GTO vs exploitative play

 
If you have been around poker for a while, you have heard of the term ‘’GTO’’. It´s actually not a new term. It´s based on the Nash Equilibrium, and was formulated in 1950 by mathematician John Nash. If you are interested in the math and history, there is a great movie about John Nash, called A Beautiful Mind with Russel Crowe. Nash equilibrium can be used to create unexploitable strategies for many different games. 
 
In recent years there´s been more and more advanced software developed, which can create solved GTO solutions for poker, and it has become an integral part of any serious poker player to understand GTO strategy.
 
Game Theory Optimal, in poker, is an umbrella term players use to describe the holy grail of no-limit holdem playing strategy, by which you become unexploited to your opponents and improve your winrate. It essentially solves the game. No longer do you have to think what to do, when you are in a limped pot vs button and flop comes AK6r. GTO tells you to bet, so you bet. It even tells you the exact sizing to use. Apply this to all spots and now you don’t even have to think. All the spots are solved like Tic Tac Toe.
 
As a player who studies with solvers and runs a GTO study group, I can confirm that studying GTO is important, but it is at the same time misunderstood and overrated. In this article I will try to explain how to study and apply GTO.
 

There are two main problems with GTO

  1. You can’t really learn it and apply to all situations, since the combinations, pot sizes, number of people is way too much. The human mind can memorize all the correct moves in Tic Tac Toe but cannot possibly learn even 10% of all GTO spots in 6+.
  2. GTO means a balanced, unexploitable play. It’s not the most profitable strategy. You can’t win versus GTO. But if you are playing at a 6-max table with four fish and one GTO bot, you will have a higher win rate vs fish, if you play exploitative.
 
Let me give an easy example. You have a player 100a deep open shoving every single hand. GTO doesn’t know that this guy is drunk and open shoves everything, so GTO charts tell us to call AQo and JJ+. But since you, as a player know his range is any two cards, you can easily call KTo and it’s a very profitable call, while GTO would consider calling KTo a huge mistake.
This is an easy example, but let me give you a slightly more complicated one. Say you are playing 6-max, everyone is 100a deep. everyone folds to HJ, he limps and CO limps and you are on the button. This is what we are supposed to do preflop.
 

GTO Charts (click here to read more about the GTO charts):

As you can see, we are never raising NON all-in with QJs, KQs. We jam KT, QTs and JTs. That is because, GTO doesn’t want to ISO and then fold to an all-in with playable hands like QJs. And GTO as HJ and CO has plenty of limp/shoves. This is what HJ is jamming vs CO limp and BTN 9x ISO (in red, picture below).
 
In real life, random players don’t jam near that much. Probably only AQo and better. So the correct play here is an exploitative one, non-allin ISO your suited KQ and play a pot with a good hand in position.
 
So naturally, the correct strategy is to mix GTO and exploitative strategy. This of course you knew already, so what is the point of writing all this stuff just to get to an obvious answer. Everyone knows the best strategy is mixed and adjusted. But here we will try to understand, in which spots you should go exploitative and in which spots you should stick with GTO, despite playing versus fish.
 

Improve your win rate with GTO

 
Let’s say you want to achieve the highest evBB possible in 6+. If you only want a nicer looking graph, there are a few small ‘’cheats’’ you can do, such as only learning one strategy (50a 6-max) and only playing that, and leaving the table as soon as it’s 5-max or as soon as you have 60+ ante. This way your bbEV will be high, but this is not the best strategy for making money, because you shouldn’t leave the table if your bbEV drops from 40 to 30, since 30bbEV is still better than your 0 winrate you have if you stop playing. So, for this scenario, let’s see how to achieve the highest bbEV on any table you decide to play.
 
Number one is, you should learn as much GTO as possible, for every spot, on all boards. But you shouldn’t follow it blindly. After learning GTO, when you are in a spot and thinking ‘’should I bet or check here’’, you should remember what GTO does. But then you should think further ‘’ok, GTO bets here, I remember. But why?’’. Then you have to remember the reasoning for the GTO play.
 
For instance, on a board of T96, turn 6, UTG vs BTN, GTO usually doesn’t double barrel as UTG when called by BTN. That is because BTN calls all 6x on the flop, and BTN has all 6x hands, while UTG doesn’t have almost any 6. So this 6 on the turn helps BTNs range. So in this spot you should remember that GTO checks, because that 6 helps BTN more.
 
Then the last step is to think ‘’Ok, I know what GTO does and why. Does this apply to this current hand?’’. So let’s say in this T966 turn scenario we have KQ, and we know that GTO tells us to check (I haven’t actually run this scenario in GTO, just using this to illustrate an example). We should analyze our opponent. Is he a regular who has studied GTO a lot and pretty much plays as he should? Then check, sure. But let’s say we are playing a nitty player that we know well. We know that he would fold J6, Q6, K6 and A6 on the flop because he doesn’t consider those hands good. And he will fold his draws (J8, etc.) on the turn, because he doesn’t want to chase draws on a paired board. Versus this player, we should realize that GTO analysis doesn’t really apply, because he doesn’t play close to GTO. That’s why checking KQ here and saying ‘’it’s literally GTO bro’’ is bad. We are losing a lot of EV when we decide not to double barrel vs this player.
 
So in each hand, you should go over these steps
 
  1. What does GTO do here?
  2. Why?
  3. Does this apply here?
 
There is a reason why regulars have 30-40a winrates vs fish, while GTO has 20aEV vs fish. It is because we are able to pick up more information than GTO – is this player ever capable of bluffing?, is he angry?, is he drunk?, timing tells etc.. Information that computers don’t know. The best regs know when to use all weapons.
 
 

Leave a Reply