Author: Peter McKenzie
Date: 13:09:15 09/26/01
Go up one level in this thread
On September 26, 2001 at 09:59:55, Adrian Smith wrote: >Hi all, > >What is the most effective way of stopping people from playing the same lines >against your computer? >I've experienced a bit of it lately and want to stop it.. > >Thanks :) I ran into this when my program was playing alot on ICC. Position learning and book learning are probably the best solutions, but they are no trivial to design and implement (of course you could just copy Crafty like some people have done). I came up with a simpler solution that helped somewhat and was extremely easy to implement. Simply add a smallish random number to your thinking time for the first few moves out of book. The idea is that this will give your program a chance of choosing a different move because it is probably thinking for longer/shorter than last time when in the same position. In LambChop, I do it this way: On first move out of book, I add a random number to the thinking time. The random number is between 0 and half the normal thinking time for the move. So if you were going to think for 10seconds on the move, you now have the posibility of thinking for 10,11,12,13,14 or 15 seconds. You could do this for the new few moves out of book too. You could also allow your random number to be negative. cheers, Peter
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.