Author: Robert Hyatt
Date: 08:57:36 11/13/00
Go up one level in this thread
On November 12, 2000 at 14:50:59, Bob Durrett wrote: >On November 12, 2000 at 13:25:15, Robert Hyatt wrote: > ><snip> > >>The same thinking applies. I am _sure_ I am going to predict his move over >>50% of the time. If he takes a long time, should I take a long time, or should >>I do a bunch of three minute searches on different moves he might choose, and >>after _his_ long think I play a move found after a 3 minute think? > > >One might argue that the computer would be able to predict a very strong >[2500+]human's move 50% of the time, but it seems most doubtful that the >computer could predict a chess amateur's move very well at all. This is one of those "who cares?" scenarios. If the opponent doesn't play the predicted move very often, he is almost certainly going to lose, no matter what he does. In those cases, accurate pondering isn't needed. And in fact, playing such a person with ponder=off really wouldn't change the game result. One other point that is getting overlooked: transpositions. If I expect you to play Nf3 and then Bc4 your next move, and you play Bc4 first, the transposition will _still_ speed up the search significantly. So even though I am pondering the wrong move, I am not wasting my time. Of course, if >the computer is playing at full strength against a human amateur, then the >computer will win 99+% of the time, so the outcome doesn't depend on whether or >not the computer thinks on the human's time at all. > >But for play against amateurs [the most of us], having the program do a >radically DIFFERENT kind of computation might make the games more interesting. >Perhaps calculation of lines during the human's turn to move could be dispensed >with almost entirely in that case. > >Whether or not a programmer would be motivated to do this is another matter.
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.