I’ve come across an individual who has been accused of, and subsequently found to be, cheating at correspondence chess. The player who accused him was informed that, “unfortunately, there is no penalty for his actions”.
How can this be? Why wouldn’t this player be banned from continuing to play in CC tournaments? Now I’m in the position of playing him in a CC tournament, and it again appears that he’s using a computer. I plan to report this once the game is finished.
What, if anything, can be done to combat this? Why aren’t there penalties for cheating at CC?
How was he cheating? Different correspondence groups allow the use of computers. CC players have always had access to databases, books, magazines, and even help from friends.
I have been playing CC events on the USCF for 4 years now and have had the pleasure of only playing against 3 cheaters(confirmed using 2 different engines). I just recently reported this 3rd individual to Mr. Alex Dunne with strong evidence, like the USCF CC Rules states, after reviewing my complaint he emailed my opponent(cheater) to stop using computer software to aid in making moves. So that must be the warning…I would guess?
[b]Penalties
The CCD may assess penalties for violations of these
rules. Penalties include, but are not limited to, informal reprimands,
warnings, reflection time reduction, forfeitures, or
withdrawal. Warnings are usually issued before more severe
penalties but the CCD may skip this step. Smooth and timely
completion of games is the main consideration. Penalties will
be assessed as necessary to accomplish this purpose.[/b]
I am the webserver TD for CCLA. I remind participants at the start of each event that the effectiveness of CCLA’s prohibition against engine-assisted analysis is dependent on the players’ individual and collective honor.
Candidly, absent egregious cases (which have to be really egregious, because correspondence chess is all about working hard to find good moves), participants in engine free events have to embrace the honor system and trust their opponents to do the same. Or play ICCF, in which engine use is allowed (and in which blindly trusting an engine is fatal).
It seems to me you might more plausibly claim to have encountered only 3 opponents whose cheating was egregious enough to be detected. How could you possibly identify a cheater who used engines to screen for, say, blunders over a certain threshold?
There have been two major articles concerning cheating at OTB and correspondence chess in recent years.
Dr. Ken Regan, “Catching Chess Cheaters”, Chess Life], June 2014 issue.
Jerry Honn, “Detecting Engine Use”, The Chess Correspondent (A CCLA publication), July-September 2012 issue.
These are detailed technical articles that discuss methods of detecting cheating. They are both worth reading. My takeaway from these articles is that it is not all that simple a task to prove cheating from the game score. It does not mean it cannot be done, only that a simple solution may not be possible.
Another consideration is the nature of current databases. All the recent games are generally the results of home preparation, i.e., with the use of an engine. So when a correspondence player finds a game in a database to follow, it may be “the use of an engine, once removed”, with apologies to genealogy terminology.
Doing a database search for a game is not that much different from reading a book, like Informant or the NIC YB, to check for novelties and new ideas. Using a computer to find things is the lowest level of usage. When a program is used to explore a variety of sequential paths rather than using one’s own mind to do it, then we are using an external tool to do our work. A relatively crude too, but a tool nonetheless. I have to assume that there is some hard evidence that this was done, though none has been put forward.
Game comparison shows a measure of correlation. A strong grandmaster or a well prepared player with a pet opening can rattle off a couple of dozen moves which are indistinguishable from computer analysis. In situations where the analytical “tree” allows one to look at forced sequences, the analysis also looks like a computer did it. Even analysis projection and working back from a positional idea may also look similar to computer analysis. On an upper level, it is hard but not impossible to think or play like a computer. We can do this pretty well and is what computer researchers sought to emulate way back when. I am curious to know how the alleged cheater was caught. Did someone see him using a program? Did he let it slip that he was using one? If the method used was correlation, that is much slimmer evidence.
It is much easier to “cheat” in CC by having a strong player vet your moves, and even your engine analysis, to find holes in it. This happens more than you think. On a number of occasions a CC player has shown me his games and we found that his analysis, and the engine checked moves, were in error because of a flaw in the evaluative regime. Programs are not as good at dealing with compensation, with looking far ahead, or dealing with chaotic situations as you would think. Neither are most humans. One of the beauties of CC is that you have time to look at paradoxical situations, weird moves, and string chunks together. The result may look like computer analysis, but is not. That is probably why many CC organizations don’t care it you use a computer. It is not only difficult to police, but is unnecessary as the CC players know that their “tool” has its limits. If you have any ego, you would rather figure the problems out yourself. I have been told that in CC winning is not everything.