Andy Stacks is the only regular who plays an approximation of that "GTO robot" style. Garrett makes some killer reads, but his opening ranges are old-school tight. I can't be sure, but almost everybody seems to be trying to put people on hands instead of ranges.
Just to be clear (and I’m sure you know all this, but for the general record) ... Just thinking about ranges isn’t in itself a GTO strategy. Ranges also are used in purely exploitative play.
I’m hardly an expert, just an interested student of the game. But as I understand GTO, at the most basic level, ranges are merely a starting point for GTO decisionmaking.
By guesstimating opponents’ ranges from various positions, and knowing your own, you proceed to estimates of what frequencies and sizes are optimal for checking, leading, calling, check-raising, 3betting, etc. in various situations. And it’s not just how often someone takes various actions from those positions, but how big or small the bets and pots are, which inform those guesstimates.
The bigger the sample size of hands with various villains, and the more attention you pay to their habits, the better the guesses. But that’s not GTO, it’s just sound poker... which may be enhanced by an understanding of GTO.
And given that there are trillions of situations, and multiway NLHE is far from solved, no human can ever really play true GTO.
Anyway, point being that the very same range estimates (which again involve a lot of imprecise guesswork, unless you have a giant sample size of hands) can be used for non-GTO play. In fact, I don’t know any serious poker theorists or authors who actually say you should try to play pure GTO all the time.
After all, GTO at its purest is a strategy for ensuring that you can’t lose, long-term. If two players both achieve perfect GTO, it would eventually become a stalemate once they played enough hands to eliminate variance. If one person is playing perfect GTO, and the other is almost-perfect, the perfect player would win—but it would be a very tiny win rate, depending on the gap in their level of perfection.
So pure GTO as practiced by humans unaided by computers is not a strategy for maximizing profit; it’s a strategy for being less exploitable. The main way this seems to have crept into real-world poker is in players attempting to “balance” their ranges and decisions. A simplistic example would be using similar sizing for a huge river bluff as you would use when you have the nuts. But also only using that sizing a certain percentage of them time, according to what you think is correct for the situation.
The best authors, like Andrew Brokos (who has two books on the topic) seem to advocate trying to understand GTO so that you can see the flaws in your opponents' games more clearly, and then adjusting away from GTO in your own play to exploit their errors and make bigger, more consistent profits.
For example, if you have noticed that someone limps far too often from early and middle positions, you might “punish” their wide range out of position by raising their early limps more often than you might otherwise when acting after them. Or limp behind in late position wider than you otherwise might. Or devise some other strategy to exploit a flaw in the game.
Or, if you notice that someone raises their button nearly 100% of the time, you might defend or reraise from the blinds at a wider clip. And then further define and narrow their ranges based on how they then behave in the rest of the hand.
Postflop, you may use ranges to come up with general estimates of how “hard” various textures hit their likely range, based on their preflop action from various positions, and then try to exploit what their perceived ranges are according to the situation.
But once again, that is not GTO per se, just basic thinking poker strategy. Just approaching hands that way does not make me or you or Andy anything like a GTO player—again, at least as I understand it as an amateur.
