Exploring Modern Poker Theory: What Is GTO?

GTO stands for game theory optimal, and it’s a whole branch of mathematics. In poker, however, the term is applied somewhat loosely and is used synonymously for the Nash Equilibrium.

Nash Equilibrium is a set of strategies where: 

• Players are clairvoyant. Each player knows every other player’s exact strategy. 
• All players are maximally exploiting each other simultaneously. 
• No player can unilaterally change their strategy to improve their own expectation.

There is a lot of math involved here but instead, let’s study a simple example of Nash Equilibrium to get our bearings.

The Rock, Paper, Scissors Game

Imagine two players, Rob and Alex. Rob has a very simple strategy. He always chooses rock. As you may have already guessed, this strategy is highly exploitable as Alex can always choose paper and crush Rob.

So clearly, Rob has to employ a different strategy. But if he starts to always choose scissors, then Alex can adjust to always choose rock. Rob has to come up with a more complex strategy that Alex can’t easily exploit.

The solution to this game is for both players to stay balanced and choose each option one-third of the time, and this constitutes a Nash Equilibrium.

GTO In Poker

Poker, is of course, way more complex than Rock, Paper, Scissors, and a full solution of the game hasn’t been found. But we can break it down into smaller, easier-to-solve pieces and use GTO principles to play stronger, more-difficult-to-exploit poker, while making it easier for us to find leaks in our opponents’ games and exploit their weaknesses.

Poker can be played for many motives, such as having fun, socializing, and pursuing glory, but the inherent goal of the game is to make money. More accurately, to generate and maximize profit.

What is profit in poker? 

Profit = Opponent Mistakes – Our Mistakes

In other words, to be able to make money in poker, you don’t need to be actually good at the game, you just need to be better than those you play against. While playing GTO isn’t a requirement to be profitable in every game, what happens when you start facing tougher opposition? 

Playing poker at your kitchen table against your drunk buddies isn’t the same as playing against the world’s best at a major final table or a high-stakes cash game.

Poker can be learned by experience without any theory study, but the process just takes too much time. Some people are faster than others, but even with infinite time there is simply too much to the game to figure it out on your own.

Just like in chess, players have been accumulating knowledge generation by generation. Over time, humans discovered that move 1. a4 doesn’t allow white to play for an advantage while 1. e4 does. Now with the aid of computers, players can easily find out what is the best continuation on move 24 or even move 60 within seconds, which is something impossible for humans to achieve on their own.

Similarly in poker, humans can quickly find out that opening a hand like 7-2 offsuit from under-the-gun in a nine-max table with 100 big blinds is really bad and will cost you a lot of money over the long run. But what about a hand like J-T offsuit? Is that a profitable hand to open or not? What about at a six-max table? Does it make any difference? What if we change the stack size? What about more complex situations like facing a re-raise out of position with 40 big blinds. What type of hands are profitable calls, and what are the best hands to go all-in with if you don’t want to be exploited by a crafty opponent? 

Now we have access to GTO solvers which are tools that run calculations for us and are useful to find approximations to Nash Equilibrium for subsections of the game such as preflop ranges, flop continuation betting strategies, and much more. But the purpose of these tools isn’t to try to memorize charts and play a robotic style, but to understand the game mechanics which will help us play better poker so we are tougher to play against while also capable of capitalizing in our opponent’s leaks.

For example, you will be at a huge advantage on the poker table if next time you are sitting UTG in a tournament with 50 bbs, you know that you can open about 15% hands (6-6+, A-J offsuit+, K-J offsuit+, A-4 suited+, and 10-9 suited+) while if you are at the button you can open about 53% of your hands (2-2+, A-2 offsuit+, K-5 offsuit+, Q-7 offsuit+, 9-8 offsuit+ J-2 suited+ and 5-4 suited+).

And you’ll know the reason why you can open more hands from the button than from UTG is simply because of the likelihood of someone waking up with a strong hand behind you when you are UTG and that you will also have to sometimes play out of position when called. While if you are on the button you only have to worry about two players and you are also guaranteed to be in position for the rest of the hand.

Ranges like the ones above are outputs from GTO solvers which use supercomputers to estimate Nash Equilibrium approximations. The solver finds them by “playing” against itself, running thousands of iterations where all players in the simulation try to maximally exploit each other.

It would be very tough for a human player to come up with anything better on their own. As such, these ranges can safely be used even against the toughest opponents (assuming you can handle yourself post-flop).

But those ranges aren’t set in stone. Let’s continue with our previous example of a button tournament range with 50-bb stacks. If you happen to understand the underlying GTO principles and you know your opponents very well, you can adjust your strategy just like Alex did in the Rock, Paper, Scissors game. Basically, if you know that Rob ‘only plays rock,’ you might be able to open many more hands from that button. As much as 80 to 100% of hands if the blinds are especially weak and passive.

This was just an oversimplified explanation of GTO and a tiny example of how it can be used in real poker, but the applications are pretty much infinite, and the heuristics that can be extracted from solver work are changing the game forever. Things like when is it optimal to use small continuation bet sizes or overbet the pot, what type of flops the big blind is likely to play very aggressively, what are good turn cards to develop a leading range, how to properly bluff catch the river, what are the best hands to bluff with in different runouts, and much more.

Poker theory is fascinating, but admittedly overwhelming. Entire books could be written on single topics, which is why focusing on the right things and understanding the game mechanics is key. GTO study tools are here to stay.

The chess computer Deep Blue defeated world champion Garry Kasparov in 1996, and now everyone looks to these chess engines to help unlock more secrets about the game. Poker will be no different. These study tools will reshape the poker world and will be part of the game for the foreseeable future. ♠

Leading poker theorist Michael Acevedo is the author of Modern Poker Theory. He has more than seven years of experience in the game as a professional player with more than $2 million in cashes. He has coached hundreds of players from around the world, including poker superstars such as Patrik Antonius, Brandon Adams, and Jonathan Little, and worked with some of poker’s brightest minds such as Jonathan ‘Apestyles’ Van Fleet, Stephen Chidwick, Fernando ‘JNandez87’ Habegger, Martin Kozlov, and Bert ‘Girafganger7’ Stevens. You can learn more about his work at FlopGTO.com, and follow him on Twitter @GTOPoker.