Notes › Detection & fair play
Note 03 · Integrity
Detection and fair play
Once you accept that the only real bot incentive sits on the Sweeps Cash side, fairness stops being abstract. It becomes two concrete questions: are the cards dealt honestly, and can the platform tell a program apart from a person? Both have well-understood answers.
Fair play rests on two layers. First, dealing integrity: a certified random number generator decides the cards, independent of any single player’s actions. Second, behavioural integrity: timing patterns, action regularity, multi-tabling fingerprints and device signals that distinguish automated play from human play. The currency model from the previous note focuses this scrutiny exactly where the incentive is — Sweeps Cash redemption.
Layer one — dealing integrity (RNG)
Before behaviour matters, the cards have to be honest. Reputable platforms use a certified random number generator: a system whose output is tested by an independent lab for statistical uniformity and unpredictability, so that no sequence of cards can be anticipated and no account can be favoured. This is the part players most often suspect (“the deal is rigged”) and the part that is most thoroughly audited. A bot cannot “read” a properly certified RNG; it can only play the same uncertain cards faster or more consistently than a human would.
That distinction matters for the bot question. Automation does not break the deal — it cannot see the next card any better than you can. What it can do is execute a strategy without fatigue, which is a behavioural advantage, not an informational one. So the defence shifts to the second layer.
Layer two — behavioural integrity
Humans are gloriously irregular. We hesitate, misclick, take a sip of coffee, vary our timing with the texture of a hand. Programs, unless deliberately disguised, are not. Behavioural detection looks for the absence of human noise and the presence of machine regularity:
- Timing distributions. Human decision times spread out and correlate with how hard the spot is; a naive bot acts with suspiciously uniform or instantaneous timing.
- Action consistency. Strategies that never deviate, never tilt and never make a human-shaped mistake produce a statistical signature over thousands of hands.
- Multi-account & multi-table fingerprints. Several “players” sharing a device, an environment or a behavioural rhythm point to one operator running many seats.
- Session and input signals. Mouse paths, focus changes, client telemetry and environment checks that an automated harness struggles to reproduce convincingly.
No single signal convicts. Detection is cumulative — a weight of evidence across many hands and sessions — which is why patient, low-volume cheating is harder to catch and high-volume grinding (the only kind worth doing on Sweeps Cash) is easier.
What “fair” means when one currency is virtual
Here the sweepstakes model returns. On the Gold Coin side, fairness is mostly about a clean, enjoyable game — there is no prize to protect, so policing automation hard would be effort spent guarding tokens that go nowhere. On the Sweeps Cash side, fairness is also financial integrity: protecting a redeemable balance from being farmed at scale. That is why detection effort, account review and redemption checks cluster around Sweeps Cash. The part of the platform that can pay out is the part that gets watched.
Putting the three notes together
The arc is simple. The currency model decides where a bot would have any reason to operate. Dealing integrity removes the fantasy that a bot can see the cards. Behavioural integrity catches the realistic threat — a tireless, regular program grinding the one redeemable currency. Read in that order, the dramatic “Global Poker bot” search resolves into something calmer: a narrow incentive, a certified deal, and a detection surface pointed straight at it.
Work on game integrity, RNG certification or anti-automation detection? If a detail here is wrong or dated, say so.
Reach out