Author: Enrique Irazoqui
Date: 16:26:45 10/20/98
Go up one level in this thread
On October 20, 1998 at 18:12:16, José de Jesús García Ruvalcaba wrote: >On October 20, 1998 at 17:55:17, Enrique Irazoqui wrote: >>It would be too easy to detect such a cheat. The time delay per move during >>auto232 games is about 2 seconds/move in DOS programs and half as much in >>Windows programs. Since all programs track also the opponent's clock, a >>difference between both clocks bigger than some 80 seconds before time control >>would be noticed. >> >>Enrique >> >But then, how do you know what really happened? >Did a program report less time for its own moves? >Or did the other program report more time for its opponent moves? Take a look at autoplayed games and you will see that while in book there is a delay of 0.5 to 3 seconds between the moment a program sends its move and the moment the opponent receives it. If thinks work normally, after the first 40 moves you will see no more than 80 or 90 seconds difference between the clock for the opponent in program A and the clock for its own moves in program B. More than this would be noticed and suspected. Since all programs play against several others and this anomaly would happen with only one of them... Don't forget either that many of these autoplayed games are watched during play. I looked at this very many times and only once I have seen an exception to all this, not very significant and probably due to a bug. One more point: to take advantage of cheating with the clock, a program would have to give to itself a pretty outrageous amount of extra time. We would all notice it quite quickly. Enrique
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.