Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: GM accused of cheating on a Chess server

Author: Russell Reagan

Date: 19:42:39 11/10/03

Go up one level in this thread


On November 10, 2003 at 21:29:15, Mike Byrne wrote:

>It's not one specific move - it 's the pattern of moves - playing the exact same
>moves as a program over a fixed number of moves.  In real life, that does happen
>over and over again in sequential games.  It's the same method that a chess
>program clone is identified.  Computers will play the same move over and over
>again in the same position.  It's identifiable and repeatable- in fact every
>program gives off a "signature" in every game that can be identfied.  No human
>will play 8 moves in a row in exact sequence of a program and then do that again
>in the next game.  Humans, GM or not, do not play like computers in that type of
>sequencing (carbon copy 8 moves in a row non/opening/non end game - essentialy
>random positions ) and then do it again the very next game.  It may happen once
>in a long while, it will not happen in back-to back games.

That sounds good, but I have to disagree based on factual evidence. I decided to
test out your theory, because it didn't sound logical to me. I assumed a player
playing the same moves as the computer would be very common. So I downloaded a
PGN from a tournament (just the first one I could find at TWIC,
http://www.chesscenter.com/twic/ecc03.pgn), and tested a few games. Here are the
results so far.

Analyzed with Ruffian 1.0.5

Game 1: P. Svidler - S. Cicak

P. Svidler statistics
41 moves played
Different move:	17 times
Streak of 1:	7 times
Streak of 2:	6 times
Streak of 3:	5 times
Streak of 4:	4 times
Streak of 5:	2 times

S. Cicak statistics
40 moves played
Different move:	21 times
Streak of 1:	8 times
Streak of 2:	3 times
Streak of 3:	3 times
Streak of 4:	2 times
Streak of 5:	1 time
Streak of 6:	1 time
Streak of 7:	1 time

Game 2: K. Fahrner - S. Cicak

K. Fahrner statistics
41 moves played
Different move:	21 times
Streak of 1:	11 times
Streak of 2:	5 times
Streak of 3:	4 times

S. Cicak statistics
40 moves
Different move:	23 times
Streak of 1:	7 times
Streak of 2:	3 times
Streak of 3:	2 times
Streak of 4:	1 time
Streak of 5:	1 time
Streak of 6:	1 time
Streak of 7:	1 time
Streak of 8:	1 time

These were the FIRST TWO GAMES that I tested, and in the first two games of the
same event we have the same person with a streak of 7 moves and 8 moves that the
computer would have recommended, had he been analyzing the game while playing
online. Maybe if I had tested thousands of games and only found two of the same
person, I would believe your theory, but these were the first two games I
tested.

Either you're wrong, or something very unprobable happened ;-)



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.