Author: Reinhard Scharnagl
Date: 08:35:06 02/22/06
Go up one level in this thread
On February 22, 2006 at 07:31:30, Thomas Mayer wrote: Hi Thomas, >just for curiosity: Your mate count given in this table, does that include mates >in the leaves ? Usually one reason why engines with legal move generation are so >fast in perft is, that they can stop perft one ply before the leaves because >they have already all the legal moves in hand. (Quark is something inbetween >here -> to test the legality it must not do a deepening because it checks it's >legality already before the move is made) I think, that using a fully informed legal move generator (that is supplying also check and mate information) is quite comparable to a pseudolegal move generator using a Perft, because the legality and additional information cost computing efforts comparable to making and unmaking a move. Thus I added that information to the built statistic, to become somehow comparable. SMIRF did this without specialized moving and removing routines on an AMD 64 X2 Dual core 3800+ (but using only one single core and only 32 Bit). The timing on an Intel P4 2.8 GHz has been even about 10% better compared to that results. I am about to rewrite my routines completely, because I have discovered a more dense data structure than that actually used in SMIRF. Nevertheless a fast move generator is not that important, as you might learn from SMIRF's still average playing strength. And the searching procedure has been my first such serious approach to chess programming, and my evaluation (still without a mobility component), too. Maybe SMIRF will also get a kind of an attack table later then ... Reinhard.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.