Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: "Fruit 2.2.1 has no partial extensions and it extends too much"

Author: Uri Blass

Date: 07:45:42 11/02/05

Go up one level in this thread


On November 02, 2005 at 09:31:13, Jouni Uski wrote:

>Yes Uri seems to be right in this. In most of my tactical test suites engines
>like Shredder, Fritz and even SlowChess are much = twice as fast! But still
>Fruit has won all matches so far. So Fruit's secret is NOT the search, but?!
>Only in my endgame test suite Fruit scored 41/43 = the best ever!
>
>Jouni

Note that fruit extends too much single reply extension but on the other hand it
does not extend mate threats and I think that partial extensions for mate
threats is also productive for tactical strength.

When I said that fruit extend too much I did not mean to say that it extends
more than other programs in everything but only that it extends too much single
reply extensions.

extending less single reply extensions and extending more mate threats can
probably be productive for fruit.

I think that the value of extending mate threats is relatively bigger in test
suites but it also has some positive value in games.

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.