Author: Wylie Garvin
Date: 02:51:20 02/07/02
Go up one level in this thread
On February 06, 2002 at 23:50:48, Edward Seid wrote: >Question regarding the announcement of Mate in X when using tablebases. > >How is X determined? My guess is that it is the summation of the number of >moves to conversion to the next lower tablebase. > >If true, then I have a scenario. > >Assume a 4-piece ending. Using tablebases, White announces Mate in 20. Using >the above definition of Mate in X, White sees Mate in 20 because the nearest >conversion to a 3-piece tablebase is 10 moves, from which there is forced mate >in 10. > >Now let's assume that White chooses NOT to play for the nearest conversion, but >instead plays for a 12-move conversion to a 3-piece ending in which mate is >forced in 6. In the second case, even though White chooses a longer path to >conversion, it actully results in a shorter Mate in 18. > >Is this possible? From your question, it sounds as if you believe that the Nalimov tablebases contain distance-to-conversion values. I think that was the case with the earlier tablebases created by Edwards, but the Nalimov tablebases actually contain distance-to-mate values. So the engine says "mate in 10" because that's exactly what the tablebase contained. In order to choose which move to make, the engine simply generates the possible moves, looks them all up and chooses (any of them) which is "mate in 9". cheers, wylie
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.