testing eval
Moderator: Ras
-
outAtime
- Posts: 226
- Joined: Sun Mar 08, 2009 3:08 pm
- Location: Canada
testing eval
Just wondering what is the minimum time control for testing eval changes? 3 0 seems to take too long... could I get any reliable data about changes made to eval at time controls less than 1 0? What would be the minimum games? Thanks.
outAtime
-
yl84
- Posts: 21
- Joined: Tue Sep 07, 2010 6:37 pm
Re: testing eval
Hi,
Testing eval or testing search, you should always look at the outcomes of the games, and verify if there are some loss on time.
If there are no loss on time, testing at 1 0 should be sufficient if you want to test a change in eval (which can appear at very low depth and thus very short time control)
My 2 cents
Nevertheless, you should search the forum to have an other answer; this subject has been discussed many times (though you restrict to testing eval changes). See for instance (but there are lot of other topics):
http://talkchess.com/forum/viewtopic.php?t=36037
http://talkchess.com/forum/viewtopic.php?t=31401
http://talkchess.com/forum/viewtopic.php?t=30677
Yves
Testing eval or testing search, you should always look at the outcomes of the games, and verify if there are some loss on time.
If there are no loss on time, testing at 1 0 should be sufficient if you want to test a change in eval (which can appear at very low depth and thus very short time control)
My 2 cents
Nevertheless, you should search the forum to have an other answer; this subject has been discussed many times (though you restrict to testing eval changes). See for instance (but there are lot of other topics):
http://talkchess.com/forum/viewtopic.php?t=36037
http://talkchess.com/forum/viewtopic.php?t=31401
http://talkchess.com/forum/viewtopic.php?t=30677
Yves
-
yl84
- Posts: 21
- Joined: Tue Sep 07, 2010 6:37 pm
Re: testing eval
This thread is also very interesting for your question
http://talkchess.com/forum/viewtopic.php?t=25461
Yves
http://talkchess.com/forum/viewtopic.php?t=25461
Yves