GTO Poker Strategic Foundations: Episode 1

The most important thing you should consider while playing poker is maximizing the expected value (EV) of your decisions. The reason for doing so is simple; the bigger the EV of your play, the more chips you'll win in the long run.

Luckily, most poker tools, including solvers, take expected value into account. In fact, whenever you use Deepsolver, it uses the data you input to find a strategy with the biggest possible expected value under given circumstances.

That's a great feature, but it also has a drawback - quite often, the strategies with the highest possible EV are complex and very hard to incorporate by humans. Fear not, though; we've got a remedy for that!

But first, if you've never used Deepsolver or any similar poker software, look at our tutorial: GTO Poker From Scratch.

How can you create a simple GTO poker strategy without losing any EV?

Let's look at the example of a button vs a big blind hand. We'll use cash game ranges for 100 big blind stack depth. The board will be dry: king, seven, deuce rainbow. Let's compare the EV of the strategies: one with five sizings available for the in-position player and the second with only one sizing available. Let's force out-of-position player to check 100% of their range for simplicity.

When the OOP player always checks, there is no need to use a complex strategy based on various sizings. In this scenario, if you are the in-position player, you can use only one sizing - 33% and your strategy will have the same EV (3.4) as the game theory optimal strategy with five available sizings.

While at the poker table, it's usually not the best idea to use complex and mixed strategies unless there is a clear incentive to do so. It's great to explore the different outputs of the solver for study purposes, but you should always focus on creating a solid, easy-to-replicate baseline strategy first.

Simplifying GTO poker strategy in practice

So, how can you transform a strategy with multiple sizings into a much simpler one? Firstly, you can use the "simplify" option when solving a hand. After setting the preferred threshold (we suggest choosing 5 to 10%) and simplifying the sim, you'll end up with the most prevalent options available.

Afterwards, you can re-run the sim with the sizing that you are the most comfortable with or with the one which is the average of the ones used before. Your results will almost always have very similar EV, yet they will be much simpler to implement at your average poker game.

Categorizing the boards is something every poker player should do

With time you'll realize that different boards have the same common properties, and you can categorize them accordingly. Let's look at the example below:

All the boards above have two broadway cards (hence double broadway), do not contain flush draw (rainbow), are unpaired, and as a result, the preferred bet sizing for the in-position player is 33% of the pot. Analyzing different boards according to the same characteristics will allow you to understand what your baseline strategy for a particular type of board should look like.

Also, remember that Deepsolver provides you with a game theory optimal ranges for every position and available stack size, but you must consider how realistic they are. Most poker players you'll play against will be far from the optimum, so you should adjust the ranges to reflect the poker games you play every day.

Creating easy-to-replicate turn patterns

All turn cards can be categorized into one of the four types: straight completing, flush completing, pairing the board or blanks. Returning to our initial example, let's evaluate how we should proceed on the turn when it completes the possible straight.

We can clearly see that when given two options (75% and 150%) solver bets the straight completing turns (A, Q, J and 9) quite aggressively. We can also simplify our strategy by choosing the in-between sizing (115%) with which our strategy would look like this:

The great thing about this option is that you can toy with different sizings and assumptions until you find the one that suits your needs while maintaining an EV close to the base strategy.

Knowing how categorizing and simplifying turn textures work, you can look at other turn cards. Remember, the value threshold always dictates the turn strategy. Let's look at how it will look for a blank, like four of spades.

Here you can see that KJ or better always bet for value, while the weaker top pairs mostly check. Of course, the more value bets we have, the more we can bluff. How do our bluffs look on this occasion?

Most of our turn bluffs have some future equity, like straights - and if we are polarized enough, we also bet our best blockers and a typical example of this is J4.

This combo has a top pair blocker, turned second pair blocker and very little showdown value. That's why this hand is perfect to bet turn with and bluff on certain rivers. We usually try to fold out better hands, so make sure you are betting polarized enough.

GTO poker on the river

Since the river is the last street of the hand, our ranges have become more defined. But we can use the same assumptions for different rivers as for turns. Let's look at the queen of hearts, a straight-completing card.

The IP player has some all-ins, overbets and medium size. Our straights like to jam, and we use our Jx straight blockers as main bluffs to even out our value-to-bluff ratio. We can also see that we are jamming pocket fours for value, thus using other 4x combos to bluff with.

If we compare this to the eight of diamonds, you can see that the previous combos mostly give up, and we are now bluffing our low missed draws. Since eight is a blank, our value region from turn doesn't change much, and we don't need to block as much. So, bluffing our lowest cards will have more efficiency as they unblock more folds.

To highlight everything, we can look at the five of spades, and here we can see some adjustments to the strategy. Since both players have flushes in their range, some of our hands get devalued. But we can still valuebet our sets for a smaller size, and most of our bluffs come from the offsuit region, blocking the flush.

A simple poker strategy executed well is more efficient than a complex one executed poorly

To summarize, simplify your strategy where it makes sense. Remember to group similar board textures and strategies for efficiency and always think about how your opponent's range can look like.

That being said, if you want to succeed in today's online poker environment, you must study advanced poker strategy. While we believe Deepsolver is the best poker tool to do so, we've also covered how other poker software work.

Do you have questions about Deepsolver or want to play GTO poker better? Join our Discord, where you'll find our helpful staff and many poker players striving to improve daily.