How artificial intelligence is conquering the game of poker

For Texas Hold’em professionals, it should be their dream to become the God of Gamblers, right? But how can you become a gambling god? It is mathematicians and programmers who are interested in this problem and have found the solution. The real masters know that playing poker is a combination of reality and fiction. But what exactly is the combination? How much of it should be imaginary? How much is real? What is the essence of poker? The actual poker player has to be able to get a good deal on the actual poker game. Whether poker players themselves realize it or not, or want to know, the core of this game has always been the maximization problem revealed by John von Neumann.

Online Poker Sites – AI Tools

Jason Koon is a friend and mentor of Seth Davies, one of the first and most loyal adopters of so-called “game theory optimal” poker. On the second day of the three-day Super High Roller tournament, I visited Koon’s multi-million dollar mansion. The house is located in a gated community, which in turn is located in an even larger gated community adjacent to a Jack Nicklaus-designed golf course. On Day 1 of the tournament, Koon paid $250,000 to enter, and four hours later, he was eliminated, but immediately paid another $250,000 to re-enter, only to lose all his chips again. He later texted me: “Welcome to the world of desperate tournaments. Just play your best – you can still get even in the end.”

For Koon, that evens out in the form of offline tournament winnings totaling more than $30 million (at least as much as the high stakes cash games from Las Vegas and Macau, he says). Koon started playing poker seriously in 2006, when he was recovering as a sprinter on the track team at West Virginia Wesleyan College. Playing poker has allowed him to make a good living, but it’s hard to win the highest stakes games all the time. He says, “I was actually a pretty mediocre player before slover (specific problem solver, in this case Texas Hold’em) came out, but when the second solver came out, I jumped headfirst into the thing and then I started improving, improving fast, fast, fast.”

Thanks to tools like PioSOLVER, Koon has redefined his game play to understand how to bet most effectively in different situations. Sometimes a small bet, say a fifth or even a tenth of the pot, is ideal; at other times, a heavy bet of two to three times the size of the pot is the right bet. And, while good poker players have always known that you need to play a combination of real and imaginary hands, slover provides Koon with a more precise frequency of employing a combination of real and imaginary hands, and depending on what he is holding, it can determine what is the best and worst hand to bluff with.

The best players are able to reverse engineer the AI’s strategies and build heuristics that apply to hands and situations similar to the ones they are working on. Even then, the amount of information they have to deal with is still huge. When I told Koon my idea that it would be like flipping through a 10,000-page book to remember as much of it as possible, he immediately corrected me: “It’s a 100,000-page book. This game is too hard.”

In fact, the scale of the data Koon has to work with is even larger than that. He has rented nearly 200 terabytes of cloud storage space to store the data – that is, the game tree data he has developed since he started working with solver. While you can’t get all the information at the table when you’re playing face-to-face with a human, that limitation doesn’t necessarily apply to poker played online. Automated bots, especially in lower-stakes poker games, have been a problem for Internet online poker sites since before the rise of solver, but now people playing online can use another screen to find AI strategies to circumvent the rules and play the best strategy. Koon said, “Anytime there’s a device that has high enough stakes to win a lot of money and that can potentially be used for good, there’s a way for people to turn it into a cheating tool.”

Koon isn’t particularly worried about cheating when he plays poker against people on the Internet, but other players aren’t so sure. Professional poker player Ryan Laplante said, “That’s the main reason I don’t play online tournaments anymore, I mean real poker tournaments.” In a recent $7,000-entry online tournament as part of the World Series of Poker, Laplante said that out of the 100 or so entrants, he recognized at least four names of players who were rumored to have been banned from other sites for using so-called “live assistants. Laplante believes some of the largest online sites are doing a good job of regulating tournaments, but he fears that as solvers become more common, the balance of power will continue to shift in favor of those who cheat to gain an advantage.

Back to Top