tmpfs, we all thought your post was an expert level. Now, less sure.
By your reasoning, GTO ROSHAMBOT would never throw rock. Rock is the aggressive action, and our unexploitable bot would never do that. Thus, if you see someone who's playing ROSHAMBO and never throwing rock, you know he's GTO and that you can't beat him. QED. Is that what you're getting at?

In the off chance that tmpfs isn't leveling, I'd urge him to consider that GTO strategies are, by definition, maximally exploitative against the nemesis (among other things).

I will not further this discussion as it just generates meaningless ranting, I will however conclude by answering the replies that seemed appropriate.

I'm not leveling anyone I do believe in what I said, and I am a profitable HU FLHE player, what happens is that we seem to have different notions of GTO.
I'm afraid I'm not too familiar with roshambot study and so I'm unable to transfer conclusions to HE I do think however that understanding GTO would pass through studying a simpler game like you are doing.

I view it this way, as a player you can choose to play it two ways:

a)Play based on defined inact decisions based on your thought process
b)Play based on your thought process as a function of villain's behaviour

or...

c)Play random (not valid in my point of view, has it requires inputs from outside the game)

So in my explanation of GTO only a) is true


Rei your definition I believe I read on Sklansky book, I do think however that if these two next conditions are true:

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

then the condition you state:

3) GTO strategy is maximally exploitative against the villain.

Can't coexist with condition 2), as to be maximally exploitative against the villain can't be done without considering his behaviour, or a sample of behaviours as it is done with bots.

I'm gonna stress again that if we are to properly analyse this problem we must establish the boundaries or conditions or rules that define the GTO strategy in a theorem like form, so that we can conclude through logic which conditions can coexist or not, this is, if what we are talking about actually exists or not. In my approach I only considered the 2 conditions that seemed to make sense to consider, 1) and 2).

Nope, reread:

The recent activity in this thread really underscores how far these forums have fallen in the past few years. Just remarkable.

in the off-chance he's not levelling and he's just being really arrogant and dumb as **** (AND FFS, NOT APPRECIATING ANY ****ING HELP AT ALL), can i know what in the hell could possibly be the reason for trying to educate him???

Since it appears that you're serious, I'll briefly wade in here and say that your understanding of GTO is completely flawed. Your thinking seems to be (if I'm understanding you correctly) that a GTO strategy must always fold kings preflop because otherwise our opponent could exploit us by only ever raising aces preflop. This is incorrect thinking, however. An opponent who only ever plays aces preflop will, indeed, make money off of us when we are holding kings, but you neglect to factor in that our opponent is also going to be bleeding money by passing on tons of profitable situations when he folds hands like QQ preflop, and the net balance is that a GTO strategy will make much more money when our opponent folds non-aces preflop than it loses when he does get aces.

In fact, by the definition of GTO, there is no strategy our opponent could take to exploit us and be +EV. Whether he plays aces only, 100% of hands, or anything in between, the best he can hope to do over his entire range is break even. Before you assert that such a strategy is impossible, it is mathematically proven that a GTO strategy exists for poker.

Also, "minimizing losses" does not mean check/folding every time you don't have the nuts. If KK wins $100 75% of the time and loses $100 25% of the time, not playing KK is definitely not "minimizing your losses" just because it will sometimes lose, because it's still +EV to play it.

It's remarkable how new posters keep showing up with authoritative treatises on topics they don't know a god damn thing about.

To me it's remarkable how that gets panties in a bunch.

So he's wrong, who cares?

^^ This

I don't really understand the pleasure in trolling yet I believe it has something to do with bunched panties.

No you aren't understanding me correctly.

So lets consider a steady-state strategy where you only play the range [KK,AA] because according to you playing the kings would be profitable. A perfectly exploitative villain with a transient strategy that over time will know exactly what you are doing, will call you preflop with every hand and extract maximum value on the following streets because you will call down kings on every street because that's your preset strategy, because you play despite villain's historical behaviour because it was defined that you should go to showdown with kings because it's supposedly +ev. By this I mean that post flop he will call or raise anytime he's ahead of aces or kings, or simply has pot odds to keep calling. Please remind that a human being isn't capable of randomizing, and a preset decision will happen 100% of the times, as you don't exploit your villain.

This being true KK is now -EV and you lose less by only playing the AA.

This example will spread to whatever range you consider, ultimately a perfectly exploitative villain will figure out your range and what actions you take and what is the perfect strategy to extract the biggest ammount of value out of you.

Like in the game that everyone tries to guess a number that is an average of the average number that the others are guessing, in which the solution tends to zero, here in poker it tends to the nuts. As the nuts have zero probability of losing.


I give up trying to expain to you guys what I'm saying as it is now clear to me that you guys either are too stubborn or not intelligent enough to understand what I'm saying, and that non of us will change our minds. This is my last reply, got games to grind, better stuff to think about.

*Note that I haven't insulted anyone here, apart from proving people dumb* (now this last line was a level btw)

I also have no desire to discuss with you, but if you want to understand why you are wrong, google or wiki "Nash-equilibrium". That is proven to be the game theoretically optimal solution, which already exists for simplified poker games (a good example is the push-fold chart for short effective stacks), and does not involve folding KK pre very often.

In poker that means that all players have adjusted their ranges and action frequencies until further adjustment gains no profit, and is a state of unexploitability, contrary to the extremely exploitable strategy you propose.

We could consider many strategies where I play horrible poker and you get to win. Would it be fun to do this?
and thus proven that the only reason to argue on the internet is because arguing is fun for its own sake. Many of us have learned a lot about poker, but not in this part of the forum. The trend continues.
Would it be over the top mean to say "that Zynga poker won't play itself"?
Among many, I hope.

Why would anyone bother responding to someone who starts off by claiming to be the only person who understands game theory in a thread that includes multiple people who've written books on the subject and then says something completely ******ed? Did you really think this was going to go better than it did?

Awesome

Sorry to bump this again but I bookmarked it about one year ago after happening across it somewhere online and only recently got to wade through it all.

I couldn't help but remember back to a video on youtube of the FT v PS HU battles from last year. It showed a few clips of Ike & Durr sitting opposite each other with a smirky Durr looking across at a stoic & bespectacled Ike (who was glued to the screen) as if looking for a live tell in the online game. Not exactly a crazy idea in itself but nonetheless ironic as it kind of symbolically ties in with this thread i.e. Professor v Hustler.

I don't know which was more bizarre: The original assumption that he was the only one in the World who "knows," or the idea to play 200k hands with no buyout clause seemingly without first working out where that could lead to i.e. losses that neither side would be willing to endure even @ 1BB/100. Not to mention the fact that subsequent non happenings have resulted in the realization that Durr almost certainly just wouldn't finish such a challenge unless he was winning. No wonder he didn't want a buyout clause!

Anyway, don't want to pick on the guy anymore.

Except to say he must of ran really really good for a long time. Just saying...

What has been bumped cannot be unbumped.

Who would have thought one simple character could convey such astonishing depths of sarcastic contempt lol. GTO trolling imo.

Yes

http://oyc.yale.edu/economics/econ-159


Nuff Said

Does durrrr still think he can beat GTO?

probably. didn't he beat the computer?

Did he ever think this?



EDIT:

Yea, I guess it looks like he did think that.

Or actually maybe not. Maybe he thought that GTO did not exist for poker.