Author: Robert Hyatt
Date: 07:19:21 09/03/02
Go up one level in this thread
On September 03, 2002 at 07:00:03, Gian-Carlo Pascutto wrote: >On September 02, 2002 at 19:32:58, Robert Hyatt wrote: > >>On September 02, 2002 at 15:53:20, Gian-Carlo Pascutto wrote: >> >>>On September 02, 2002 at 13:12:33, Robert Hyatt wrote: >>> >>>>Yes. but suppose the positional scores are +/= three pawns, as I have seen >>>>out of fritz? >>> >>>So...what's your point? What do the positional scores have to do with this? >> >>Where have you been during the discussion? >> >>To back up and repeat again... >> >>mtd(f) does poorly when an evaluation can produce a signficant "swing" in >>the score after only one more move is made on the PV. IE a big variance >>due to king safety, pawn majorities and potentially won endgames, or whatever. > >This isn't as much related to having big positional scores per se, but >to the stability of the evaluation. I'm pretty sure Fritz's evaluation >is rock solid. > >-- >GCP This means that the program is not very "aware" positionally, which I don't buy looking at analysis posted in CCC. An additional ply should show you more about a position, both tactically and positionally. If your eval doesn't pick up on that, there is a problem... If it does pick up on it, then it makes mtd(f) implementations "hunt" more than they should. Which was my point all along. I've seen large eval swings from fritz results posted here... I have seen it not vary much at all, too. But those seem to be more tactical than positional...
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.