Author: blass uri
Date: 03:04:33 09/03/99
Go up one level in this thread
On September 03, 1999 at 05:04:27, blass uri wrote: >On September 03, 1999 at 02:47:58, Michel Langeveld wrote: > >>On September 03, 1999 at 01:14:14, Richard Bean wrote: >> >>>this is for the kasparov vs the world game. >>> >>>I have sent email to Bob Hyatt & Eugene Nalimov, >>>but they may be asleep. >> >>I wish anybody has but I don't think so because to generate this you need al >>these 6 men tablebases. >>KQPKQP >>KQbKQb >>KQbKQn >>KQbKQr >>KQbKQq >>KQbKQp >>KQnKQn >>KQnKQr >>KQnKQq >>KQnKQp >>KQrKQr >>KQrKQq >>KQrKQp >>KQqKQq >>KQqKQp >> >>The six tablebases which exist now are in kind of >> >>KBBKNN etc. >> >>No pawns yet. > >I think that you can save a lot of space and usually get the right result if you >assume that underpromotion is illegal. > >In this case you need only KQQ vs KQQ and KQQ vs KQP. > >It is interesting to know how many mistakes they are in the KQP vs KQ tablebases >if you build them assuming that underpromotion is illegal. > >I guess that the result is correct in more than 99% of the cases. I also guess that most of the mistakes are when pawns are on the 7th rank so for practical purpose you can trust wrong tablebases(assuming no underpromotion) when the pawns are not close to promotion. Uri
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.