How Do You Control Online Cheating?

Two articles caught my particular attention this week, both addressing online cheating where you would least expect it—in chess and cycling.  

As the pandemic has forced many activities and competitions online, the need to police cheating in new environments has become critical.  If there are a certain number of bad apples in any barrel, as my coauthor Marc Epstein and I argue in our new book ROTTEN: Why Corporate Misconduct Continues and What to Do About It, then multiple strategies to limit it must be deployed.  Relying on the character and good will of the competitors is an inadequate answer.  

One story addressed cheating in online chess, which has gained huge popularity due to the Netflix limited series The Queen’s Gambit.  When more chess tournaments and competitions moved online early in 2020, the two most prominent venues, chess.com and lichess.org, stepped up their anti-cheating measures, requiring in one case that two cameras be trained on the player, one from behind.  Even a few grandmasters have been found to be cheating.  This being the online world, complex algorithms and artificial intelligence are being employed to detect cheating.  

The other story is from the world of online cycling.  Unethical competitors have engaged in old-fashioned doping, but also manipulating the data sent by their bicycle monitors to show they have raced faster and further than they actually have.  Race organizers have tried to police doping by asking all competitors to be part of a testing regime, by requiring racers to use identical stationery bikes with the same monitoring equipment.  Further, they are compiling records of all a competitor’s races and trials so that anomalous times can be detected. 

Controlling bad apples requires either (1) appealing to the honor and ethics of the competitors, (2) reducing the opportunities to cheat, (3) reducing the value of cheating, or (4) raising the penalties for those who are caught.  Online and real-world communities have generally defaulted to the first, preferring to hope that an honor system works.  This is increasingly combined with greater shaming, publicly condemning those caught cheating.  This seems more effective.  Reducing the opportunities to cheat involves setting rules and procedures which can be monitored closely, such as the system of cameras or use of sophisticated and sometimes expensive surveillance systems.  Reducing the value of cheating would suggest smaller purses, though the desire to attract more players makes that unlikely.  Raising penalties for cheating can be effective, including lengthy and even lifetime bans for cheating.   In November alone, chess.com has closed more than 18,000 accounts for misconduct.  Enforcement processes are expensive, as Facebook, Twitter, and other social media platforms are finding.  And the cheaters and purveyors of misinformation always seem to be one step ahead of enforcement.  

If companies cannot reduce the prizes to be won, they will have to use the other three types of strategies simultaneously to have any hope of sustaining trust in the online competitions.  As in other “tragedy of the commons” situations, the few who are seeking a personal advantage may destroy the “commons” enjoyed by the many.

Relevant articles:

The Real Queen’s Gambit: Catching Chess Cheaters

Virtual Cycling and Real Cheating: Cracking Down on ‘Digital Doping’