How to Find and Prove Collusion and Soft Play Online in 2026

Collusion and soft play are among the hardest integrity problems in online poker because they rarely look obvious in a single hand. One strange fold can be a mistake; one missed value bet can be fear or inexperience. The difference in real cases is repetition: the same two accounts repeatedly meet, avoid major confrontations with each other, and apply pressure in ways that consistently harm third players. In 2026, the most practical approach is evidence-first: collect specific hands, summarise a few measurable indicators, and submit a report that an integrity team can verify from its own records.

What collusion and soft play look like in real hands

Collusion typically means two or more players coordinating decisions to gain an unfair edge, while “chip dumping” is a related behaviour where one player intentionally transfers chips by losing on purpose. Many poker rooms treat these as serious violations because they undermine fairness for everyone at the table. When you report suspicious play, it helps to use these terms accurately and stick to observable actions rather than assumptions about motives.

Soft play is usually quieter than chip dumping. The hallmark is selective passivity: a player who competes normally against the field becomes noticeably risk-avoidant only versus one specific opponent. The pattern often appears in multi-way pots, where two suspected partners fail to challenge each other and instead choose lines that allow both to survive while squeezing out a third player.

Collusion can also show up as coordinated pressure. Examples include repeated “protective” raises that make little strategic sense unless the raiser expects a particular response from the other account, or sequences where one player creates a situation that forces the third player into a difficult decision while the suspected partner is effectively shielded. The key is that these actions become meaningful only when they happen repeatedly between the same accounts.

Hand-level red flags you can document without guesswork

Focus on hands where the line is both unusual and asymmetric. For instance, a player may routinely 3-bet or isolate with certain holdings against the pool, but repeatedly declines to do so when a specific opponent is involved. Document the spot: positions, stack sizes, board runout, and why the deviation is striking compared with standard play at that stake and format.

Watch for repeated “avoidance” on later streets. This can look like frequent checks in obvious value spots, a reluctance to thin value bet rivers, or repeated decisions that let a partner realise equity cheaply. One such hand proves nothing; a cluster of similar hands against the same opponent is far more persuasive, especially if the same player shows normal value-betting behaviour versus everyone else.

Timing and sizing can add context. Instant folds in high-frequency defence spots, repeated small bet sizes that keep pots capped only against one opponent, or strangely consistent “safe” lines can help illustrate a pattern. Treat these as supporting details rather than the core proof, because the most convincing evidence is still the strategic pattern across many hands.

Statistical markers that separate variance from coordination

Hand narratives are helpful, but numbers make the case stronger because they reduce the chance you’re reacting to variance. In practice, you want to show that a player behaves one way against the general pool and materially differently against one specific opponent. That “pair-specific deviation” is often more informative than overall aggression or overall win rate.

Start with interaction metrics: how often the two accounts sit together, how frequently they reach flops in the same pot, and how often their pots stay unusually small compared with comparable situations versus other opponents. In tournaments, also track whether their most suspicious hands cluster around high-leverage stages such as the bubble, pay jumps, or late final-table play.

Be realistic about sample size. A small number of hands can point to something odd, but it rarely proves intent. The strength comes from consistency: repeated table-sharing over time and repeated behavioural differences that persist across sessions. Even if you cannot reach a huge hand count, strong frequency evidence plus a set of representative hands can still justify an investigation.

Concrete checks you can run from your own hand history

Create a simple comparison for Player A: baseline statistics against the pool versus statistics specifically against Player B. Useful indicators include VPIP, PFR, 3-bet rate, fold-to-3bet, c-bet frequency, turn barrel rate, river bet frequency, and showdown rates. You are looking for “drops” or “spikes” that appear mainly in the A-versus-B pairing.

Add money-flow context. Track net chips exchanged between the pair, and note whether the transfer happens through suspicious lines such as repeated overfolding, inexplicable passive play with strong holdings, or repeated avoidance of value bets. In cash games, compare average pot sizes when A plays B versus average pot sizes when A plays others at the same table.

Describe your findings as comparisons, not verdicts. A careful phrasing helps: “Player A’s aggression is normal versus the pool, but against Player B it falls sharply across N hands, and the most extreme deviations occur in late-stage tournament spots.” This gives an integrity team a clear starting point without forcing you into claims you cannot prove directly.

How to report suspected collusion so an integrity team can act

A useful report reads like a structured case file, not a rant. Integrity teams need exact identifiers: who, where, and which hands. Your goal is to provide a small set of high-signal evidence that can be validated quickly against the room’s full database, rather than sending a huge unfiltered dump.

Include the essentials: screen names, stake and game type, cash table name or tournament name/ID if available, date and approximate time with time zone, and a list of hand numbers or IDs. Then add 5–15 key hands that best illustrate the pattern. Choose hands that show repeated selective passivity, coordinated pressure in multi-way pots, or strange deviations that happen mainly in the suspicious pairing.

Keep privacy and fairness in mind. Avoid claiming you “know” two accounts are the same person unless you have direct proof. Focus on observable behaviour and repeatable patterns. Also, do not publish other players’ hand histories publicly if rules restrict distribution; send details through official support channels so the operator can investigate using internal logs.

A complaint template you can copy and what to expect next

Template opener: “I’m reporting suspected collusion/soft play between [Player A] and [Player B]. Across [dates] in [game type/stake], they repeatedly shared games and showed consistent risk-avoidance versus each other, combined with lines that increased pressure on third players. I’m providing key hands and a short statistical comparison.” This makes the purpose clear in two sentences.

Then add a compact evidence section: (1) table-sharing frequency or session overlap, (2) the list of key hands with brief notes on why each hand is unusual, and (3) the baseline-vs-pairing statistical comparison. Make it easy to skim. If you attach screenshots, use them only to support context like seating and stacks, not as your main proof.

What happens next in 2026 often depends on policy: many operators will confirm receipt but may not disclose details of enforcement or investigation outcomes. That’s normal. Your best leverage is the quality of your submission: specific hand IDs, a short set of representative hands, and a measured explanation of the repeated pattern. If the behaviour continues, submit follow-up evidence in the same format so the case remains easy to verify.