I think a problem with an obsessive game theoretic approach is that it can cause you to focus so much on your own hands rather than what your opponent is doing and how they play. If you memorize things incorrectly you could be making serious blunders.

Another thing is the current GTO approaches only consider 1 bet size. I don't think it has been proven that you must select all of a certain hand type in 1 bet size. Perhaps it is game theoretically correct to bet smaller with the nuts on a certain flop, but jam with a strong yet vulnerable hand.

i don't know why you believe that, it is certainly not true.

By that I meant only solve for one bet size at a time. 1/2 pot bet vs. 3/4 pot bet separately rather than simultaneously.

That's still not true. With Pio or Similar, you can give it 10 different bet-sizes to choose from. It will take forever to find a solution, but it's definitely possible for it to do the calculations, and put various combos in multiple bet-size buckets at different frequencies.
It's only Snowie that uses 1 size for its entire range, and my personal belief is that using one size for each spot (whether it's small or large) doesn't sacrifice much EV overall. e.g. Maybe shoving the nuts wins 15.5bb, whereas betting 25% of pot wins 13bb, but the nuts is only a tiny part of your range, so missing out on 2.5bb once every 10,000 hands or so by picking a sub-optimal size isn't gonna kill you.

We may have to agree to disagree. MDF is not equilibrium, just simple math based on pot odds.

How would you know (unless you yourself are a perfect GTO player)?

Maybe it just looked like they could be exploited. Maybe they called 10 times in a row in a spot where GTO says 33% call/66% fold...just because their random number generator just went that way this session. That would look super-exploitable to you even though the player is perfectly GTO.

Until no limit hold'em is solved it's a pretty good assumption.

My view of GTO is that it is just a very good tool for improving win rate. Like that old joke, I don't have to outrun the bear.

Why use mdf as an example of gto/exploitative then?

Oh, wow didn't know that. Seems pio is cooler than I thought.


Fishing and YouAreAwesome, I think this can be settled quite easily.

Fishing used MDF, which is a game theoretical concept to build a strategy. He never mentioned anything about exploitation, but I'm guessing that is implied. Fold more than your MDF strategy, if your opponent is reluctant to bluff. Call more than your MDF strategy, if your opponent is a chronic bluffer.

Also, nobody plays exactly GTO. I think it is pretty ridiculous to claim otherwise. I mean, theoretically there could be a pink bunny with 5 eyes living on Saturn, but I think it is safe to say that there isn't.

You're not gonna (or at least shouldn't) defend MDF in all parts of the game tree.

Fishing claims to use GTO/exploitative styles mixed together. But to show this gave an example of a hybrid MDF strategy. He said nothing of exploiting. He said nothing of GTO. He did not qualify his initial "mixing" claim. In fact, if you are defending his argument by saying that the exploitation is implied, then he is adding nothing to the conversation (sorry to be blunt). This point has already been stated multiple times ITT. In fact, all he would be saying is that MDF is a GTO strategy that can be used to exploit people. Two problems. One, he didn't actually say this. Second, who cares, it's obvious MDF can be used in certain circumstances to help make decisions.

For the second thing, about no one playing GTO, this is not what he said. He said, "I haven't seen it" as though he would notice it if he did see it. Anyway, nice attempt to blur the points as though I just "misunderstood" what he was saying, but nope.

We can assume that fishing plays exploit but works in a little GTO to be more balanced than he might be while purely exploiting.

This is fine, leads to better poker, and more profit.

I would suggest hybrid players just play GTO but reserve the last action in the hand as a “exploit double check” of the target opponent.

Also those comments by the other poster had little to do with OP's original questions (I realize that other people may disagree on that point). Which were actually good questions, or at least had the potential of prompting a thoughtful, lively, and informative discussion.

But, as has been proven innumerable times, it is virtually impossible to have a thoughtful, lively, and informative discussion if GTO is in the title of the thread (or makes an early appearance).

I'd like to add that too strong of a focus on GTO can restrain creative thinking. There are a lot of spots where you can make some pretty cool moves that GTO players just ignore and play a strategy that they memorized off of some chart. Sometimes, I believe people try and over focus on GTO out of fear of being exploited. I think it is really important to do something and be aware of "Hey, I'm doing something that could be really bad if my opponent adapts correctly. If he outplays me, he outplays me. I am willing to take that chance."

I think too many people revert to it when it really should be the back burner strategy. On a river situation in an unusual WA/WB situation they will call with their top bluff catchers and fold their bottom bluff catchers because they don't want to play a guessing game. I think people should play the guessing game. They should try and determine their opponent's value hands, bluffing hands, and what they are checking back then make a decision accordingly. If it is close, then they can go the game theoretical route. A lot will just go the game theoretical route. This skips an entire thought process that I think is crucial for development as a poker player.

I have talked to some really good pros and what I've heard is that a primarily GTO approach will make you win, but a primarily exploitative approach is what will make you win big.

For me personally, I probably should focus on more game theoretical sound moves because sometimes I can get a bit too crazy and end up spewing. For most players, I think they are going too much of a theoretical route and becoming robotic in their play style.

I played at a table where a guy was opening nearly 100% of hands and barreling a lot and the other players hardly adapted. I think this is a huge mistake for them. I ended up calling off his stack with queen high on a scary river and he had 62o. There were a lot of hands beating me on that river, but he had too many bluffs for me to fold. If I was too focused on a pseudo GTO strategy, then I wouldn't have even thought about calling.

I mean, I get that I should adapt to the times and all that, which hopefully I am somewhat doing. I actually don't like how GTO has changed the games though. It is like an inevitable plague in the poker world.

QED

Exactly. My post was about how to maximize profit. MDF is a great tool for that. I should probably avoid mentioning GTO in future post. Riles people up. Against players that make mistakes GTO is not the same as maximum profit.

MDF is a great tool for playing the river heads up out of position, as well as heads up raising wars on all streets*, but that's where the usefulness ends.

*because if you open the betting when heads up, you should have a betting range that can withstand action without allowing immediate profit. However, this does not apply to flop or turn checking ranges, nor do busted draws that cannot beat a bluff need be considered in heads up out of position MDF calculations.

thanks for contributing everyone

as someone who grinded/studied alone for many years it is exicitng to make threads like this in 2+2 and read all your responses, poker can be such a solitary game, there is a lot of value in being part of a community

the way this thread unfolded highlights what motivated me to ask those original questions: game theory offers relatively complex solutions, and i think there is room to simplify or improve the way we implement these concepts 'in-game.'

These conversations about GTO always seem to be marred by definitions of exploitative and optimal play, or confusion about how to utilize game theory. The fact that these threads always go off the rails is an example of the complexity that GTO has created.

I know most of you have a firm grasp of all these definitions, and how to use game theory 'in-game.' However I think there are certainly areas where we can improve the way we utilize game theory on a practical level.

Its great to read the discussions on bet sizing, heads-up based solver solutions, MDF, frequencies, bet sizing, etc. These are exactly the places I think we can improve/simplify our implementation of game theory: how to solve for multi-way scenarios more effectively, how to find the appropriate frequencies of a mixed strat in-game, how to simplify MDF in game, how mixed-bet sizing impacts our strategies etc.

I am reminded of a quote: 'things should be made as simple as possible, but no simpler.'

I have reached a different conclusion. Using simulation in 6 max no limit, I have been able to dramatically improve the win rate on every street and every position. I first evaluate hands ( flush draw,Top pair, etc. then reduce the calculated MDF by the hit rate (over simplification). What remains is the strongest hole card range that I continue with (not fold). I am sure it works, but I am equally sure that there is a better way. Maybe I misunderstood your post.

Hey I see how the thread went went this direction, if this thread continues I hope it can stay on topic

like i said before, the fact that these discussions usually go off the rails is exactly why I wrote those original questions: gto is complex, many struggle to implement the ideas, and there are definitely ways to improve the way we understand/utilize GTO.

Meeting the balance of frequencies that solvers create is difficult in game, you cannot run sims for every hand, and randomizing can also be difficult.

There MASSIVE simplifications that can be used in an optimal strategy which do not forfeit hardly any EV.

For example if you plug a board into PIO, the solver might give you a 80% bet frequency with a small size, and 20% check freq.

If you use the node-lock option, and use 100% bet frequency with a small size you will often see that very little EV is compromised from this strat option, perhaps 5% (this is dependent on the board and range though). There are many nodes/board types where this kind of simplification can be created. Then you might change that node-locked sim and incorporate 10% check.

These kinds of solver simplifications can be created in a variety of ways, there are many ways to drastically reduce the complexity of a GTO strat.

Instead of having to meet those complex frequencies that a 'perfectly' balanced solver solution offers, you can reduce the complexity by simplifying the strategy without compromising any significant EV. Plus the fact that you will be able to implement this strategy more easily will actually increase the effectiveness of the strat.

im sure many of you who use solvers already know about these kinds of simplifications, but I bring this up because this is an example of a way in which GTO can be improved IN GAME