each run gives slightly different output. For example Leela's search is deterministic on single thread. Here is Komodo 12.3 on 1 CPU in starting position: First run: 0.02: ...

Thanks. BTW, what's the relation between depth and number of threads? Is depth 55 on 20CPU deeper than depth 59 on 1CPU? I've never understood paralell search in SF.

And here is latest Stockfish after 82 hours on 1CPU. I had to abort the search. r1bq1rk1/pppp1ppp/2n2n2/8/3BP3/2N5/PPP2PPP/R2QKB1R w KQ - 1 8 82:29.13 8.Be2 (głębokość 59, 1/40), 423,9G pozycji (1,4M na sekundę), 2,6M końcówek 0.00: +1.36/3 8.e5 Nxd4 9.Qxd4 0.00: +2.39/4 8.e5 b6 9.exf6 Nxd4 10.Qxd4 ...

So now i't trying to find absolute best move in the following position: r1bq1rk1/pppp1ppp/2n2n2/8/3BP3/2N5/PPP2PPP/R2QKB1R w KQ - 1 8 I think there are two serious moves: Be2 and Bxf6. I'm running analysis on my machine with latest Stockfish dev (1CPU, 4096MB hash, Analysis Contempt: Off) and after ...

Thanks, I think it's quite convincing now.

I get your point Louis but there is a difference between 'much worse' and 'lost'. It's clear that black is much worse here but is he lost? This is what Uri asked in the first post, I belive. There are many opening positions where engine gives +1.80 and later it occurs that it's not possible to conv...

the opening is refuted for practical purposes. if other lines (ruy lopez, petroff, etc.) give black 0.40, why would black hasten to a 1.80 score? :D If Ruy Lopez and Petroff give black 0.40 then why people keep playing the king's gambit 2.f4? Again, after 2.f4 white is much worse but is he lost? We...

So it seems that the main goes like: 1.e4 e5 2.Nf3 f6 3.Nxe5 Qe7 4.Nf3 d5 5.d3 and now black has a choice: 5...dxe4 6.dxe4 Qxe4 5...Nc6 5...Be6 So if we want to refute the opening - and I belive that there is a refutation - we should analyse these three positions to a great depth. Note that after 5...

So it seems that the main goes like: 1.e4 e5 2.Nf3 f6 3.Nxe5 Qe7 4.Nf3 d5 5.d3 and now black has a choice: 5...dxe4 6.dxe4 Qxe4 5...Nc6 5...Be6 So if we want to refute the opening - and I belive that there is a refutation - we should analyse these three positions to a great depth. Note that after 5....