Just three cunning acts of cheating can drastically shift the odds in chess, turning an evenly matched game into an almost guaranteed victory, a new analysis reveals. This highlights the need for effective systems to detect and prevent cheating in chess.

Daniel Kellen from the University of Haifa, Israel, simulated 100,000 games using the advanced Stockfish chess engine. Famous for its unmatched performance, Stockfish can outplay any human world champion. In these matches, two computer engines competed at an average skill level (1500 Elo rating), while powerful interventions occurred from a more skilled computer with an Elo score of 3190—higher than any human’s.

Typically, playing as White confers a slight advantage with a 51% probability of winning, primarily due to the first move advantage. However, Kellen’s findings showed that leveraging a computer system like Stockfish for guidance on just one move amplifies that probability to an average of 66%. If a player resorts to three instances of cheating, the chance of victory escalates dramatically to 84%.

“We anticipated that the first act of cheating would raise the winning probability to 55%, with a second nudging it to 60%. However, cheating three times escalated the odds to an astonishing 84%,” Kellen notes. This is a crucial insight for players seeking victory.

A single, well-timed move suggested by a superior chess engine (approximately 30 moves into the game) can improve winning chances by 15 percentage points. In contrast, random interventions throughout the game yield a mere 7.5 percentage point boost.

The analysis employed a system that intervened only if the suggested move significantly enhanced winning chances compared to the player’s original choice. The thresholds for cheating became more stringent as gameplay progressed. “Initial acts of cheating are somewhat lenient; however, scaling up your cheating necessitates a marginally greater advantage,” Kellen points out.

Such tactics create a “camouflage measure,” according to Kellen, which enables cheaters to evade detection by the automated systems employed by online chess platforms. These systems may mistake a remarkably effective move as an exceptional human insight rather than a product of computer assistance.

“Proving selective cheating through manual analysis poses significant challenges, as a single strategic hint from an engine can determine the outcome of a game,” emphasizes Kellen. Kim Shu from the University of Mainz in Germany finds this research compelling.

Kellen clarifies that his work does not aim to facilitate cheating but rather to assist chess platforms in recognizing the threats posed by subtle cheating methods. “Understanding your adversary is crucial,” he concludes.

As online chess gains popularity, the community must intensify its efforts to uncover cheating instances, notes Shu. “An effective anti-cheating strategy demands a multifaceted approach,” he advises, suggesting the analysis of behavioral patterns, move timings, and comprehensive histories of online chess accounts.