Suppose there are two poker bots. One is codes to play a gto poker that is way more advanced then what we know.(it doesn't adjust based on opponents though. It has mastered balance only mathematically).

The other bot has no strategy of its own but is given the first bots strategy. I.e. it knows what first bot is doing with all ranges on all boards.

If I am not mistaken, bot 2 should demolish bot 1 because while bot 1 is perfectly balanced, with this much info bot 2 can just exploit the range gaps...

By. Range gaps I mean if bot one has a particular range to bet call a particular raise size and a range to bet fold to a raise, then if we know these ranges precisely, we can choose a size that is unexploitable for US.

so atleast theoretically perfect exploit beats a perfect but static GTO.

The point of this thread is that often we see opponents playing so many medium hands both value and showdown very standard, but we are at say the highest end of our showdown value so we call.

While playing gto is the start for poker, true gto is about finding the maximum ev in the current spot against the current opponent using his tendencies.

Basically if you are able to perfectly exploit an opponent in a spot, that is the real GTO.

I think there was a post about this long back called shania but I'm not sure if it is really used to the maximum.

I think that if you are serious about improving as a player, we should open our creativity beyond the standard way to play.

We should find a way to have almost any line and any size in every spot. If we put this much effort in analyzing each spot, you discover that poker theory has barely been scratched. It is almost completely open to research and new methods of approaching the game.

Hope my rant was clear. Don't mean to offend anyone here.(it's 2018 better be safe an apologise preflop)

Google nash equilibrium. Games are solved based on each player knowing the equilibrium strategy of their opponent.

I didn't understand the rest of the post, sorry

If bot 1 really is playing GTO, bot 2 cannot destroy it, but can only break even at best.

He's saying that bot 1 is playing GTO only with his ranges and no other info but bot 2 is playing GTO with it's ranges while knowing what bot 1 is doing with its ranges. Obviously bot 2 is crushing bot 1 because of the informational advantage .

OP, I don't remember who coined or where the "shania" post is (likely in the stickies). Shania's doing things for metagame purposes.

+1 the bot is playing against itself and tries to increase its ev. Gto is achieved when it's no longer possible.

No, GTO is about finding nash equilibrium. It is a concept that did not originate from poker. Game theory and it's concept are applied to many different games and strategic problems. What you describe is perfect exploitation

I think if bot 2 can demolish bot 1, then bot 1 hasn't reached nash equilibrium in all situations, so isn't truly optimal.

Bingo.

OP, I'm going to randomly throw rock, paper, or scissors 1/3 of the time each. If you built a bot that you programmed to know my strategy could it ever exploit or beat me over a meaningful sample size?

Yes. Except for theoretically. Or in actuality.

If bot 1 is playing GTO, and bot 2 knows bot 1's strategy and plays perfectly against it, then both bots break even.

But . . . but . . . a truly exploitative bot could find the weakness in your strategy and use it to crush you . . . !

I'm saying if you are able to make some sort of read on a players tendencies or range in a spot, throw your gto out the window and exploit the s**t out of the opponent.

Till we reach atleast 200 NL or even much higher, we can do with exploitatively changing our ranges, lines and sizes because while everyone can say this is gto, unlike FL holdem, this game has not been solved.

We are worrying too much about am I high up enough in my range etc. Whereas a lot of those decisions are tremendously marginal.

If we spend more time trying to win more ev in every spot, then our game will eventually overtake the said "losses" you make in experimenting with changing styles.

Thanks for simplifying! If our math is moderately ok but our information is of high quality, we can own opponent.

Maybe we should spend more time trying to supposedly soulread opponents than just balancing.

Noone ever said that you couldn't deviate from GTO to exploit your opponnent... But you don't have accurate info about your opponnents tendencies in all spots all the time. What to do when you have no info? And francly, your villian reads and population reads can often be wrong.

Also, in order to know what a mistake looks like, you have to first know how a perfect strategy looks like. How do you know if villian is calling or folding too much, if you don't know what the right amount is? Studying GTO will make you better at identifying smaller unbalances in your opponnents -> you can better exploit.

This is a VERY different statement than your OP.

Learning how to balance ranges and choose betting frequencies very often isn't about experimenting with changing styles, it is about plugging leaks.

If an opponent is deviating from optimal play then the highest ev decision will be a corresponding exploitative play, yes.

Your OP was incredibly convoluted if we were supposed to get that from it but whatever.

There are no patterns to discern from either GTO bot. There are no missing ranges to exploit. The bot just plays each hand perfectly, and gives zero thought to you or any other player or bot.

Good luck playing a corresponding exploit strategy against that.