what's all the hype about this gto stuff? it is neither the best nor the most profitable way to play poker; there's actual proof of that, so why even bother?

To negate any edge a superior opponent has over you

is there any good masters where i can study deep game theory?

completely understand if this gets no response:

Just having one of those moments where I can't see the obvious... need some perspective.

assume two gto bots face off in a HU match over an infinite or significantly large sample. Assume they both know how to implement gto strategys for every potential node and do so for the duration of the game.

I believe it is ok then to assume that arising from this match there would be numerous situations were one bot checks to another on the river.

Grant me the ability to tamper with one of the bots only when this situation arises. I understand that only pedalling the nuts in this situation loses the value gained from value bets etc. and would not profit over the pure bot.

Assume though that the bot continues to bluff and valuebet within the realms of their gto understanding for all hands that fall outside the term "nuts" on the river.

At this juncture we intervene and make the bot overbet jam all nutted hands. Given that bots don't respond to bet sizing etc. and just operate at optimal frequencies. Would one be right in saying that the tampered bot would beat the pure bot, negating variance??

Or does overbet jamming nutted hands simply become the gto strategy?

if this has already be dealt with earlier ... you're well within your rights to tell me to EAD...it's what I would do!

Oh hey durrr! Nice to see you have a gimmick bro!

Optimal frequencies take sizing into account. If these bots aren't doing that, they aren't GTO bots.

If the bots' current strategy, which involves jamming x% of nutted hands, is in fact GTO, then deviating from that would be sub-optimal.

ahh they take bet sizing into consideration... I obviously got confused somewhere along the line rei ty!

action... I don't think you realise how much you just insulted durrr

what do you mean?

I think I'm the only who seems to have a clear understanding of what GTO actually means, so I'm going to post my explanation.

First let's define properly GTO:

1) GTO is defined as the strategy that minimizes losses.
2) GTO is defined as the strategy that doesn't exploit villain's behaviour.

Therefore, if these hold true and get verified always, then:

Using the GTO strategy you must check in every action and fold every time you face a bet, this until you hold the nuts, then bet and raise all the way.

-You can never call or raise because that implies that you estimate the likelyhood of your opponent having a worse or better hand than yours, therefore you'd be exploiting your opponent's profile. (conditions 2 not verified)

-If you would choose to always call or always raise despite your opponent's actions you're not minimizing losses. (condition 1 not verified)


So in conclusion, GTO is always leading you to lose the game against someone who exploits you by simply betting, always break even or win against another opponent who uses GTO as well, this is, that also just checks or folds until time runs out and both are forced all in, in cash game the two GTO players will break even as there's no time limit and considering no rake. Also this means that you can only play aces preflop.


Hope you guys enjoyed my explanation and please let me know what you think.

ok

orly

yarly

If Phil Galfond doesn't use it, I dont want to learn it.

Oh my god you're totally right what have we all been doing in this thread this whole time??

D- lvl

nice first post tmpfs

<< my reaction

Lol my first post and I'm already being mocked.

But I tell you what, I'd like to see someone describing GTO through conditions like I did making it a well defined system rather than putting meaningless dialog text.

The idea that everyone seems to have about it is that there's a group of actions that you can make and minimize exploitation on the long run against any villain, well, tell you what, that is not possible, your opponent will always be able readjust to your strategy searching for a nash equilibrium, which will not happen as poker has way to many variables for it to happen.

What everyone is assuming is a sort of an average skill for the villain and then saying that the villain skill is not accounted for. This is a 2 people game (minimum) and therefore your action is always a function of the other player actions, you can't play random as you would need an input from outside the game. You can only just check or fold until having the nuts like I explained in my last post as you make no meaning of the villain's action.

The pot doesn't start at zero. There are blinds/antes. As a starting point, a player must defend his fair share of the blind/antes to avoid being exploited. Then as more money goes into the pot from there, the player must defend their fair share of whatever is already in the pot.

tmpfs go back and read the thread as several people have come in and said similar things to what you have said and it has been explained to them why they are wrong.

Why must you defend? You're assuming agro behaviour from the villain. To defend is to exploit, this is, you're exploiting the exploitation, thus it's not GTO. You're assuming an average skill for you're villain. What if he only attempts to steal with aces? What if suddenly all players in the world only steal with aces than you should never defend the blind, besides he can just call around.


I read no explanation similar to mine, all the posts assume an average skill for the villain and or bluffing possibility. You can never bluff without knowledge of your villain or sample of villains. What if suddenly all the players in the world would become calling stations that call 100% of the time. Bluffing would become meaningless. You can't assume nothing about your villain using an ideal GTO. What you guys are all describing are counter exploitative strategies for specific levels of villain skill that are themselves exploitative strategies.

the problem is that you have a huge misunderstanding or no understanding at all of the fundamentals of game theory, and yet act as if you did

I'm all for helping people who want to learn, e.g. someone who reads the thread and then has some questions. But when the opening line is: ... then I don't really feel like making the effort.

If you have actually read the thread properly you are either a moran with zero reading comprehension or you are trolling. Go back and try again before asking any more questions/******ed statements that have already been answered/dis-proven multiple times.

This guy convinced durrrr he could beat the bots and look what happened to durrrr

Kind of funny, but only if no joke.