Articles
About Online Chess Cheating
Have questions about chess cheating? We have answers!

About Online Chess Cheating

CHESScom
| 514 | Chess.com Help

Cheating in chess happens when someone decides that they do not want to play using their own brain, and they turn to using help from a friend or a chess engine to find better moves. This is obviously unfair to the other player, and ruins the experience of the game. It's a selfish and thoughtless thing to do. Fortunately, our research shows that fewer than 0.2% of players cheat in online chess, so it is much less common than some people think. 

While cheating (aka "Fair Play" on Chess.com) does include numerous forms of abuse, including sandbagging and multi-accounting, this article deals primarily with the most common form of cheating: using a chess engine for assistance.

Chess.com closes more than 800 accounts every day for cheating. And while that may seem like a lot, more than 70,000 new players join Chess.com each day. Still, people are understandably concerned about cheating (though cheating paranoia can be just as much of a problem!).

As a company, Chess.com is continuously investing in new technology and processes to combat cheating. It also helps when the community is informed about cheating, and to that end, this article addresses many of your most common cheating-related questions. For more information regarding fair play, see our support article.

Chess.com has closed over 550,000 accounts.

In Chess.com's history (calculated October 5, 2020) we have closed more than half a million accounts for cheating! Our projections predict we will reach 600,000 accounts closed before the end of 2022 and one million accounts by early 2024.

Currently, we are averaging over 800 closures per day for engine-use alone and an additional 120 closures per day for sandbagging, rating manipulation, and other types of fair play violations.

Closures per month Chess.com
Closure rates have ebbed and flowed as chess experiences periods of higher and lower popularity.

As you can see, Chess.com takes cheating very seriously, and while these numbers seem to be high, there are more than 20 million people playing chess on Chess.com each month, so the actual percentage is fairly low.

Of those closed, more than 500 were titled players.

While titled players would seem to have no need to cheat, there is much evidence that some do!

Of the cheaters we have detected to date, 51 (10.2%) are grandmasters. 98 (19.6%) are international masters, and 160 (31.9%) are FIDE masters. More than 90% of these players are male. 

Title players cheating on chess.com pie chart

We have written confessions from many of the players we've closed, including from some inside the top 100 in the world.

Most closures are now automated.

In order to meet the needs of our community, we've ensured that the majority of closures (77% last month) are done without the need for a human to review the case. Human review takes time, which is worse for our members because it will take longer for blatant cheaters to be closed.

We have created a system that detects suspicious cases based on a number of factors and closes them by interpreting what our systems are saying. We can't go into much more detail than that because unfortunately, the cheaters are reading this too!

Automated closures have ramped up over the past couple of years in response to the surge of new players on the site and now comprise roughly 75% of the total closures in a month.

Percentage of automated closures in the past year
On average, 70% of all closures are now automated.

We have also set up automated sandbagging detection, which first warns and then closes the member for detected violations. It closes on average 120 accounts per day and warns 250.

Chess.com employs 20 fair-play team members.

Chess.com is dedicated to preserving the integrity of the game. To that end, we invest hundreds of thousands of dollars in combating cheating each year.

This number includes six full-time analysts, four researchers, three grandmasters, and three developers dedicated to improving and maintaining our systems. Many other staff members also contribute to protecting the game, including those on the events and support teams.

You likely know (or are...) a chess player who has cheated.

Many chess players think that cheaters aren't representative of the chess community. The unfortunate truth is that, as community members ourselves, we can say that several of our clubmates and chess friends have been closed for cheating. Chess.com staff personally know and like many Chess.com members that we have closed for cheating. If you are well-connected in your local or regional community, statistically, you probably also know someone who has been closed for cheating, whether the case was public or private.

Anyone can report cheaters; all reports are acted on.

Every member on Chess.com can report a member for cheating. Each report guarantees that the reported member's most recent games will be reviewed (which may result in an automated closure in clear cases) and (if the case is complex) a fair-play team member.

If you have additional information about a suspected cheater (such as a suspicion that the member only cheated in a specific event), please include that information. It can improve our ability to build a clear case.

We often receive complaints regarding members that are deemed to be clearly cheating and have not been closed. For many reasons (including youth), Chess.com may elect to close member accounts privately.

CAPS is not cheat detection.

Chess.com regularly receives complaints about members who have an excellent accuracy score in a game. We are thrilled to see our members using game reports, but the accuracy scores that are provided are not a part of our cheat detection, and they shouldn't be considered immediately damning evidence of cheating. Even beginning players can achieve very high accuracy scores if their opponent errs early, and they capitalize on the clear mistakes.

We will grant you that if a beginner has a 99% accuracy against a grandmaster, we will be very suspicious... However, in general, accuracy is NOT evidence.

Chess.com's methods are robust.

We had our algorithms vetted by a Harvard statistician, Natesh S. Pillai, in 2016, who confirmed that our systems were functioning in the manner we intended. We have also given the tour of our systems to dozens of strong titled players (multiple inside the top 20) and tournament directors and on June 3rd, 2020 we received the full endorsement of USChess after a thorough review from experts in statistics on the ratings committee.

Chess.com uses a wide variety of evidence to prove cases of cheating. We look closely at the level of play, but we also consider multiple behavioral factors and additional data regarding member accounts. Naturally, our methods are highly confidential as revealing specific detection methods only makes it easier for cheaters to avoid detection. However, we can say that our methods consider many additional variables and are significantly more sophisticated as compared to third-party tools.

Additionally, Chess.com acts to deter cheating whenever possible. Players in official prize events are required to be on camera, to share their computer screen, and to provide a 360-view of their environment. On-site proctors may also be required.

Permanent banning is sometimes impossible.

In addition to closing cheating accounts, Chess.com can and does implement IP bans and hardware bans to prevent the creation of new accounts, but these methods are technologically limited. Most of the players we block do not return, but some knowledgeable cheaters (and spammers) are able to circumvent these methods on our site or any other site.

We never close accounts based on the say-so of strong players.

Should the ghost of Bobby Fischer himself descend from on high and declare that a Chess.com member is cheating, we will not close that account unless our fair-play team confirms this to be true.

False positives.

There's nothing worse than when an innocent player is unduly closed. At the same time, the chess community demands decisive action and strong responses to cheating. This is where our task becomes truly difficult. The first point to be made is that the rate of false positives detected by our algorithm is intentional. If we set an extremely high threshold for evidence needed to take action, we will almost never make an incorrect closure, but we will also allow tens of thousands of cheaters to continue cheating. Alternatively, if we set a low standard of evidence, we will catch almost all cheaters, but we will also falsely close many innocent players. We believe we are balancing these competing goals optimally, closing cheaters quickly and confidently with a very small number of false positives.

In response to valid appeals, we overturn approximately 0.03% of closures. That means that in August, we would anticipate that about 5-6 cases out of 18,000 closures may be overturned in light of a compelling appeal backed by sound evidence of clean, if exceptional, play.

Chess.com is always willing to defend its decisions in court and to our members. We receive numerous threats of legal action each month; we are not intimidated by these threats.

What can I do?

Chess.com NEEDS your support to combat cheating in chess. Even if Chess.com caught every single cheater within 24 hours, cheating would still be a major problem. We want to PREVENT cheating, not punish it.

As a parent, coach, organizer, or community member, you can help by communicating openly and honestly about the problem of cheating. Especially, in the case of scholastic chess cheating, our experience shows that support from parents and coaches can do a great deal to discourage our youngest members from cheating.

Additionally, please don't be discouraged! We are getting better at catching cheaters! Cheaters are not getting better at cheating


Past Q&As

More from CHESScom
Are We Entering A New Golden Age Of Chess?

Are We Entering A New Golden Age Of Chess?

How To Watch The 2024 FIDE Candidates Tournaments

How To Watch The 2024 FIDE Candidates Tournaments