Leveling & GTO Poker Strategy

The most interesting poker is played when two players are deeply thinking about their opponent’s range, what they are likely to do, what they likely think the other player has, and how they can craft a play based upon that. This idea of “leveling” in poker gives strategy extra depth, and is a window into understanding GTO poker as well.

While leveling can be used to derive incredibly profitable exploitative plays, it can also help us understand the baseline of GTO strategy. Push play to see how leveling, exploitative, and GTO play all meld together.

What Is Leveling In Poker?

Many players consider this aspect of the game “real poker.” Leveling is what creates the cool plays we see on streams, and is very usable in every poker game. Leveling in poker is about which plane of thought process a player is on. 

The Levels Of Poker Strategy

Here are the basic levels in poker:

• Level 0: I am not thinking
• Level 1: What do I have?
• Level 2: What does my opponent have?
• Level 3: What does my opponent think I have?
• Level 4: What does my opponent think that I think that they have?
• Level 5: What does my opponent think that I think that they think that I have?

Most beginning players are somewhere on level 0 or level 1. They are unable to think about their opponent’s range or what their opponent’s actions mean. These are very easy players to play against. We can play totally straightforward against them and use our ABC poker strategy. Our play gets interesting when we start playing against players on higher levels.

Winning The Leveling Game

One of the best ways to win the leveling game is to remain one step above our opponent. Being too many levels above our opponent will have us making plays that are unnecessary against a particular villain (which creates FPS, or fancy play syndrome). If we are playing against a calling station on level 1, we just need to be on level 2. We shouldn’t run complex bluffs on them because they will not understand what we are representing and we will just be burning money.

As we move up and grind into higher levels, we will eventually start running into players that can and will think on higher levels. There are two major ways to combat them: 

• Think one level over them: This can be very tricky, especially against players that are playing on level 4 and 5. It will be hard to figure out when they are adjusting and will make our life quite difficult. It will oftentimes result in having to risk a lot of money in more uncertain situations (as that opponent would be able to make more correct plays against us).
• Think two levels under them: If we can figure out which level they are playing on, playing two levels under them is relatively simple. It actually brings us back to playing more ABC poker, which exposes less of our stack and puts us in fewer situations where we could make really expensive and bad mistakes. Now if we are just one level under our opponent, they will crush us. But if we can constantly stay two levels underneath them, we can make some serious money with relative ease.

At this point, we can really see how balancing and leveling go hand in hand. Take a preflop situation where a player steals from the CO and it folds to us in the BB.

Preflop Leveling

If our opponent is on level 3 and thinks we can think that high, he probably thinks “OK, this player understands that I am stealing. I thus expect him to resteal me more liberally, which means I am prepared to increase my 4-bet % to combat that.” So because he thinks we are balanced to an extent in our restealing frequency, we can actually play more straightforwardly. 

We could actually make it so that our resteal range is purely for value (only hands we would stack off preflop with). Or, we could try to jump up a level and resteal/5b shove our entire poker range thinking that his 4b range is wide enough.

Which option is easier and safer? The option that has us only 3-betting for value? Or the option that has us risking our entire stack with only a hazy idea as to how correct it is?

What if after time we figured out that our opponent was adjusting and starting to fold to most of our 3-bets? This means we could use that information to start 3-betting more of our range until he adjusts. Then we just need to figure out if he is adjusting by stealing less, steal/4-betting more, steal/defending more, or not adjusting at all. All of these things are very easy to see, and they are things we should be taking notes on.

While this kind of dynamic doesn’t happen often, it can put us a helpful step ahead of our opponent. A default way to handle this situation is to play tighter at first, take the information we have, take notes on how they respond when we use our strong range (is he folding? 4-bet/folding? 4-bet/stacking? Etc.), and then adjust to how they are responding.

Leveling And GTO Poker

In my new book GTO Gems, we explore the concept of mixing strong hands as checks postflop. It’s easy to look at solver data, especially in simplified models, and assume that strong hands should always be bet along with a certain ratio of bluffs. 

However, applying simple leveling steps can help us see what actual balanced play is.

Postflop Leveling

Let’s look at an example where we are on the river in a perfect polarization setup. The big difference in this model is that the defender is going to have some additional air combos (whiffed draws) on a board of K♦ T♦ 6♠ J♥ 6♦.

In an earlier model, the solver suggested the aggressor should fire 100% of their value combos. But as the defender’s range gets updated with these additional air combos, it’s logical the aggressor’s range (both value hands and GTO bluffs) should get updated in kind. Here’s an example of how the levels can step between each player’s strategy.

Step 1: The defender begins to realize that the aggressor is always folding after checking the river. The defender then exploits the aggressor by always bluffing the river with busted draws, and checking behind with their bluff-catchers.

Step 2: The aggressor begins to realize that the defender is bluffing a huge amount when checked to on the river. The aggressor calculates that they can now generate a higher expected value by checking value hands on the river and inducing a bluff.

Step 3: The defender suspects that the aggressor is simply not folding anymore after checking. So the defender starts reducing their bluffs. 

Step 4: The aggressor realizes that the defender is not bluffing as often when checked to, and that checking value hands no longer maximizes EV. The aggressor shifts most value hands back into the betting range. However, since the defender has shown they are capable of exploiting an undefended checking range, the aggressor continues to check a percentage of value hands. It will no longer be +EV to try and push the exploit too far, since there will be a breaking point where the defender realizes they are being exploited and lowers the EV of the aggressor by countering.

Step 5: The defender does in fact realize that the aggressor is once again folding too much after checking (although not folding quite as often as before). The defender starts to bluff the river when checked (but not quite as aggressively as before). The defender knows that the aggressor is capable of countering if the bluff-to-value ratio gets too far out of line, so the defender does their best to subtly exploit the aggressor without making it too obvious.

Step 6: The aggressor is smart enough to notice that the defender is bluffing slightly more often than correct when checked to on the river. The aggressor then subtly shifts a few more value hands into their checking range. Perhaps the defender will not notice and will continue to over-bluff when checked to.

Step 7: The defender picks up on the fact that the aggressor is not folding quite often enough for bluffing to be profitable. The defender subtly starts bluffing less often than is correct, but does not push things as far as they did previously.

Step 8: The aggressor notices the small adjustment in the defender’s bluff to value ratio and shifts a minuscule amount of value combos back to the betting range where their EV will be highest. 

Step 9: The defender realizes that the aggressor is now over-folding the river by a minuscule amount after checking. 

Notice how the two players iterate against each other, using smaller and smaller exploitative responses at each step. We might visualize this as displayed in the following diagram:

If the diagram represents smaller and smaller exploitative adjustments, what is represented by the red dot at the very center?

This is the point where there are no additional exploitative opportunities available for either player. Each player is completely balanced and cannot increase their win-rate by deviating from their current strategy.

In game theory lingo, this is referred to as a Nash Equilibrium. This term describes the game state where no player can increase their win-rate by deviating. In poker terms, a Nash equilibrium is achieved when two players are playing perfect GTO poker against each other.

Leveling & Mixing

Note that the aggressor has no choice but to mix strong hands into their river checking range. If the aggressor fails to do so, the defender has an exploitative bluffing opportunity.

However, the aggressor cannot place all of their value hands into the checking range because:

• The aggressor will miss out on the additional EV that comes from value betting.
• The defender can stop bluffing as an exploit.

The only way to maximize the EV of every hand is for the aggressor to have a perfect mix of bets and checks with their strong holdings. Without this perfect mix, the defender can generate an exploitative counter-strategy, which in turn lowers the EV of the aggressor’s hands.

Thus, we see that mixing is a crucial part of maximizing EV.

Balancing Against Fishy Players

If our opponent is not capable of finding exploitative counter-strategies, mixed strategies lose much of their value. If the EV of betting and checking is the same, betting and checking in any proportion will result in the exact same win-rate against a non-adjusting opponent.

However, we often find the EV of various actions is less likely to be the same against opponents who fail to adjust. Exploitative poker often makes heavy use of pure strategies as a result and does not make use of mixed strategies as often as a GTO approach.

Against very good players, we might not be able to get away with overly-obvious pure strategy exploits. The two players in our above model tried to exploit as hard as possible while still flying below their opponent’s radar.

Trying to exploit our opponent too hard will actually lower our EV if they are capable of quickly deploying the relevant counter-strategy.

Don’t Level Yourself

You will never play against a perfectly GTO human opponent. In fact, most opponents you will play against will cap out at level 2. 

Understand that exploitative opportunities are the true winrate creators, and avoid trying to perfectly emulate solver output against terrible opponents. If you know a terrible opponent is highly likely to zig, then make sure you zag. Don’t zig as well just because you think a solver would approve.

This is why we wrote GTO Gems. The point isn’t to get bogged down into solver minutia, but rather to get the big-picture takeaways from a GTO strategy and implement the good ideas while ignoring the intricate ideas that create more confusion than good. 

If that sounds up your alley, grab your copy from Amazon today and leverage our years of solver exploration into actionable ideas you can begin using ASAP.