Author: Robert Hyatt
Date: 14:46:51 05/30/00
Go up one level in this thread
On May 30, 2000 at 14:01:39, Aaron Tay wrote: >On May 30, 2000 at 10:25:45, Robert Hyatt wrote: > >>On May 30, 2000 at 06:55:59, Ray MacFadyen wrote: >> > > >>You should get all the 3/4 piece files. These total less than 200mb. To that, >>add kr*kr where *=pnbrq. That will get you close to 1 gig total. > > >Just wondering why do you consider Krqkr(isn't this nearly always a simple win) >important and KrnKR (draw?) important? I would think KrpKr would be most >important followed by KPPKP . Also KQPKP . KnKpp and KBKpp ... There are under-promotion issues that can make you look silly. If you don't have kqrkr, then you won't know it is a mate. Most programs will then dance around and leave the pawn on the 7th, because it is a mate sitting there, but an unknown after it promotes. Crafty can handle this case correctly, but most programs cannot. And if you are going to be using any program other than Crafty, then you ought to have all the tablebases in a "set" to avoid such problems... > >KrBKr might be drawish but is rumoured to be difficult to defend, but I would >think those I sugguest above are more important. > >Some Rook endgames just don't seem as important. > >>If you have more space, I would grab 'sets' with a "p" in them, as I suggested >for krp vs kr. > >Agreed. > > > >>ie kbpkb, kbpkn, knpkb, knpkn and so forth. Don't forget the promotion cases >>if you want optimal play... > >Yes..This IMHO should have priority over KrbKr and KrNkr..Dispute the fact that >Rooks endgames are common.. Don't forget the promotion problem. You don't want to have your program draw in a totally won position, because it would not promote.
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.