Author: Simon Finn
Date: 07:25:40 08/23/02
Go up one level in this thread
On August 23, 2002 at 09:57:09, Gian-Carlo Pascutto wrote: >On August 23, 2002 at 09:49:59, Gian-Carlo Pascutto wrote: > >>On August 23, 2002 at 09:35:40, Frank Phillips wrote: >> >>>[D]7k/7p/8/8/1pP2P2/p7/3K4/8 w - - > >Maybe a fix for this is to add knowledge that 2 freepawns that are >2 files or more apart cannot be stopped and thus are runners. > >This wouldn't help for the static eval here, but it would help >because right now the search sees quickly that black has seperated >runners, but it takes a lot longer to see that white has them too. >The heuristic would draw the score closer to zero faster. > >Are there any exceptions to the rule in the first paragraph? Yes. If the distance between the pawns is only two files they can be stopped if they're still in their own half of the board. For example, unsupported White pawns on c4 and f4 can be stopped by a Black king on e6. White to move: 1. c5 Kd5 2. f5 Kxc5 3. f6 Kd6 etc. In fact, the king can even capture the pawns: Black to move: 1... Kf5 2. c5 Ke6 3. c6 Kd6 4. f5 Kxc6 5. f6 Kd6 ... The pawns win if they're more than two files apart or if they've both crossed the centre line (and aren't subject to immediate capture). So, in the original position, this knowledge shouln't kick in immediately, but only after a short search. Simon > >-- >GCP
This page took 0.01 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.