Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How to evaluate KQ vs KR?

Author: Uri Blass

Date: 05:14:09 05/06/04

Go up one level in this thread


On May 05, 2004 at 14:35:10, Vasik Rajlich wrote:

>On May 05, 2004 at 07:12:22, Tord Romstad wrote:
>
>>On May 05, 2004 at 05:11:53, Vasik Rajlich wrote:
>>
>>>On May 04, 2004 at 12:40:09, Tord Romstad wrote:
>>>
>>>>But in fact, I prefer not to use TBs even at the PC, for aesthetical
>>>>reasons.  Using more than 1 MB of disk space just in order to play KQKR
>>>>endgames perfectly is really ugly.
>>>
>>>Of course same could be said for alpha-beta, and searching 20 million positions
>>>just to move a rook to an open file. (Ok, 5 million for Gothmog.)
>>
>>Aesthetics is always a subjective thing, of course.  My personal preference
>>is to always let the engine figure out as much as possible by itself, based
>>on general knowledge without too many complicated special cases.  A huge
>>lookup table containing the exact number of plies to mate for all KQKR
>>positions is about as far from general knowledge as you can possibly come.
>>
>>Tord
>
>Well, for KQKR I'm sure you can in a few hours come up with something effective,
>even at low search depths. However, for things like KRPKR and KPPKP, tablebases
>are a really nice solution that you won't easily replace.
>
>Note also that 20 elo points is nothing to scoff at. If you speed up your engine
>by 40%, that's about what you'll gain, and we've seen how far some people will
>go to get this. You're probably at the point with Gothmog where you'll happily
>work for a month to get a ten-point increase.

No

Based on reading an old post of dann Corbit Tord probably can get more than 10
elo improvement only by a better compiler, but it seems that he does not care
about small linear improvement and go for the big exponential stuff.

Uri



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.