Search found 478 matches

by Robert Pope
Fri May 24, 2019 9:08 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: QSearch perft
Replies: 24
Views: 4440

Re: QSearch perft

I think the difference is due to the fact that you don't generate check-evasions, because check-evasions will contain non captures too. So you have to generate evasions when the king is in check. You have to do this in quiescence too, otherwise your quiescence will make very big errors. Do you have...
by Robert Pope
Wed May 22, 2019 4:00 pm
Forum: Computer Chess Club: General Topics
Topic: We need more GPUs to improve the BIG LC0:)
Replies: 7
Views: 2219

Re: We need more GPUs to improve the BIG LC0:)

Damir wrote:
Wed May 22, 2019 2:56 pm
Why offer more GPU support, when Lc0 networks have not been updated for almost a week ? :o :o
This is a different network. If your GPU can't go towards the Lczero.org net, why not support this one for a bit?
by Robert Pope
Mon May 13, 2019 2:28 pm
Forum: Computer Chess Club: General Topics
Topic: Opposite Color Bishop Endgames
Replies: 43
Views: 4757

Re: Opposite Color Bishop Endgames

It's all in your mind, which apparently is running completely amok. Virtually nothing you say is remotely connected to reality, especially when you think you can read other people's mind. Apparently it outrages you that LC0 is pretty poor at tactics, as far as chess engines go. And obviously you th...
by Robert Pope
Thu May 09, 2019 9:05 pm
Forum: Computer Chess Club: General Topics
Topic: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU
Replies: 40
Views: 6824

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

They are claiming that the net is in essence a book as well as an evaluation function. Of course, in many ways, Stockfish is as well , they just spend little time tuning the part that would help discriminate between lines in the opening. But the size of the SF evaluation function is very small (the...
by Robert Pope
Thu May 09, 2019 5:54 pm
Forum: Computer Chess Club: General Topics
Topic: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU
Replies: 40
Views: 6824

Re: Bluefish vs Leela in TCEC , who will win? BF =170 Threads, Lc0 =2x GPU

People on this forum have said that LCzero has a 40 MB pre calculated cache of moves . I wish that the TCEC would run an experiment letting SF use the latest Brianfish or else disabling LCzeros cache of moves. LOL... Leela is open source, anyone can go and check for that. There's nothing like that....
by Robert Pope
Mon May 06, 2019 4:23 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: A bug is squashing me. HELP!
Replies: 20
Views: 3539

Re: A bug is squashing me. HELP!

I think you have your if (tree[i].score > alpha) { if (tree[i].score >= beta) { return beta; backwards. You need to check for a beta cutoff first. EDIT: Or on second thought maybe it doesn't really have to, but you are adding an extra "if" statement for beta cutoffs. If you are always returning the ...
by Robert Pope
Mon May 06, 2019 4:17 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Time reduction adding a/b to mini/max?
Replies: 9
Views: 1684

Re: Time reduction adding a/b to mini/max?

With a mean branching factor of 30, minimax should take about 30^6 = 729 million nodes to search to depth 6. alpha-beta search with random ordering should take around ((30 - 1 + sqrt(30^2+14*30+1)/4)^6) = 305 million nodes. The expected time taken for alpha-beta given your minimax time of 30 second...
by Robert Pope
Fri Apr 26, 2019 5:01 pm
Forum: Computer Chess Club: General Topics
Topic: Is it possible to have and EPD string wider than 79 chars
Replies: 4
Views: 671

Re: Is it possible to have and EPD string wider than 79 chars

That's only 79, though, isn't it? I think that is going to be the max, since the only way to get it longer is to add castling flags. And a castling flag would mean there is a 2 space gap, eliminating the extra character advantage.
by Robert Pope
Fri Apr 19, 2019 7:34 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: No standard specification for Perft
Replies: 22
Views: 4594

Re: No standard specification for Perft

Huh. Though if you read the actual article about RCSE-1, all it says is that it calculated the number of depth 3 leaf positions at 8902, which is equivalent to what perft(3) does, not that they invented or implemented the perft command. Somebody back in 1500 wrote down all 20 legal opening moves. Th...
by Robert Pope
Fri Apr 19, 2019 6:04 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: No standard specification for Perft
Replies: 22
Views: 4594

Re: No standard specification for Perft

If the move generator only generates pseudo valid moves, then leaf node have to be validated and thus probably make/validated/unmake or copy/make/validated. Am I wrong ? That is true. However, the wiki says to only generate legal moves. It doesn't say to only generate legal moves, it says to only c...