How Chess Sites Can Fairly Address
Suspected Unfair Play

 Last week, following the Chess.com issue covered here, I was approached by Ben Johnson of the Perpetual Chess Podcast, and one of the questions he asked me was “How do you think chess sites should deal with suspected cheaters?” At that time I told him it’s hard for me to answer this question since I’m not an expert. But having been thrust into this situation, I pondered on his question further and now believe I can offer a fair and feasible solution. I’ve taken into account that chess websites are not required to reveal any of their methods used to detect unfair play.

Chess.com and other online platforms have to recognize that for any unfair play detection process to be considered fair, it has to be fair to the accused player as well. Otherwise, they perpetuate an unfair situation in their effort to ostensibly promote fair play in chess.

This is by no means a definitive solution, and most likely portions have been discussed before. Instead, it’s something that can be used as a basis to begin the improvement process.

• To begin with, suspected titled players who the site thinks are using unfair means should be given additional options and not just solely the Confession option:

• Submit an explanation for the specific games in question.

• If the site still has issues, then agree to play games with a webcam on Skype or Zoom, till a certain sample size is collected. If the site so deems, the games can be played under the supervision of a site admin.

• Determine if the level of play in the monitored games is drastically worse, which can be defined as perhaps 25% worse than the average performance of the prior 3 months before the incident was notified. The threshold to define drastic can be whatever makes sense based on some objective guidelines.

• The player is free to play all tournaments and any prizes won during the monitoring period will only be paid out if the eventual outcome is in player’s favor.

• If the level of play and results remain similar or within range, then the site has to restore all privileges and make amends to improve their system.

• If the level of play is worse than the acceptable range, the player can be prohibited. The player may still be innocent but at least the site has made an attempt at due process. The prohibition will just have to be an unfortunate consequence of a system misevaluation.

Is this a tedious process? Sure.

But accusing a titled player of using unfair means is a very serious claim, and it’s worth putting in some more effort towards ensuring a fair process.

In the previous posting, I related how I was told by Chess.com VP Danny Rensch that a single set of games in 2015 were the only ones that they were calling into question.

Here is my proposal to Chess.com:

•I’m ready to play using a private webcam for the next 3 months or whatever reasonable duration is required.

•If my performance is considerably worse than the last 3-6 months, then I will accept the prohibition. One indicator is that my current blitz rating would begin to fall dramatically and stay down.

•If my performance is similar or within range, then Chess.com system has made an error and I should be justly compensated for I revealed the weaknesses of their system, helped them improve it, took all the nonsense, and invested a lot of energy and time to dismiss the erroneous claims. If Chess.com is 100% sure, then they will never have to worry about paying me.

•I also need to be informed about what triggered Chess.com to lock my account based on games played ~3 years ago. Why now? Why not when it happened or a few months from then? Also, the exact games from 2015 that are being questioned should be identified to me. The mystery must be solved, for as it is, using 2015 games as a reason to lock me out in 2018 sounds downright strange.

The biggest risk for me is that the data is adjusted to fit the conclusion, but that can be addressed. My best indicator remains my blitz rating.

The problem I foresee is that after claiming to be 100% sure, it can be very hard for anyone to backpedal if my performance is within the range. Admitting to false positives may be a high hurdle to overcome.

But false positives exist, and I’m far from being the only one.

Here’s an article talking about the possible errors Chess.com made in another situation. It’s a comprehensive and thoughtful read, and if nothing else just read the statements from the players who were accused. I recall looking at the FM’s games in 2015 when this incident occurred, and always believed he was wrongly accused. Many others shared that sentiment. I was unable to look at the IM’s games, though, as his account was closed before I could do so, but his explanation speaks for itself.

I can’t help but think that if I and others were able to come to a reasonable conclusion at that time that the FM games appeared genuine, then why couldn’t Chess.com figure that out or approach the issue with a healthy dose of skepticism in their systemic conclusion?

Who is providing that human understanding to the games at Chess.com when such key evaluations are done?

After my last blog post, I was also approached by another Junior titled player who admitted to giving a fake confession just so he could play in the ProChess league.

In most cases, the cost of such false positives is insignificant to Chess.com; there is a minimal impact on them for making errors and losing a few titled players here or there. In addition, the Confession is a zero-cost strategy that has worked when many, but not all, of the times the players are indeed guilty.

As Danny Renschsaid,

“We have received confessions from GMs north of 2650+ FIDE, and you would never know it because we work with them to protect their reputation, allow them back on our site, and, after honoring our requests and “serving their time”, they have all gone back to enjoying Chess.com, knowing they are playing on their second, and only “last chance” account.”

The irony of this statement is not lost on me as I stand, alongside a few, fighting to defend my innocence for a set of games played ~3 years ago, while some guilty GMs amidst us continue playing on Chess.com and no one knows about these 2600+ players.

The second account option offered by Chess.com works best only for the guilty for they can stop the unfair part and return to normal.

Players who fake confess take a gamble.

Even if I did make a “fake confession,” and resume playing with a second account, Chess.com might return with the same issue, for I can’t change the way I naturally play.

I know I’m innocent and have never used any unfair means. So it’s easy for me to put out this proposal to Chess.com.

Now I don’t even care about using the Chess.com site, as long as I still have access to the PCL.

But I raise my voice at the reason and manner in which I’ve been dismissed – accused of engaging in unfair play and told that my only option is a confession. I question the interpretation of the data and the conclusion that Chess.com arrived at, which I know is erroneous. In addition, I feel Chess.com is pressured into such evaluations anytime their marquee GMs have a worse than expected result.

Incidentally, I’m now playing primarily on LiChess which is an absolutely terrific site from a user experience standpoint.

I stream on LiChess regularly and did so during the Titled Arena a few days ago. There I had the opportunity to play by far the best player in Classical, Rapid, Blitz, and Online Bullet, World Champion GM Magnus Carlsen!

And I played him 9 times!

Playing him is always a thrill that probably even Kingda Ka cannot match.

Won’t spoil the results, but you can check out the archived version of the stream  

If you just want to know the result, click here. But that’s no fun!

Now it’s time to move on and get back to my Chess. I will see you on Li Chess and Twitch!

If you wish to stay updated, click on the Follow Twitter button.