Search found 163 matches

by nionita
Sat Sep 16, 2017 5:02 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: tt move vs null move
Replies: 21
Views: 5941

Re: tt move vs null move

Barbarossa does null move at d==1. I just made a branch to try this. But it will take days until I can see significant Elo results. I will report when I have them. Now the results are clear: doing null move for d >= 2 brings 9 Elo (after 52000 games) in Barbarossa. So doing null move also at d == 1...
by nionita
Sat Sep 16, 2017 7:55 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: tt move vs null move
Replies: 21
Views: 5941

Re: tt move vs null move

nionita wrote:Barbarossa does null move at d==1. I just made a branch to try this. But it will take days until I can see significant Elo results. I will report when I have them.
Now the results are clear: doing null move for d >= 2 brings 9 Elo (after 52000 games) in Barbarossa.
by nionita
Sat Sep 16, 2017 7:28 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: LMR and killer
Replies: 9
Views: 2453

Re: LMR and killer

In Barbarossa since the beginning I did reduce the killers but had in mind to extempt them some das from LMR. But when I implement it, it was a lost of a few Elos. So now Barbarossa reduces the killers just like any other move.
by nionita
Fri Jun 16, 2017 8:07 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: On-line engine blitz tourney June
Replies: 41
Views: 7200

Re: On-line engine blitz tourney June

The monthly on-line engine blitz tourney for June will take place on: Saturday June 17, 3pm Boston time (21:00 Amsterdam time ) To connect: winboard -zp -ics -icshost winboard.nl -icshelper timeseal -fcp ENGINE.exe -fd ENGINEFOLDER -autoKibitz (for UCI engines, add -fUCI) It is possible that I will...
by nionita
Fri Jun 16, 2017 8:05 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Check extension vs LMR
Replies: 26
Views: 7642

Re: Check extension vs LMR

In micro-Max King Safety and Check extension are not just dependent on presence of the Queen, but on game phase (basically total non-Pawn material). If this drops below a certain value (I think <= 2 Queens + 1 minor), the King gets drawn towards the center, and check extension is switched off. I ca...
by nionita
Fri Jun 16, 2017 6:06 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Check extension vs LMR
Replies: 26
Views: 7642

Re: Check extension vs LMR

Excellent point, this one with (no) check extension in end game. I will try this one for sure. Unfortunately this one was a clear loss of at least 10 Elo with 95% (more probably 20 Elo, but I don't want to spend more CPU on it to know exactly how much). The condition was: if move checks && side to ...
by nionita
Tue Jun 13, 2017 10:31 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: End game and Null move
Replies: 8
Views: 2162

Re: End game and Null move

hgm wrote:Could you win KPK and KRK before that?
KRK yes, KPK often. But I did not do extensive tests with KPK.
by nionita
Mon Jun 12, 2017 9:58 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: End game and Null move
Replies: 8
Views: 2162

Re: End game and Null move

When I suppressed null move for STM having no piece a few months ago, the win was 6 Elo.
by nionita
Sat Jun 03, 2017 2:46 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Check extension vs LMR
Replies: 26
Views: 7642

Re: Check extension vs LMR

I never tried the latter, because I very much doubt that it has enough generality. If you look at a Crazyhouse or Shogi game, for example, it is full of checks on unsafe squares. Usually to lure away pieces that prevent check drops elsewhere. It is of course a good idea to distinguish checks that a...
by nionita
Sat Jun 03, 2017 1:21 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Check extension vs LMR
Replies: 26
Views: 7642

Re: Check extension vs LMR

For SEE not to work out is very unusual , I have tried removing the SEE test from many engines and it is always better with SEE. Every engine is different, but would suggest double checking your implementation. Without SEE, the search blows up eventually. A built in bench command will see a fairly ...