Does the person sitting next to you at the poker table think they are playing a fair game?

I’ve come to the view that examples like the above are mostly irrelevant. Actual cheating at online poker is not what killed the game. What became toxic is a widespread perception of cheating.

The first thing to understand is that most cheating is unsophisticated and, as a result, is fairly easy to catch - by the reputable companies at least - through internal tools and analysis, not by the players who think they’ve spotted it. The poker rooms have access to vastly more data and context than any individual at the table, so they catch vastly more. That was certainly the case at PokerStars, which invested a lot in good people, systems and tools in this area. What you also got was a fair number of false reports:

And that is the key. Players are correct to conclude that cheating exists. They are mostly wrong about whether it happened to them (unless they were a victim of the UltimateBet/Absolute Poker scandal, probably the most egregious example from poker history). There are only a small minority of long-term winners in online poker, so even the average player, someone who is genuinely competent, is a net loser over time. People are terrible at recognising their true position in the poker/gambling hierarchy, even when the data is clear. Put those two things together and you get the now-common perspective. They didn’t lose because they were worse, they lost because others are cheating. And the thing about perspectives is that they are very hard to change.

Despite scandals and less sophisticated detection tools, back in the boom years the damage was fairly localised. There were enough easy games to go round that the genuine incidents of small scale collusion, and the imagined ones, never made a real dent in how the product felt. Now you’ve had over a decade of players churning, the games are far harder, and the odds of crossing paths with someone who has soured on the whole thing are much higher. The number of people carrying a negative view reached critical mass years ago. When you were the only one among your friends who held it, it stayed moderated, or internal. Once it’s a chunk of your social group, and almost ubiquitous at your Thursday night home game, it becomes the default mood music, and it reaches the ears of new players quickly.

This is not a problem unique to poker, and it is worth watching where prediction markets are heading. Retail participants lose to better-informed, better-resourced counterparties for entirely structural reasons, and they will reach for entirely structural villains to explain it such as insider trading or front-running. PMs are their own worst enemy in this regard, running markets on events which have already happened, or that are quite clearly open to manipulation or privileged insider information access. Shayne Coplan often mentions that insider knowledge is a feature, not a bug, but this is a luxury opinion of an operator. Players, poker or otherwise, have a certain baseline expectation of fairness (reasonably or not). Once enough people are convinced the game is unfair, their perspective on the product will not be changed by appeals to theory. And unhappy players are vocal.

In the previous posts in this series, Not neutral and Not for me, the main issues were driven by operator choices. In this case they aren’t, and they have no real answer to it. They can talk up modern ML and AI detection, tighten their monitoring of the environment their clients run in, and catch more. None of it touches the weight of opinion. Players seeking an unfair edge have always made it an arms race, and every time they surface a more sophisticated method of cheating, they also confirm the suspicion that there was something to find. Detection and reassurance have always pulled in opposite directions.

Do modern AI tools make things even worse? Almost certainly. There are research papers suggesting LLMs can beat solvers in certain spots, and it is genuinely unclear how a site would respond to a player using one for decision support, or whether they could reliably detect it at all. But I think the more likely outcome is simpler: the perception that players ‘teaming’ with an LLM are winning will take hold, and at that point there will not be enough fingers to plug the holes in the hull.

And, as with every other part of this series, the ‘pros’ have done a spectacular job of making it worse. Ghosting. Stables. Selling equity in ways that quietly distort the incentives in major tournaments. Quietly colluding against high-profile whales. The very base of the poker pyramid probably isn’t aware of most of this, but the more time you spend in the ecosystem, the more apparent it becomes. When even the figureheads are overtly flawed, the whole system suffers.

So is there anything the sites can do about it? I don’t think so. It is the same clash that runs under everything else - the nature of the (existing) business against the nature of the product. The mass of players who churned out and hardened into a cynical wall of voices are growing in size every hand, every day. The sites have nothing to say to them, other than a reactivation email and a deposit bonus. The biggest shame is that so many of them still love the game. They just hate the product.

Philip Atkinson, CEO, May 2026