Today we’d like to share key insights about working with GTO software from one of the most experienced online poker players, our longtime business partner and friend - Mikhail Shalamov. Mikhail is currently a PokerStars PRO Online player, a regular of 6-max zoom cash games at NL500 limit, and one of the most popular twitch poker streamers. At PokerStars Mikhail plays under the nickname "innerpsy", and he shared some of the key strategies he uses when studying with GTO software below (note we’ve translated his advice into English).
Basic principles of working with GTO software.
Many players try to avoid putting in the time to study theory. There is a false impression that since you can't directly earn money while studying or practicing, that study time is some kind of useless and rather boring endeavor. In reality, probably 95+% of poker players will gain much more overall EV if they increase the time they spend working on their game away from the table rather than grinding away a higher volume of hands.
But, to actually gain EV from your efforts, you need to study efficiently. Deconstructing a 400bb deep hand where you had to fold the second nuts is certainly fun, but it will have the same effect on winrate as rubbing a lucky rabbit's foot during all-ins. If instead, you focus on improving your decision making in common and important situations, like adjusting your c-bet and check frequencies in single-raise pots, that is a very different matter.
Today I would like to share my experience of working in GTO trainers, which have long been the main tool for fine-tuning the game for me. They bring efficiency to an otherwise boring educational process, while being as close as possible to a real game. Below I will try to deduce a set of rules that will help you make the most of your sessions against a GTO-machine.
1. Systematic approach.
Try not to mix different spots in one training session. For each study session, choose a spot based on how often you had problems at the tables in that situation, how long ago you trained it, and how many mistakes you made the last time you practiced that spot. You should not train exclusively on the button vs bb in a single raise pot, although this will be one of the most common spots, do not forget to play in early positions and tight ranges.
2. Relevance of results.
It is important to understand which results will be most applicable and which ones are best left out. The more assumptions that accumulate in the players' ranges, the more strange solutions will be obtained. For this reason, I almost never train 4bet pots because the opponents' ranges and postflop play are too different from what the GTO Solver shows, so I prefer to analyze such spots in Simple Postflop in more detail with non-GTO ranges. I also devote less focus to studying the river - usually a significant deviation from what will be in real play will accumulate by the river in the GTO ranges, and the results will often be surprising and not particularly applicable. However, it is always helpful to carefully understand why a GTO strategy plays the river as it does, even if you might elect to play differently at the tables. From here, by the way, one of the main points comes out:
3. Consider each click.
Do not click buttons as if you are racing to supernova elite status against a GTO simulator. Before each action - for example, a continuation bet or a check as a preflop raiser, try to mentally talk with what frequency and what size do you think the cbet with your hand is, and also suggest why you’re choosing a given action. For example, you might say to yourself, "a low suited board is more suitable for the big blind range, which means our hand versus range equity will not be high, but nevertheless our hand should be bet at least 40% of the time due to good backdoors ".
After making a decision, be sure to check your assumptions with the real result, adjusting our thoughts before clicking accordingly, eg. "damn, in fact, a check is always played. Apparently, the vulnerability to check-raises on such a board does not allow us to effectively bet with this hand, and such bluffs are carried over on the turn or river when we have a better idea of the equity of our hand and the equity of our opponent's range."
4. Frequencies!
Poker now is a game of correct frequencies. Starting from preflop, almost any action can theoretically be optimal if done with the required frequency. Postflop frequencies are also important. Therefore, it is very important to teach yourself how to work with the randomizer. At the tables, I always show a random number from 0 to 100, just like in the Simple GTO trainer for the desktop. For the GTO Sensei app, a randomizer will also be implemented soon, so in this regard the working environment will be identical.
Determine for yourself how the scale will work (I have 0 - the most passive action, 100 - the most aggressive). Before each decision, when making a suggestion about the frequencies of the game with our hand, try to predict what frequencies will be optimal here, eg. "we 3bet pre, the board came A52r. I think that a c-bet with AJ will be placed here in about half of the cases. When using strategy based on two c-bet sizes, we should significantly favor a smaller size so if the randomizer is less than 50, I check. The remaining 50 I would split between 1/3 of the pot 40% of the time and 2/3 of 10% , so at 50-90 we put a third on the randomizer, from 90 to 100 - 2/3 of pot." Similarly, choosing our action, we need compare our choice with the GTO action and try to draw certain conclusions and adjustments in thinking, eg. "Damn, AJ is bet almost always and often for a larger size, apparently weaker aces are more suitable for protecting checks, and AJ is better played aggressively as a value hand."
5. Pay attention to your opponent.
The double efficiency of working with the GTO Trainers is achieved by the fact that we can check not only our own, but also the opponent's line (on the desktop version of GTO Trainer you can see the opponent frequencies as well as the chosen action, I hope they will also add it soon in the GTO Sensei). After the hand is played, during our analysis, we should also pay attention to the opponent's line and try to do all the reasoning for him, too, after the fact.
6. Detailed analysis.
Sometimes it is not easy to understand the logic of GTO Strategy, in such cases, do not be lazy, and instead take the time to open a spot in a Simple Postflop and see how the entire range is played, and not just the specific hand. This will make it easier to make assumptions about the reasons for certain actions and develop a mechanism for making the right decisions in such situations in real time at the tables.
7. Develop heuristics
The goal of the above steps is to allow you to understand the general principals, and not to focus on remembering the irrelevant specifics. Don't remember that AA checks 90% on a 568cc board in a single raise pot bu vs bb. It is necessary to remember the conclusion (overpairs that do not require protection from overcards are bet on the flop on dynamic small boards quite rarely, the lower the overpair, the more often the bet is played). This will make it easier to make decisions in the game based on the lessons that you learned when working in the GTO trainer.
This article was originally published in russian language (we tried to translate it as accurately as possible) Mikhail's telegram channel, where you can also find a lot of fun, useful and entertaining poker content. If you liked the article, we recommend that you subscribe to all of Misha Inner's social networks, since the content of each social network is often unique and not duplicated: