Do you need to be GTO? I hear this question often from my students. If you read the first article in the series, then you already understand the difference between exploitive (sword) and unexploitable (shield) solutions.
If you haven’t read that article yet, you probably should, but the short version is that attacking your opponent’s range and tendencies is your sword. It will do the most damage, but also leave you open to damage yourself if your opponent is better than you think.
Unexploitable solutions are your shield. They may not do as much damage, but you are always safe when using them. You can still lose a hand, and often will, but in the long run, you can’t be outplayed.
What can GTO do for me?
The first problem is that many people have this idea that they have to become a robot to use game theory in their play, or that game theory requires lots of complicated math and logic solutions to play a simple hand. That just isn’t true. On the other hand, understanding what solvers and game theory solutions tell us about the game can profoundly influence how you see the game and change how you play even the most basic hands.
The flop continuation bet is a good example. Just a few years ago, most pros were making continuation bets more than 90% of the time. Solvers have shown us that a lower C-bet frequency, more like 60%, can work really well.
The reasons for this change in C-bet frequency would need their own article, or maybe even a book, but the short version is that if we C-bet all the time, we are exploitable and, if we play well post-flop, we don’t need to rely on constant continuation bets. A good player can bet based on the flop and their opponent’s range, delay their bets until the turn, or even check all the way through and induce bluffs more often.
This style of play gives us more options, more tools, and less predictability while also getting us closer to being “solver correct.”
Knowing that good players aren’t C-betting as often as they used to will help us with our own game and help us recognize when other players have done their homework. A player who’s been cashing regularly for years, but still bets the flop 90% of the time when they’re the preflop raiser, is definitely behind the times and we know a bit more about how they play.
We also know to put a ton of heat on their C-bets if we have any kind of reasonable draw because solvers have taught us that this is profitable.
Bet sizing is important too
It’s not just the frequency of our C-bets that solvers have changed. It’s also the size. After working with a solver just for a day or two, you’ll find that GTO play often relies on much smaller bets than most players make. It’s not uncommon for a solver to recommend a bet of 20% of the pot, or even less, on many boards. On a board that is all the same suit, a bet of as little as 10% can be your best option.
When I discovered this a few years ago, I gave it a shot. My experience with new strategies is that it takes time to get used to them. A new play puts you in spots you aren’t familiar with, and it often takes time to get comfortable playing in different situations.
With smaller C-bet sizing, I was comfortable within a few days, and now the smaller sizing is completely standard for me and it has increased my profit.
Learning about smaller C-bet sizing not only helped make me more money directly, it also helped me learn to think about the game differently. I found myself playing deeper into hands, seeing more turns and rivers, and learning how to take advantage of my opponent’s mistakes on later streets. It put me in positions to make more money and made me think about how to get into more of those spots.
While I certainly have memorized specific spots and learned how solvers play in the most common situations, I didn’t need to do that to learn a great deal from working with a solver. Just making smaller C-bets opened up a whole new facet of the game for me. And this is just one small example of how understanding game theory optimal plays was helpful.
GTO is a great defense
Understanding how solvers tend to play also helps you get better at guessing how a solver might play a hand, which gives you a nice shield to hide behind when you’re playing against a great player and don’t want to get slaughtered. When playing with really strong players who I know have spent time with a solver, I try to picture how a solver would play the hand and do my best to emulate that. If the greatest players have an edge on me, it’s not by much, and I can still pull out my sword and attack the weaker players instead of playing like a solver all the time.
I could go on at length about how it is valuable to work with a solver and think about why it’s recommending certain plays. If you don’t understand some of what solvers are doing, and why, the game is passing you by.
So the first answer to “Should I be GTO?” is yes, sort of. You should definitely learn about it, understand it, and take what you can from working with a solver or a database of pre-solved hands. But there is a second part of the answer.
Solvers don’t know everything about poker
It’s a misconception that we must either be a GTO player or an exploitive player. Sometimes the answer to “Should I be GTO?” is a resounding no. There are lots of reasons why just playing like a solver all the time isn’t the most profitable way to play.
First, if the game is so tough that you need to be a robot, churning out solver solutions just to win, then you should be finding a softer game. If everyone plays perfect GTO all night, the only winner is the house, and everyone at the table is paying rake and wasting their time.
Second, there are many times when an exploitative solution is much more profitable. If your opponent shows you an opening, and you know you can hit them hard, then use your sword and end the fight rather than using your shield. Eventually, solvers will be generating exploitive solutions based on all available data, but that’s a long way out, especially for brick-and-mortar poker.
Solvers also don’t know what your ROI is in a tournament or how badly your opponent might play on a later street, or if they have a tell that you might spot when the river card falls. They don’t know how much a tournament’s payout structure affects the correct play with hundreds of people left.
A solver also doesn’t know that you might go on tilt if you make a certain play and it goes badly, costing you a lot of money as you play badly for the next hour. And it doesn’t know that your opponent might go on tilt as well if a play works out the right way. It’s well worth sacrificing a tiny bit of equity on the chance that you can send an opponent steaming off for three more buy-ins in a cash game.
A solver doesn’t think about the meta-game either. If making a play that’s clearly not solver-approved, like betting 70% of the pot on the flop, is close in equity to the solver-correct play — and there are good players at my table who will recognize that I’m not making a solver-correct play — then it may be well worth it to convince them that I don’t play a GTO-based game so they’ll underestimate me in the future.
GTO is a great foundation
There are many more reasons why you shouldn’t just play like a solver all the time, even if you could. But GTO is an excellent foundation. If you start with a solid GTO-based game and alter it when you find a good reason to do so, you’ll be a strong player. In fact, this is exactly what most of the best cash game players in the world are doing.
I find that running solutions by hand can be a big help in finding this balance. I use a spreadsheet that I built to calculate the value of different plays, an equity calculator (I like Equilab myself), and an ICM calculator if it is a final table or sit-and-go hand.
Analyzing the hand myself really helps me think deeply about what’s happening. What is my opponent’s range? What do they think my range is? How will they respond to different plays? How much equity will I have if we get all the chips in? How often will they fold? How will my flop play affect my choices on the turn and river?
Answering these questions, and many more, away from the table helps me think about them more clearly when I’m back at the table and in a similar situation.
Once I’ve analyzed a hand in this way, I often go to a solver and see what the GTO solution is. It helps me understand even more about how the hand works and, sometimes, gives me insight into new ways to play. But, the solver rarely changes my decision about how the hand should be played unless I’m up against a truly world-class player where I need to be GTO to avoid being exploited.
So, the answer to whether you need to be GTO is both yes and no. I think that’s fitting since the answer to almost every poker strategy question is “it depends.”
Can I learn to play perfect GTO and rule the poker world?
Good luck. Memorizing every solution is impossible for a human as we just don’t have enough memory. The best players could probably give you an approximate GTO solution around 80% of the time, and these are brilliant people who have spent years studying the game.
Can I learn to beat GTO players?
Technically, no. If someone is playing perfect GTO strategy, then the best you can do is to play the same perfect GTO strategy back at them and break even. Even then, in most games, you’ll lose since there’s rake in cash games and a house fee in tournaments. But, no one plays perfect GTO so far — and no one will until we have computers implanted in our heads — so you can certainly learn to beat strong GTO players, but only by becoming a strong player yourself.
This sounds hard and I don’t know where to start, should I just give up?
No, of course not. When you first started learning about poker, it was a complex game that you knew nothing about. Remember how exciting it was to learn how to play and figure out new strategies? Embrace that feeling and treat GTO study the same way. It’s a whole new world of strategy and a whole new set of tools you can learn to use to beat your opponent. It’s not memorization if you’re doing it right — it’s learning new strategies.
Would a solver beat the best players in the world?
Yes. A strong poker AI has already beaten the best players consistently, and an actual solver that was given time to solve each hand before it made its decision would beat any human.
Can I just use a solver to play online?
While there is definitive evidence that this is happening online, it’s not a good idea. Not only is it against the rules and unethical, but it’s a lot of work and the site will catch you and confiscate your bankroll. This is a dangerous game. Don’t play it.
Get better at poker and win with your brain. It’s much more satisfying and a lot less risky.
In a future article in this series, I’ll be talking about the problem of solvers being used in online poker and what sites are doing to prevent it.
Where do I start learning about GTO and solver-based solutions?
There are lots of training sites and software solutions for this, with new ones popping up every day. I like Learn Pro Poker and their application Range Trainer Pro for an overall training solution, GTO Wizard as a solver database that is easy to learn and use, and the DTO app which you can use to quiz yourself and improve your game on your phone.
For the ultimate solution, get a subscription to Pio Solver, which has a steep learning curve and is fairly expensive, but worth the time and money.