Daniel “Jungleman” Cates talks to Paul Phua Poker on exactly what is GTO poker and how do we play GTO effectively.
It has been suggested that Daniel “Jungleman” Cates has a brain like a super-computer, which is useful for us because game theory optimal (GTO) poker play is a challenging subject to get your head around, and harder still to incorporate comprehensively into your game.
That should come as something of a comfort to any of you who suspect you might be losing out because you don’t employ the solutions and principles of GTO poker, because the truth is, not a great many players do, at least not completely.
Poker is way too complex a game for the average human to be able to apply GTO to consistently, and even the smartest pros like Jungleman need computer simulations to figure out GTO plays in most situations.
So, why then does GTO matter, and, more importantly, what on earth is it?
What is GTO?
Game Theory itself doesn’t apply only to poker, or even exclusively to games in general. Loosely speaking, it’s the science of logical decision making, and its real-world applications can be used to inform almost any complex decision-making process, particularly those which, like poker, involve incomplete information and unknown quantities: difficult military, political, or business choices, for example. Mathematician Roger Myserson describes game theory as “the study of mathematical models of conflict and cooperation between intelligent rational decision-makers.”
What is GTO in poker?
Since a game such as heads-up NLH involves so much incomplete information, it cannot have a perfect strategy, or “solution,” like chess. What game theory optimal solutions do is approximate the closest thing to a perfect strategy, which is defined as a strategy that cannot be countered – or exploited – by another strategy.
As Jungleman explains: “If you play GTO strategy and another player plays any other strategy, if your strategy is truly GTO, then you’re gonna win some amount of money. It doesn’t try to make the most money possible, it’s just a strategy that cannot be beaten. And actually, it would be like quite a bit of money if you played enough hands. However, it’s really tough to play GTO.”
The very latest, most cutting-edge poker artificial intelligence programmes analyse billions of hands, or in some cases trillions, played against themselves, in order to ‘learn’ the most optimal plays in each situation. Now, they’re starting beat the best humans, while teaching us a thing or two.
How do you play GTO poker?
While its very unlikely that many poker players will suddenly start playing perfect GTO poker, studying game theory optimal simulation programmes like PioSOLVER can be hugely instructive, helping to fine-tweak your range selection, or give you an idea of the optimal continuation bet frequency on the flop, for example. It can also make you rethink aspects of the game you perhaps previously took for granted, as Jungleman explains.
“A lot of players feel the need to like always fast play,” he says. “Sometimes there are other ways of thinking of things that work out better than always fast playing. Like you should have some traps in some really-not-obvious situations if you’re playing optimally.”
Or, he says, players believe they should ‘bet for protection,’ when they make a pair on the flop, for example, which, according to GTO, is back to front.
“They have this idea that, ‘Oh, I’ve got to put my money in before, like, bad cards come out…’ or whatever. But the simulations do it the other way. They say, ‘I don’t wanna put money in until good cards come out.’
“If you look at the simulations, you’ll see that a lot of ways hands actually play, they actually would rather wait for favourable run outs and then put a lot of money in. So a lot of new players have this backwards.”