AI a defence as well as risk regarding ‘advantage play’ in land-based casinos: Differential Labs

Artificial intelligence (AI) and smart-table data can help land-based casino operators identify suspected “advantage play” in real time, while the player using it is still on the gaming floor, according to Mana Azizsoltani, head of analytics at Differential Labs. The technology could enable operators to compare the “count” for a particular side bet with players’ actual wagering behaviour, Mr Azizsoltani told GGRAsia. The executive said it was “something like a data arms race”, as advantage-play teams could be receiving a constant stream of card data via a camera, microphone or some other method. “But the casino also has a constant stream of smart-table data coming in in real time,” he added. “Used correctly, the casino can understand the ‘count’ for every side bet – when it is hot, when it is not, and everything in between.” According to Mr Azizsoltani, a casino “can then compare the count with what players are actually doing”. It can “identify someone who only wagers when the count is favourable, increases their wager as the count gets stronger, mainly bets near the end of the shoe, or wins unusually often across several side bets,” he noted. “Most importantly, AI can flag that behaviour in real time, while the player is still at the table,” he stated. That would give surveillance and table-games staff an opportunity to respond before a player had inflicted losses on the casino and left, said the Differential Labs executive. He added that real-time detection was particularly important in relation to “portfolio counting”, where a player might move between bets depending on which one offered the strongest advantage at a particular time. Differential Labs, a specialised analytics and AI consultancy focused on the global casino and gaming industry, recently examined the emerging risks associated with advantage play on baccarat side bets in a series of blog posts. The analysis generated a lot of interest within the industry, said Clayton Peister, managing director and analyst at Differential Labs. The consultancy said computer-aided counting and AI could allow advantage players to identify favourable wagering opportunities more accurately and wager more efficiently. Depending on the methods used and the circumstances, such activity could potentially amount to cheating or criminal conduct. The firm said baccarat side bets accounted for between 40 percent and 60 percent of baccarat revenue at many casinos, making them an increasingly attractive target for advantage players. Mr Azizsoltani said in his remarks to GGRAsia that another potential defensive application involved examining how chips move around the floor. “With chip attribution, the casino can understand who is passing chips to whom, how much value is moving between different players, and who is consistently a net giver or receiver,” he noted. “Advantage-play teams often share chips or bankroll, so this can help connect players who may look unrelated when the casino considers them individually,” Mr Azizsoltani added. Smart-table data could also accelerate surveillance investigations, the executive suggested. “If an advantageous shoe has been hit, surveillance traditionally has to pull the video and manually take notes on every card and wager.” Smart-table data allows the casino to reconstruct and analyse that shoe almost immediately – showing when the opportunities appeared, who wagered on them, and how their wagers changed,” he explained. “It turns what used to be a very tedious manual process into something that can be done quickly and, in many cases, automatically,” Mr Azizsoltani stated.