Guide · Kapittel 10 Complete guide to chess fair play and anti-cheating measures for tournament organizers. FIDE anti-cheating regulations, detection methods, phone policy, and arbiter procedures.
Cheating in chess has existed as long as competitive chess has. What changed in the 2010s was the accessibility of chess engines strong enough to guarantee winning moves in any position — and the smartphone made those engines pocket-sized. This chapter gives tournament organizers a practical framework for prevention, detection, and response.
Modern chess engines (Stockfish, Leela Chess Zero, Dragon by Komodo) play at levels no human can match. A player consulting even a mid-range engine app on a phone during a game has an overwhelming advantage. Unlike performance-enhancing drugs in sport — which improve performance probabilistically — a chess engine consultation is potentially decisive on every move.
The stakes are high for organizers too. A single confirmed cheating incident at your event damages the reputation of your club, discourages future participants, and — if the event was FIDE-rated — may trigger an investigation that delays rating submissions for all players.
The good news: most cheating at club level is preventable with simple, consistently enforced measures. A well-organized event with a clear phone policy and basic access controls eliminates the most common vectors.
| Type | Method | Frequency | Detection difficulty |
|---|---|---|---|
| Engine consultation (phone) | Checking moves on a phone or smartwatch during play or during toilet breaks | Most common | Low — phone rule, search |
| Engine consultation (accomplice) | A third party outside the hall analyzes the position and signals moves (coded signals, vibrating devices) | Rare at club level | High — requires surveillance |
| Pre-game preparation fraud | Accessing opponent's preparation through unauthorized means (stolen files, social engineering) | Very rare | Very high |
| Score manipulation | Falsifying result slips, agreeing to false results with opponent | Occasional at team events | Low — cross-check scoresheets |
| Rating fraud | Deliberate underperformance to maintain a lower rating for sandbagging | Rare; mostly online | Statistical analysis required |
For over-the-board club tournaments, phone/device consultation accounts for the overwhelming majority of real incidents. The other methods require either sophisticated coordination or are more relevant to online chess. Your anti-cheating effort should be focused accordingly.
FIDE Laws of Chess Article 11.3b (effective 2023) establishes a zero-tolerance phone policy: if a mobile phone or electronic device makes any sound or is found to have been used during play, the game is lost for the player who owns it. There is no provision for warnings or second chances.
The policy applies to any sound — a notification, a ringtone, a vibration audible to the arbiter. It also applies to devices found to have been used, even silently. The arbiter does not need to prove that the device was used for cheating purposes — use of the device alone is sufficient for the forfeit.
Practical enforcement: the policy must be announced before round 1 begins. Players must be given the opportunity to surrender their phones to a designated collection point (a bag or box at the registration desk). Players who keep phones on their person do so at their own risk.
If you do not announce the phone policy before Round 1, you cannot enforce it during the tournament. A player whose phone rings after an unannounced policy may successfully appeal any forfeit decision. Always announce it — in writing in the regulations and verbally before the first round.
For unrated club events, you may adopt a softer policy (e.g., a warning before forfeit) but you must still declare the policy explicitly beforehand.
The most effective anti-cheating measures are those that make cheating physically difficult before the game begins.
Provide a labelled bag or numbered envelope system at registration. Players surrender phones before entering the playing hall. Issue a receipt. This eliminates the most common cheating vector entirely.
Only players and arbiters in the playing hall during rounds. Spectators watch from outside a physical barrier or through a separate viewing area. This prevents signal-based accomplice cheating.
Have every player sign a form confirming they received and understood the phone policy and anti-cheating regulations before round 1. This creates a legal record and reinforces awareness.
Require players to leave bags, backpacks, and coats in a designated area away from the boards. Only water bottles and writing materials allowed at the table. Reduces concealment opportunities.
For rated Standard events, implement a procedure for toilet breaks: players must inform the arbiter before leaving, and leave their clock running. Some events require an escort for the duration of the break.
For significant prize events, handheld metal detector wands scan players before entering the hall. This detects hidden electronic devices that pass visual inspection. Standard at national and international events.
When suspicious behavior is observed during a game, the arbiter has several tools available. All searches and inspections must be conducted with dignity and with a witness present.
Direct observation. An arbiter positioned to watch a player's posture, eye movement, and behavior around the clock can identify suspicious patterns: a player who consistently takes unusually long breaks immediately after making strong moves, or who repeatedly leaves the hall at critical moments.
Phone search. Under FIDE Anti-Cheating Regulations, an arbiter may request a player to empty their pockets if there is reasonable suspicion of device use. The player must consent or refuse — refusal to be searched is considered evidence of possible cheating and the arbiter may forfeit the game.
Toilet area check. If a player is suspected of consulting a hidden device during breaks, the arbiter (or a designated deputy) may inspect the toilet area for devices that have been pre-positioned (taped behind cisterns, hidden in bins). This requires a witness and should be documented in writing.
Always have a witness present during any search. Never touch the player — ask them to empty pockets themselves. Conduct the search away from other players to avoid embarrassment if no device is found. Document the time, nature of the search, and result in your arbiter's log.
A player who is searched and cleared deserves an apology. A false accusation that is handled respectfully does far less damage than one that is handled clumsily.
Statistical detection compares a player's actual moves to the moves recommended by chess engines. The key metric is centipawn loss (CPL): the average difference, in centipawns, between the engine's top choice and the move the player actually played. A very low average centipawn loss — especially over many moves — suggests the moves were computer-assisted.
To put this in context: a 2700-rated grandmaster typically has an average CPL of 20–40 per game. An average club player (1500–1800) might have a CPL of 80–150. A perfect game played by a 1500-rated player with 5 CPL is statistically implausible and warrants investigation.
Limitations: statistical analysis alone is never sufficient to prove cheating. Some positions have only one reasonable move; a player who finds that move adds nothing suspicious to their statistics. Short games, games with many forced moves, and games with engine-recommended positional sacrifices all produce misleading statistics. Statistical evidence must always be combined with circumstantial and physical evidence.
A 1600-rated player at your tournament produces three consecutive games with 2–4 average CPL across 40+ moves each, including complex tactical positions where the engine's top choice was deeply non-obvious. Each game ends in a win against players rated 300+ points higher.
This is a statistical anomaly worth documenting and reporting. It does not prove cheating — but it justifies closer physical monitoring and, if it continues, a formal referral to the federation's anti-cheating officer.
Document everything: preserve the game scores, note the time of breaks, and log any physical observations. Your job as arbiter is to document and report — not to adjudicate.
FIDE's Fair Play Commission (FPL) — formerly the Anti-Cheating Committee (ACC), renamed in 2020 — is the body responsible for investigating cheating cases at international level. At national level, your federation's ethics committee handles cases. The FPL maintains approved tools and procedures for statistical analysis.
The primary statistical tool recommended by FIDE is Chessbase's anti-cheating module, which calculates engine correlation metrics and provides a probability assessment. FIDE also uses proprietary analysis developed in cooperation with academic researchers in statistics.
Importantly, FIDE anti-cheating analysis considers not just raw engine correlation but also: the difficulty of the positions (easy positions inflate correlation naturally), the player's rating history (a sudden performance spike is more suspicious than consistent improvement), and contextual factors (behavior, access to devices, prior incidents).
FIDE classifies tournaments into three anti-cheating protection levels based on the event's importance, prize fund, and player pool. The required measures escalate with each level:
| Level | Typical events | Required measures |
|---|---|---|
| Level 1 — Basic | Club events, local tournaments, unrated rapid/blitz | Phone zero-tolerance policy announced before play. Visual supervision by arbiter. Players must leave devices in a designated area or switched off in bags. |
| Level 2 — Enhanced | National championships, FIDE-rated opens with significant prizes | All Level 1 measures, plus: random searches with hand-held metal detectors, transmission delay of at least 15 minutes on live boards, designated toilet areas monitored by staff. |
| Level 3 — Maximum | World Championship cycle, Olympiad, Grand Prix, major FIDE events | All Level 2 measures, plus: mandatory screening of all players before each round, 30-minute broadcast delay, CCTV in playing and rest areas, dedicated anti-cheating officer, real-time statistical monitoring. |
As a tournament organizer, determine which level applies to your event and implement the corresponding measures. When in doubt, apply the next higher level — over-protecting is always better than under-protecting.
As a tournament organizer or arbiter, your role when you suspect cheating is to document and report — not to investigate or adjudicate. The formal investigation belongs to the federation and, if necessary, FIDE.
Anti-cheating measures work best when they are embedded in a culture of sportsmanship rather than experienced as surveillance. Players who feel respected and trusted are more likely to report suspicious behavior they observe than players who feel they are under constant suspicion.
State your values clearly. Include a fair-play statement in your tournament regulations — not just the rules, but the reason. Something as simple as: "This tournament runs on trust. We ask everyone to compete honestly, and we take fair play seriously to protect the integrity of results for all participants."
Apply rules consistently. Nothing erodes fair-play culture faster than selective enforcement. If the phone policy forfeits a strong player's game, that forfeit must stand. If it forfeits a beginner's game, that forfeit must also stand. Consistency is the foundation of trust.
Educate rather than punish for minor infractions. A junior player whose phone rings because they forgot to turn it off is not a cheater. Issue the forfeit (it must stand), but take a moment afterward to explain why the rule exists. Education prevents repeat incidents far more effectively than punishment alone.
Recognize honest play publicly. Just as you post the standings, consider acknowledging players who report their own errors (e.g., an illegal move they made that the arbiter missed) or who demonstrate exceptional sportsmanship. Fair play deserves the same visibility as results.
Every result stored and exportable as PGN — ready for statistical review if needed.
FIDE 2026-kompatibel sveitsisk paring, 28 tiebreak-systemer, helt gratis.
Opprett gratis konto