Data-driven analysis and visualization

But what if you still need more insight? No problem. The 'Absolute' function gives you an even more in-depth view of the strategy for every possible hand. The space of each cell is described by the normalization feature, following the mathematical description. This allows you to make informed decisions and play your best game.

Our Simplification algorithm manipulates the Gameplan in a way that minimizes EV loss when removing an action, or maximizes EV gain when adding an action.

This algorithm goes beyond the Nash algorithm, allowing for strategies that are not necessarily balanced. When you remove an action, the hands that would have used that action will be directed to the alternative action with the second highest EV instead.

To see how it works, let's consider a Raw GTO-Strategy:Removing the raising actions is a good way to simplify, even if the R 17.4 has a frequency of 14.5%. If we remove the raising actions, the new strategy will only have a fold/- and a callpart, and we will not lose any relevant Range-EV.

However, if we remove a further action, our Range-EV will decrease significantly. On the other hand, if we go back to the Raw GTO-Strategy and directly remove the callnode, keeping the raise actions possible, we will only lose 2bb/100 with this strategy.

The Simplifier algorithm shows you how to take advantage of several options for simplifying your GTO-Gameplan.

On average, there are 6900 action nodes for a 2-Bet spot, with over 16 million possible strategies. With every scenario, there's a chance to make a mistake that could reduce your win rate. That's why it's crucial to keep your game plan as simple as possible.

We've simplified the game plan by comparing each runout with every other runout. Each runout has a specific strategy, defined by the frequency of possible moves, weighted by a parameter that represents the relative value of different lines.

For example, an all-in frequency gap is weighted higher than a frequency gap for a check.

We used a 2-dimensional clustering method to find the minimum number of different clusters, without losing Range-EV for any player. The result is that the 49 turn runouts become grouped or clustered. Each cluster has a representative card with the smallest strategy difference between all cards in the cluster.

Each player has their own Runout Clustering, which is likely to be different from the other players in the same situation. Here are three examples of different levels of simplification:

In a Cashgame 3Bet MPvsBU, the strategy for BU simplifies from 49 to 2 different strategies on turn without any EV loss. (96% simplified)

In an MTT 25bb SRP BUvsBB, the strategy for BU simplifies from 49 to 12 different strategies on the river (75% simplified)

In a Cashgame 2Bet BUvsBB, the strategy for BB simplifies from 49 to 21 strategies on turn (57% simplified)

It's interesting to note that you can have the same strategy for two different cards in a cluster, even though they have a big difference in their Runout-EVs. This shows that there can be multiple, different reasons for having a particular strategy.

We save the result of each Runout Clustering calculation on our backend, so you only have to calculate it once. Understanding the different cluster structures opens up a new door to understanding poker.

Here's how it works: like mentioned the Made-Hand Distribution classifies hands into 18 different groups based on their properties. The properties of a hand are related to the board structure and can have a big impact on your strategy. With this feature, you'll be able to see the relationship between the perfect strategy for a specific spot and the hand class, like having a flush draw or set.

With this feature, you'll be able to see the relationship between the perfect strategy for a specific spot and the hand class, like having a flush draw or set.

And it gets even better - the feature provides two different distributions: one that shows how often a hand group results in a specific action, visualized by the color of the hand class bar, and another that shows how likely it is for you to have a specific hand class in your current range.

With the Made-Hand Distribution, you'll have all the information you need to understand which type of hand class you have in a certain action, and how likely it is.

The two graphs are anti-symmetric, allowing you to quickly get a full understanding of the game at a glance.

For example, if the runout is a spade, a minimum in the graph for the in-position player might indicate that they have fewer flush draws than the out-of-position player. This information is essential for making informed decisions on the next street.

This enables you to compare strategies for different flops. It comes with two filter functions, suitedness and pairing. You can order any value from high to low or vice versa, making it easy to determine the best or worst flops for you. The average values are displayed at the top of the table.

With this feature, you can easily find out things like: on which type of flops you should use a big bet or check back most of the time, which flops connect well with your range or your opponent's range, and how the Range-EV is connected to your preferred strategy.