Search found 1308 matches

by Tord Romstad
Fri Mar 30, 2007 11:40 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SEE on non-capture moves in main search
Replies: 62
Views: 80898

Re: SEE on non-capture moves in main search

So you think he uses SEE to get a _real_ estimate of whether a capture wins or not, and then orders winning captures using MVV/LVA, adding more overhead? Yes, that is precisely what I do. The additional overhead isn't a problem at all. My program is so slow from the outset that I can add almost any...
by Tord Romstad
Fri Mar 30, 2007 11:30 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SEE on non-capture moves in main search
Replies: 62
Views: 80898

Re: SEE on non-capture moves in main search

For the non-capturing moves after the hash move and killers, history move ordering performs better than no move ordering, as expected. But trying to refine this by using SEE values does not work. I have tried to sort the moves into two groups, first searching non-captures with SEE value zero ordere...
by Tord Romstad
Thu Mar 29, 2007 12:43 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Kibitz score reporting in server play
Replies: 16
Views: 18427

Re: Kibitz score reporting in server play

hgm wrote:CentiPawns? Joker uses Pawn = 256! :lol: :lol: :lol:
It does? I thought I was the only one to use pawn = 256. :)

I convert the scores to centipawns before printing them, though. Otherwise many GUIs would resign on behalf of my engine every time it lost a couple of pawns.

Tord
by Tord Romstad
Thu Mar 29, 2007 12:39 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SEE on non-capture moves in main search
Replies: 62
Views: 80898

Re: SEE on non-capture moves in main search

Well, my gut reaction to this is that it would not matter, exactly because there are so few PV nodes. Are the sub-trees from PV nodes really that much bigger then from all nodes? I would guess that this matters only a factor ~2. The sub-trees below PV nodes are certainly a lot bigger, but exactly h...
by Tord Romstad
Thu Mar 29, 2007 10:13 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SEE on non-capture moves in main search
Replies: 62
Views: 80898

Re: SEE on non-capture moves in main search

So singling out a good one would be much more effective than discriminating against SEE<0. If even a good one does not do it, you are likely to be in an all node and any further effort is wasted time. This is true for non-PV nodes, but not for PV nodes. Try to measure the relative frequency of the ...
by Tord Romstad
Thu Mar 29, 2007 10:03 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SEE on non-capture moves in main search
Replies: 62
Views: 80898

Re: SEE on non-capture moves in main search

I think the problem is this: with things like hash moves, good captures, killers, etc, by the time you have tried all of those, either you have already failed high, or you are never going to fail high at this position. Perhaps, but if it were as simple as this, wouldn't the use of history counters ...
by Tord Romstad
Tue Mar 27, 2007 5:54 pm
Forum: Computer Chess Club: General Topics
Topic: WCCC 2007
Replies: 24
Views: 30648

Re: WCCC 2007

Reread the charter, there's no such thing in it :twisted: Tony It must have been 'conveniently' rewritten then. It has become the European championship, in reality. Not that strong North American programs stay away in total, just that most of them do. It would make sense to me to rotate by continen...
by Tord Romstad
Wed Mar 07, 2007 6:56 pm
Forum: Computer Chess Club: General Topics
Topic: so what is a fair GM handicap vs Rybka?
Replies: 54
Views: 129995

Re: so what is a fair GM handicap vs Rybka?

In NFL football, games are made theoretically even by give one team points. Example, a bad team may be given a 21 point handicap to make the contest even. . with similar time controls to the current match, what handicap would the GM need so the match is considered even? Maybe none of the above. Giv...