Head Games: Understanding Game Theory

The Pros: Vanessa Rousso, Matthew Janda, and Derric Haynie

Craig Tapscott: Why learn how to play exploitative poker? Perhaps share an example or two.

Vanessa Rousso: Let me give you an example of exploitative poker. It folds to your opponent on the button and you’re in the small blind. Every time it is folded to this player on the button he raises. So you know the range of hands he is raising with is much wider than the normal range of hands that you would raise with. It is clear that he is not adjusting for image or trying to gain credibility by folding every once in a while. He is deviating from what would be optimal strategy. Optimal strategy would take into account your own image and folding every once in a while and you wouldn’t raise every single time it folds to you. Let’s say we see the loose button raiser get to showdown with 8-2 offsuit and he got there because he flopped three eights. Since we know he is deviating from optimal strategy, we decide to exploit this tendency. Next time he raises we are going to three-bet him. The reason that is a profitable play is because the range of hands he will call a three-bet with is about the top 15 percent of hands and 85 percent of the time he will fold. Even the 15 percent of the times he decides to play, he probably only reraises you five-to-seven percent of the time. The rest of the time he calls your three-bet and our continuation bet takes the pot down a majority of the time. Or you can get lucky and flop something. There are many ways to win there, which makes that particular play profitable. That is a profitable, exploitative play. The interesting part about playing exploitatively is you have now opened yourself up to exploitation. Back to our original example: You’re in the small blind and the same scenario unfolds as you reraise the button raiser. But the smart thinking player to your left decides to four-bet you light, knowing that you know that the opponent to your right is raising every single time. He has exploited your exploitation of that loose player to your right. You would not have opened yourself up to exploitation if you played “optimally” and only raised with a reraising range of hands. So that’s when you know you’re playing exploitatively. Fundamentally by definition, if you are playing exploitatively you are opening yourself up to exploitation by deviating from optimal play.

Matthew Janda: It’s important to learn how to play exploitative poker because that’s how you’ll make most of your money against weak opposition. The vast majority of poker players are hobbyists rather than professionals and nearly all of these players have significant leaks which can be exploited. While it’s of course possible to beat these players by just playing solidly ourselves and ignoring their tendencies, that won’t be the best way to take their chips. It’s also worth noting that the exploitative strategies which work best against weak opposition are often relatively easy to implement. For example, when playing against a loose player who loves to call but hates to fold, it’s easy to see our best strategy is to value bet aggressively against him but rarely bluff. While we could of course still beat him by ignoring our read and playing against him the same way as we would against a strong player, we won’t win nearly as much money this way. Furthermore, since we’re likely sitting at a table full of other players who are trying to win this opponent’s stack as well, it’s crucial we try to win his stack as quickly and efficiently as possible before someone else does.

Derric Haynie: Exploitative poker is the root of how to win at poker, while game theory optimal (GTO) is the root of how to play poker. Your entire goal in any poker game is to find situations where your opponents make mistakes, or errors, and capitalize on them. If no errors are being made, then everyone must be playing GTO, and no money can be made. If that is the case, you are better off just sitting out of the game entirely. So assuming you are in a game because someone at the game is making mistakes, you are absolutely better off capitalizing on these mistakes by playing exploitatively, in order to maximize your profit. Against bad players, exploitative plays might include bluffing less, value betting more, increasing your turn c-bet percentage, etcetera. Against good opponents, your “exploitative” play is generally going to be to play GTO. Against great opponents, GTO and exploitative are synonymous, because the best play you can take is also the GTO play; if you differ from GTO, you will be making less money. But, everyone makes mistakes, and just because you are going into a hand versus a great player with a GTO strategy, does not mean you have to stay GTO throughout the course of the hand. You may come across a spot where you feel they have differed from GTO themselves and now you can adjust in order to exploit their mistake. Once again, playing exploitatively means recognizing when your opponent is making a mistake, and adjusting your strategy to capture value from that mistake.

Craig Tapscott: Why invest time learning about game theoretically optimal poker?

Vanessa Rousso: I believe it’s more important to start with a good base by learning sound optimal strategy, and then branch out to exploitative poker. You can’t sit down and just try to exploit a player here and a player there. Because the things you would need to understand to play an exploitative style well are things you would only understand if you knew how to play optimally. While it is practically impossible to achieve the ever elusive truly “optimal” playing style, it is possible to tweak your game to further approach actual optimal play. The reason why is this strategy will be so much easier to implement than exploitative strategy, mainly because you can use it against anyone. That is true even if you don’t find a leak in an opponent’s game. You will minimize your losses and if you do it well enough you may actually achieve a profit. But I don’t want to ignore my response to the first question. Playing exploitative poker is where the actual profits come from, at least as a professional poker player. But my advice is to start with the core fundamentals of optimal play.

Matthew Janda: A game theoretically optimal strategy is a strategy that will beat every other non-optimal strategy. Many players are currently trying to tune their game to be closer to what they believe is game theoretically optimal (no one knows exactly what the optimal strategy looks like), because as they play against better opponents with fewer leaks it becomes more important to try to play one strategy that will beat all opponents. In other words, if the opponents you’re playing against have very few leaks, rather than try to find their leaks and exploit them, it may be better to instead focus on analyzing your own game and make sure you have few or no leaks yourself. While learning how to play close to an optimal strategy may sound tempting, it is important to keep in mind that implementing a close to optimal strategy is extremely difficult. It takes a long time and a lot of hard work for most people to develop a deep understanding of how poker works, and even once someone understands many important theoretical concepts, it is another challenge entirely to be able to implement them quickly and accurately while playing. Because of this, it’s often best to start with learning basic exploitative strategies first and then slowly beginning to learn the theory that often guides many exceptionally skilled players to play the way that they do. 

Derric Haynie: As I stated earlier, GTO is the root of how to play poker. It’s true that you can go your whole life as a professional poker player and not know a thing about how to play GTO. In fact, I haven’t really been a GTO player most of my career. But GTO play is very important to your overall understanding of the game and extremely beneficial to the decisions you make. You should always start with a GTO strategy, and move towards exploitative as opportunities present themselves. This means, if you come into a game where you don’t know any of the players, you should have a default strategy — GTO — that you use until you see mistakes. Once you recognize a mistake, like someone playing too many hands, for instance, you can devise an exploitative adjustment to capitalize on that mistake. Also, any time you encounter a situation in which you are simply not sure what is going on, you should be able to revert back to a GTO strategy. Daniel Negreanu might call this “the 23rd level,” but what it really is, is just introspectively taking a look at your range and devising a strategy from the hands within your range that prevent that range from being exploited. If that range is for some reason unbalanced, or you find yourself in a spot where you just “can’t do anything but fold,” one of two things could have happened:  a) you made a mistake earlier in the hand, and should look to study and correct that error to avoid making it in the future. b) you took an exploitative approach, which left you open to attack on certain board runouts. Essentially you got unlucky. You should still analyze the spot and make sure the exploitative approach you took is better than a more balanced one. Understanding GTO poker is better than understanding exploitative poker, because you can always revert to the default — to the GTO — strategy and it will marginally capitalize on your opponents’ mistakes any time they differ from GTO. But when you understand that, for instance, your opponent is not bluffing this river nearly as often as they should be, there is no reason to call with a GTO range, to stay “unexploitable,” just make an exploitative fold, save the money, ♠