Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Robert Deep Blue question

Author: Thomas Mayer

Date: 13:42:35 01/13/03

Go up one level in this thread


Hi Bob,

> I once played (using blitz, prior to Cray Blitz) in a human event, and I made
> an ill-advised change between rounds.  The effect was that in a certain
> case, it would play the _worst_ possible move (sort of like reversing alpha
> and beta).
> It made this amazing looking sacrifice that was 100% unsound and the
> opponent refused to take it.  He asked after the game "if I had taken the
> queen would I have gotten mated?"  I responded "nope, you would have just
> won a queen."
> He was amazed...

well, that remembers me to the first version of Quark that was playing on ICC -
it had a big bug in the alpha beta algorithm and cutoffs at extremely wrong
points. Anyway it achieves a quite good rating on ICC compared to it real
strength because most humans did not take the offered queen or other material.
After fixing that bug Quark did play quite better against comps, but anyway it
took a long time to get the same rating on ICC. I think most people think: "If
the comp let me take the queen there must be something fishy... so better I do
not take it".

Greets, Thomas



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.