Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: computer programs cannot see a very simple draw in a pawn ending

Author: jonathan Baxter

Date: 17:04:30 06/23/98

Go up one level in this thread


On June 23, 1998 at 12:47:52, Bruce Moreland wrote:

>
>On June 23, 1998 at 09:24:16, Robert Hyatt wrote:
>
>>On June 23, 1998 at 03:19:13, Jouni Uski wrote:
>>
>>>This is no problem for Fritz4.01 - it displays 0.00 immediately (47 ply).
>>>
>>>Jouni
>>
>>if it does, it's wrong, and it's a bug.  It takes 100 plies to reach the
>>50 move rule, most programs will take 101 plies.  Unless, for some reason,
>>fritz thinks black is better, as white can force a draw quicker by not
>>running all over the board.
>
>You're really quick to pull the trigger on this.
>
>If Fritz returns 0.00 on this, it's right.
>
>If it does it in 47 plies, you can't just immediately declare that it is a bug,
>perhaps he is doing something tricky to know the truth, or is using a heuristic
>that worked.
>
>It shouldn't be discounted as a bug, rather, if you are interested, it should be
>studied.
>
>bruce

I agree. It wouldn't be *too* difficult to code for exactly this type of
position. Kings and blocked pawns = draw if there are no entry squares for
either King. But then you'd expect to see a draw at 0 ply, not 47. The 47 ply is
just how long Fritz searched---the heuristic eval just gives 0 in this case (as
does KnightCap's) and as do a bunch of other programs I bet.

Jon



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.