Search found 1191 matches

by mjlef
Tue Jul 10, 2007 6:56 am
Forum: Computer Chess Club: General Topics
Topic: Are these persons incompetent ?
Replies: 52
Views: 19977

Re: Are these persons incompetent ?

Hi Uri, Well I don't agree about the "simply stole". I don't think there is anything simple about that and besides it would be a criminal act. Maybe even in Russia, they must have some laws there. I wrote some more. But I think I'd better stop speculating now, as that serves not much purpose. Regar...
by mjlef
Thu Jul 05, 2007 1:13 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Delaying Extensions Idea (does anyone do this)?
Replies: 12
Views: 6152

Re: Delaying Extensions Idea (does anyone do this)?

OK, here is my next screwy idea: Try to order the moves you search based not just on the likelihood of them being found best, but also based on how big a sub tree they will generate. In cutoff nodes, many moves can often lead to a cutoff. So, instead of say searching a capture move which might caus...
by mjlef
Wed Jul 04, 2007 6:33 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Delaying Extensions Idea (does anyone do this)?
Replies: 12
Views: 6152

Re: Delaying Extensions Idea (does anyone do this)?

I just tried this formula: score:=constant1-constant2*(subtreenodes) shr depth; (shr is Pascal's version of >>) And used this score instead of the history values for move ordering, and it did pretty well in fixed depth searches, generally saving at least 10% of nodes depending on the position. Note ...
by mjlef
Wed Jul 04, 2007 12:35 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Delaying Extensions Idea (does anyone do this)?
Replies: 12
Views: 6152

Re: Delaying Extensions Idea (does anyone do this)?

Another possibility would be to calculate the branching factor under this node, thus getting a small, depth independent, value. But I'm not sure it'd be a good idea to ignore depth this way. The bigger the draft of the hash node, the more reliable the information it provide is. Anyway, testing shou...
by mjlef
Wed Jul 04, 2007 10:54 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Delaying Extensions Idea (does anyone do this)?
Replies: 12
Views: 6152

Re: Delaying Extensions Idea (does anyone do this)?

Anyway, that is the idea. If anyone implements this, let me know. I think an efficient implementation would need to probably keep roughly the same move ordering (grouping winning captures first, then killers, then non captures, then losiing captures). But in each group, moves which are not going to...
by mjlef
Wed Jul 04, 2007 5:48 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Delaying Extensions Idea (does anyone do this)?
Replies: 12
Views: 6152

Re: Delaying Extensions Idea (does anyone do this)?

There is an idea in here which was the subject of a paper by Donald Michie I think. (Unfortunately, I lent my copy of David Levy's 'Computers in Chess' to somebody and never got it back, but the paper was reproduced in there. If anybody in the forum has a copy they could check the authorship and ci...
by mjlef
Tue Jul 03, 2007 7:12 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Delaying Extensions Idea (does anyone do this)?
Replies: 12
Views: 6152

Delaying Extensions Idea (does anyone do this)?

OK, here is my next screwy idea: Try to order the moves you search based not just on the likelihood of them being found best, but also based on how big a sub tree they will generate. In cutoff nodes, many moves can often lead to a cutoff. So, instead of say searching a capture move which might cause...
by mjlef
Mon Jul 02, 2007 12:48 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Ed Does it Again! (and a question for Ed)
Replies: 11
Views: 6683

Ed Does it Again! (and a question for Ed)

I see Ed has finished up his writeup on how Rebel plays chess: http://www.top-5000.nl/authors/rebel/chess840.htm Great stuff! Now I owe him even more meals. One section which concerns me a bit is this: <quote> Rule no. 1 in pawn endings: the side with a pawn up will win the game. The below code (or ...
by mjlef
Thu Jun 28, 2007 3:49 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: So far, only Glaurung 2 epsilon can find, within 21 ply
Replies: 13
Views: 6612

Re: So far, only Glaurung 2 epsilon can find, within 21 ply

Yes, you must have some kind of "unstopable passed pawn" rule in K&Ps endgames. Being out of the square of the pawn gets a lot of them. Another one is if you have two passed pawns, draw a square on the board as tall as the files apart the pawns are (a square). If this is the case, then one of the pa...
by mjlef
Wed Jun 27, 2007 6:58 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: So far, only Glaurung 2 epsilon can find, within 21 ply
Replies: 13
Views: 6612

Re: So far, only Glaurung 2 epsilon can find, within 21 ply

As a human, I can see instantly that NxN wins. (the fastest?) note: No table bases or bitbases used. Please do not suggest them as a valid answear to this question. I do not even know if they are warrented here. Glaurung 2 epsilon takes 20 ply to determine this. TogaII can not determine it, given 2...