I knew when I took on the task of creating the Advanced Poker Training GTO-Style bots that this was going to be a unique and collaborative process. Our team was thrilled with how the bots turned out for our recent Beta release. They play tough and the advisors give you insight into how GTO-style decision-making works.
But as always, our APT community has provided invaluable feedback to make these bots even better. Once we got this feedback, I sat down to refine and update the bots. In a few weeks, we will unveil the new and improved updated bots. So you’ll need to be a bit patient, but that patience will be rewarded.
OK, the bad news is that the bots are going to be even HARDER to beat. But on the positive side, the advice you get from the advisor bot will be that much better.
Changes to the APT GTO-Style Bots
We tweaked the bots in a number of directions, but here are four notable changes. There are others we aren’t even going to tell you about (bwahahahahaha!)
- When the bots open-raise pre-flop and get 3-bet (or there’s a call and then a squeeze), the bots will be more sensitive to the position of the player that makes the 3-bet. Example: If you raise UTG, getting 3-bet by the UTG+1 player is much strong than getting 3-bet from the button.
- Similarly, when facing a 3-bet cold, the bots will be more sensitive to the position of the opener AND the 3-betting player. Example: If you’re on the button, facing a raise from UTG and a 3-bet from UTG+1, you’re facing a stronger range than when facing a raise from UTG and a 3-bet from the cutoff.
- On the flop, the bots will do a better job determining which flops are good/bad for their range. They will adjust their betting/calling ranges accordingly. As the aggressor, they’ll also be using different bet sizes in line with what solvers recommend. For example, using a small bet size with virtually their entire range on certain favorable flops.
- Similarly, on the turn and river, they will be better at determining which turn/river cards are good/bad for their range, and will act accordingly.
As you can see, we’re teaching the bots to adjust to even finer situational nuance. As we have iterated the bots, they are getting closer to wedding GTO theory with how players actually execute that theory.
Looking Ahead
Playing GTO-style poker is hard. Programming bots to follow GTO parameters has been equally challenging. But the learning potential for APT members has been well worth the time. If you can develop some of the skills that a GTO approach dictates, you will become a very dangerous player. Even in lower-stakes games, where exploitative play generally reigns, being able to mix elements of GTO play will make you a more complete opponent.
Note that you don’t need to dive in the deep end right away and play the pro-level GTO bots. If you start with our lower-skill-level games, you can get exposed our GTO advisors, and watch as other bots don’t always execute GTO strategy well. Work your way up in difficulty level as you improve your ability to execute and react to GTO play. As the APT GTO-Style bots evolve, so will your understanding of GTO play!
