Author: walter irvin
Date: 19:57:12 08/29/99
Go up one level in this thread
On August 29, 1999 at 21:02:51, Mark Young wrote: >On August 29, 1999 at 20:34:52, walter irvin wrote: > >>i wish you were not right , but you are .pondering on can make huge difference >>.one case would be . >>1. program A makes a predicted move by program B .the time control is 30 sec >>move ave .program A in this case used 45 sec ,program B moves instantly ,just >>for this game lets say program B plays a different move than predicted by >>program A .now program A is forced to start from scratch #1 burn more time #2 >>allow program B to look further and further ahead .now depending on a programs >>ability to prdict the other programs moves and the simplicity of the position >>.it could turn out to be worth as much as 200 + elo ,especialy when you factor >>in time trouble .there are lots more situations where this could create problems >>.so i guess 1 machine is not a real way to test program elo . > >pondering worth 200+ elo! :) not always for sure ,but in certain situations it could be .more so if a program is able to predict the right move ,in a fast time control ,less so at slow time controls .could be worth alot if it is able to secure a time advantage in a close game that would lead to a win on time .it would be worth less in a game where you had to make x number mores in certain time .alot more in games like 5 0 or 15 0 ect .time management is extremely important .also just knowing the other player is short of time is of much importance .maybe the only true way is a auto player and 2 computers .then there should be no question .also i have not tried it yet , but i'll bet pondering is even worth more on same program .because then the program will have very high % of right guesses . at real slow time control i dont think it would matter that much at all .
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.