A Game Of Cat And Mouse: An Interview With Poker Game Integrity Expert Baard Dahl
Cards, Cash and Guns
There have always been concerns about game integrity in poker. In the late 1950s, Doyle
Brunson and his motley crew of fellow Texas road gamblers travelled from town to town, seeking
out the biggest games, carrying decks of cards, wads of cash and a cache of guns. Crisscrossing
the southern states of the USA, they outplayed the fish, outsmarted the cheaters and, on
occasion, pulled out their weapon of choice. “In those days, you needed not only good cards, but
also courage. Sometimes you could win money, other times a bullet”, recollected Brunson in his
autobiography, ‘The Godfather of Poker’.
In the modern era, poker security has come to mean something different, at least, for the most
part. There are still marked decks, rigged decks and shameless colluders but there are also botfarms, dream-machines, super-users and ghosters. This subject has always been an area of
personal interest to me but even more so since Dara O’Kearney and I left WPT Global after the
hiring of admitted cheater Ren Lin.
For us both to sign with another brand, we needed total confidence that it took game integrity
very seriously. The issue of real-time assistance (RTA), in particular, has become an existential
threat to online poker. The idea that someone, somewhere, prompted by a machine, is playing
either ‘perfect’ or ‘perfectly calibrated exploitative poker’ undermines the basic premise that the
game rewards superior decision making. That is why we both leaned heavily into an old
conversation that we had with Baard Dahl, the now head of iPoker who spent more than a decade
with PokerStars, rising up the ranks to become Head of Game Integrity.
Poker Security: How Online Poker Sites Detect Cheating
Dahl stopped by for an interview on ‘The Chip Race’ in 2024, helping us delve into the topic of
poker security. One of the first things that he made clear was that poker security teams are not
primarily reactive. Player outrage and social media storms are rarely the instigators and never the
drivers. The overwhelming majority of investigations begin internally, triggered by data, patterns,
and tools long before players notice anything amiss.
Games are policed but not by a room full of former crushers, furiously scrolling through hand
histories to spot suspicious folds. Since poker security primarily operates in the aggregate, a
single outrageous decision is unlikely to be the tell-tale heart. 95% of the time, the bad actors
are discovered because of their frequencies over hundreds of thousands of hands. That statistic
matters because it suggests that most cheating is detected without player involvement, which is
precisely how a good security system should function. Like a good referee, you only notice it
when it fails.
Dahl explained how his day-to-day role was not chasing individual cheaters but rather setting the
strategy, anticipating future threats and making sure that his teams were equipped to deal with
them. One of those teams was focused on bot detection while another looked at collusion. While
poker knowledge was essential across the board, data analysis was often more important than
theoretical understanding. Bot investigations, in particular, rarely involved looking at individual
hands but instead relied on massive datasets, statistical correlations and long term behavioural
patterns.
Why Proving Cheating Is So Difficult
One of the most revealing parts of the discussion concerned uncertainty. Players often assume
that cheating investigations are clear cut and that guilt is obvious once the right data is
examined. The reality is much messier. Many cases involve conflicting signals, with some data
point suggesting wrongdoing while others point toward innocence. This is especially true in RTA
cases, which can take an exceptionally long time.
Messier still is the burden of proof. The tools may strongly suggest solver usage, but the real
question becomes whether the evidence would stand up in court. If the answer is no, then even a
player who is likely cheating does not have their funds confiscated. That distinction is
uncomfortable but necessary. Barring a player is a serious action but confiscating their money is
more serious still. Bear in mind, for professional players, being barred can effectively end a
career, and Dahl is clear that this reality weighs heavily on investigators.
Checks and balances exist precisely because the cost of error is so high. This also explains why
sites are cautious about transparency. While players understandably want reassurance, revealing
too much about detection methods risks turning security into an arms race. This is a dilemma for
poker security teams because, while secrecy is beneficial, certain disclosures might be a
deterrent.
The Role of Players in Poker Security
When pressed on the contributions made by players, Dahl did acknowledge that they still had a
very important role. He reassured us that every accusation made by a player is investigated by
PokerStars and he was adamant that players need to keep being vigilant, reporting suspicious
behavior whenever and wherever they see it.
Dahl also emphasized the importance of being part of a culture that shuns those known to have
cheated. He specifically criticized the strain of thinking that treats cheating as cleverness rather
than theft, comparing it to other competitive environments, such as athletics, where cheaters are
considered pariahs.
Dahl also assuaged the fear that such a ‘call-out’ culture could lead to endless false accusations,
saying that there is and always will be noise in the system but that player reports, even imperfect
ones, have often provided vital starting points. They narrow the search space and give
investigators threads on which to pull.
Can Online Poker Ever Be Fully Secure?
Poker security is ultimately about trust, and trust is cumulative. It is built slowly through
consistent behavior and consistent policies. One of the reasons that PokerStars has retained
credibility, despite numerous scandals elsewhere in the industry, is the institutional memory
within its game integrity team. Each investigation left behind lessons, patterns and tools that
were applied retroactively.
Dahl describes this as iterative learning, saying how single cases revealed nuggets of information
that reshaped and refined his team’s detection methods. Over time, those small improvements
compounded, allowing them to revisit old cases and look at them through new lenses. It is
analogous to cold cases being reopened with advances in forensic science.
Perhaps the biggest takeaway from our interview was the uncomfortable truth that poker will
never be immaculately secure. It cannot be. That means the goal must be resilience, not
perfection. The players understand that it is a constant game of cat and mouse but what they
need are sites that detect most cheating, deter some of the rest, have systems that are selfcorrecting and do absolutely everything that they can to make affected players whole again.
Comments