Author: Dieter Buerssner
Date: 14:16:02 03/17/01
Go up one level in this thread
On March 17, 2001 at 12:20:37, Ernst Walet wrote: >>> 8/8/pq1PQ1kp/6p1/8/7P/5PK1/8 b - - [Me] >>I tried this position with Yace. Something strange happened: [seeing black loosing for some iterations, then seeing almost drawish score again, after which seeing black loosing again ...] >Well, no problem for CM8K, could it be the case that you maybe use incomplete >sets of tablebases? What make you think, that this may be caused by incomplete TBs? Actually, I have a complete set of 4-men TBs and no 5-men TBs with Qs. I have some 5-men TBs, so I really have an inclomplete set. But, I think, no engine should have any problems with incomplete sets of TBs for positions, where no promotion is possible (or for positions, that are not in the TBs, and TB positions are only reached deeper in the search tree). BTW. I tried hard, to exclude those artifacts in Yace for positions that are in TBs, and for which promotion is possible, and for which no, or not all promotion TBs are available. Perhaps Uri is right, and some deep threats of perpetual checks are possible in this position. I do some search extensions for this. Sometimes this helps, sometimes it can hurt, because in general, with more extensions a smaller average search depth will be reached. Ín the PV, I have posted, you can see a rather long line with many checks. Regards, Dieter
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.