GTO - game theory optimal - poker solvers are all the rage right now. They have transformed how the game is played at the very highest level, and now that effect is trickling down to the rest of us mere mortals.
Game theory optimal is a term that has been thrown around for over a decade now. Before it was simply balance, then people were trying to construct GTO ranges and strategies, and then solvers appeared on the market.
But how useful are they really in 2021? While the world’s elite players have no doubt mastered the use of solvers, and memorised the results of much of what they learned, are they important to the average player? The answer has to be no.
Game Theory Optimal Versus Exploitative PlayThere are only two approaches to how we play poker - balanced and exploitative. If we play against a perfect opponent then it stands to reason that we must play a GTO/balanced game plan against them.
However, unless we are sitting in Bobby’s Room at the Bellagio in Las Vegas surrounded by the likes of Dan “Jungleman” Cates, Phil Ivey, and Gus Hansen, then we are going to be up against players who make plenty of mistakes.
Game Theory Optimal Poker SolversThe solvers we see on the market now allow us to input a range for both us and our opponents. We can also choose what bet sizes the solver should consider on each street.
Once the program has this data it will compute a strategy that is unexploitable by any opponent. One observation that we will immediately notice is that a mixed strategy is almost always thrown up.
This means that we must take one action with one combo a certain percentage of the time, and the rest of the time we take a different action. This effect is to do with balancing but at the end of the day it costs us profit.
Why Solvers Might Not Be the Best Use of Our TimeAs improving players we must make the best use of our time to improve our skills. This will almost certainly be breaking down hands and confirming that the line was as our default strategy dictates. Niches such as making sure we are c betting correctly, as well as probe betting and floating betting.
Getting to grips with this will be far more productive than ploughing through what a solver will tell us when in all likelihood we won’t even have the time to properly memorise the results.
The number of hours required to make proper use of a solver is insane. High stakes elite players have such a high standard of fundamentals already that for them this is just about the only way left that they have to improve.
Will Poker Be Solved Soon?When the use of solvers became commonplace people were posting on forums across the internet that it was the beginning of the end. They opined that there would soon be a perfect strategy posted online for all to use and we would all be left losing to the rake.
Of course, this has turned out to be completely ridiculous. Just as with the game of chess, NL Hold’em has too many permutations to calculate using a modern home computer, and even an industrial use supercomputer.
We just don’t have the computing power today to come to a concrete conclusion for the complete game tree.
Even if we did have such technology, for one, can you imagine trying to memorise such an amount of data. And secondly, it’s not practical to make print outs for use in game either. This leaves us with the bot problem, which while still a nuisance is not yet known to be able to use solvers in real time. Or store such an amount of information to retrieve in quick time.
The conclusion for how game theory optimal solvers deal affect poker in 2021 is that they are helping to perfect strategies for the very best players, but aren’t much to worry about for the average player.