Perft Results

Programming Topics (Computer Chess) and technical aspects as test techniques, book building, program tuning etc

Moderator: Andres Valverde

Perft Results

Postby Dan » 09 May 2011, 07:33

Hi all, I've finished a move generator and am now testing Perft and my results are these:

Perft(1): 20
Perft(2): 400
Perft(3): 8902
Perft(4): 197742
Perft(5): 4897256
Perft(6): 120921506

Now, after some Googling with my numbers, I see I'm not alone in these (wrong) results. Is there a known systematic error that, apparently not only I, have made?
Dan
 
Posts: 12
Joined: 26 Mar 2011, 16:21

Re: Perft Results

Postby Dan » 09 May 2011, 15:13

Update:

I found what was wrong; there was no legality checking to my moves. I guess I have to do that before I go ahead and make them.

:-)

Best
Dan
 
Posts: 12
Joined: 26 Mar 2011, 16:21

Re: Perft Results

Postby H.G.Muller » 11 May 2011, 08:24

As most engines use pseudo-legal move generation, it might at first glance seem useful to also have perft counts for pseudo-legal moves. Problem is that there is no unique definition of what is pseudo-legal, and that it differs per implementation what the move generators produce. For instance, is castling through check pseudo-legal or illegal? How about castling out of check? Some move generators will generate pseudo-legal moves where the King wanders into check, but not moves with pinned pieces. Some move generators do switch to legal move generation when they generate check evasions. There just is too much variety. So this is why people stick to counting legal moves; those are unambiguously defined.
User avatar
H.G.Muller
 
Posts: 3453
Joined: 16 Nov 2005, 12:02
Location: Diemen, NL


Return to Programming and Technical Discussions

Who is online

Users browsing this forum: No registered users and 15 guests