Mid/High Stakes Limit Hold'em Poker Forums

Yes but isn't that entirely the point?
If our opponent is playing so that he is exploitable we want to adjust our strategy to exploit him, if his counter adjustment puts him more towards a GTO then Sonias play also become more GTO.

Basically, against all but brain dead opponents Sonias ranges would be balanced, for a certain value of balanced.

I guess the counter argument is that if Sonias is adjusting her play, then given a random run of cards from a truly GTO player against Sonia, if Sonia adjusts at all, she should be a losing player (if only marginally).

I think there are 2 things to take from it, we can't look at a hand from Sonia in a vacuum and expect it to be the GTO way to play the hand.
But the adjustments from GTO play Sonia does make vs. a player like Polaris are probably very minor so if we play our ranges like Sonia we are most likely not making any measurable mistakes.


All I am spending a fair bit of time on saying is, if there is any merit to checking back a balanced range on certain board textures from a GTO standpoint, then I think in the 'real world' there is an even greater reason to do so, because not only can playing closer to optimal never hurt us, but we are also putting our opponent into a situation he is likely much less familiar with given that checking back flops HU is not the ordinary play.

I agree with the last section.

But earlier on you said that if someone would play balanced then Sonia would play close to balanced too (right?), but I don´t see that since if someone who plays balanced ch/r a flop just at the right threshold for us to be indifferent to call down or fold. We might just call down every single time since our opponent doesn´t adjust. So as long as we don´t make stupid mistakes we can play in any way we want against a balanced player. Because he has already neutralized the EV of all decisions (except the braindead ones)

I agree with the last section.

But earlier on you said that if someone would play balanced then Sonia would play close to balanced too (right?), but I don´t see that since if someone who plays balanced ch/r a flop just at the right threshold for us to be indifferent to call down or fold. We might just call down every single time since our opponent doesn't adjust. So as long as we don´t make stupid mistakes we can play in any way we want against a balanced player. Because he has already neutralized the EV of all decisions (except the brain-dead ones)

Ah I see.

I was making the assumption that Sonia's default play would be GTO/balanced and she would adjust incrementally to the opponent.
So in your example if our opponent is making Sonia indifferent then Sonia would/should play GTO, but if he starts CRing slightly more then Sonia will incrementally call down or re-bluff with a wider range until we reach the threshold where the opponent starts readjusting.

Also I am not sure I understand exactly what is going on, the way I think of it there are no 'taking an action some of the time', just ranges.

Thus there is no indifferent to calling or folding just hands that are profitable call downs and hands that are not, so our range stays GTO.

Keep in mind some of the value in a flop and turn call come from the times the river will go check/check and we win the pot, thus it will always be just 1 hand at the edge of our range that is indifferent to our action in terms of EV, at least as far as I understand it.
(the argument here is that if hand A is neutral EV on a call then a hand stronger than A will be +ev)
I think this is also covered in the mathematics of poker but I could be wrong, I wasn't very adept at the game when I looked through it.

So if calling becomes more profitable then we widen our calling range, we would still be balanced in the sense that we are folding, calling and raising some of the time with a 'balanced' range.


Even more so you can argue that it is likely no hand is going to hit point blank neutral EV, it is always going to be an ever so slight winner or loser, so we are never truly indifferent to our action.

Sorry, but there is a lot of misconceptions about GTO play in your post. (or I´m not understanding you right)

First of, GTO play has nothing to about what is more profitable. So if villain only bluffs but never valuebets GTO play would still have us fold some hands. This you understand, right?

Second. Now if someone else is playing GTO and betting on the river it doesn´t matter if you call or fold your bluffcatcher since he has already made you indifferent to the choices. This is understood, correct?

Having said that it could still be that Sonia plays quite GTO when up against a GTO player. But of course his GTO style might be quite different to a truly GTO player. But that is where I am unsure.

Lastly, a GTO player does (with a high degree of certainty) play a mixed game where she does one actions some of the time and some actions some of the time. This is to ensure that she has those specific cards on different branches of the game tree.

I think there are a couple of reasons why the 100% c-bet has been so succesful in the past in especially 6max. First people has not been aggro enough OOP. U just bet flop, they call (or fold) and then u have full control of the hand on the turn. I mean if opp doesn't XR 88 on 223, because they fear u have aces, betting is safe.

Secondly I thnk 6max players mult-table and that a 100% c-bet is an easy strategy to implement and that it is often good enough. When u play many tables simplifying the game has a lot of merit. In some lower stakes games raisng a standard range pf, bet twice and decide what to do on the river can be enough to win.

IMO there is no chance a 100% c-bet can be correct in theory. The reason is that there is no essential (for this concept) difference between the flop and he turn.

Players that never caps OOP can do so because of the 100% c-bet. The player with initiative gives back the bet on the flop 100% of the time. Now imagine playing vs somone like DH. You sacrifice 1sb pf. Then he sometimes X back the flop. Meaning u miss one more street of value. This is the same principle as the FC raise. DH gets to see four cards with a "draw" (a hand with eqity, that is a dog vs ur range). But if he improves on the flop he can then choose to get his value instead.

Now all we hav to figure out is how to do this in practice ! And DH is on the right track IMO.

Second. Now if someone else is playing GTO and betting on the river it doesn´t matter if you call or fold your bluffcatcher since he has already made you indifferent to the choices. This is understood, correct?

This is correct in sipmler gams but in LHE ur bluff catching range is pretty wide. And ur opponent often has value betting hands that are within ur bluff catching range. Thus the so called bluff catchers will have different EV on a call. U need to find the so called "threshold hand". That is the hand that is indifferent between calling and folding.

I guess u allready know this. Just wanted to clarify.

First of, GTO play has nothing to about what is more profitable. So if villain only bluffs but never valuebets GTO play would still have us fold some hands. This you understand, right?

Yes

I am out on deep water here, and I concede you are a far better poker player than me but I think we are agreeing from different angles.

Second. Now if someone else is playing GTO and betting on the river it doesn´t matter if you call or fold your bluffcatcher since he has already made you indifferent to the choices. This is understood, correct?

Yes but what I was getting at is this is because there are no more betting streets after the river.
On the river we can make the assumption that an opponent would never bet for value with a hand worse than A and as a bluff with a better than B, so we can view the range between A and B as having the same EV.
However doing so deviates from GTO play, in that if we call with a weaker hand and fold a stronger we are not playing the most optimal strategy.
Just because a strategy is exploitable does not make it optimal, we still want to call with stronger hands than we fold in case our opponent bets a hand he should have checked.


It also doesn't hold for calling a bet on the flop or turn, in that again, giving that calling with a given hand is neutral EV, then calling with a stronger hand would be +EV. Because there is a chance we will win improved if no more betting goes in on later streets.
I do not think we can reduce a range of hands to equal EV over multiple streets.

When the choice is between calling and folding then we never fold a +ev hand and we never call (or raise) a -ev hand, doing either action would deviate from GTO play.

Lastly, a GTO player does (with a high degree of certainty) play a mixed game where she does one actions some of the time and some actions some of the time. This is to ensure that she has those specific cards on different branches of the game tree.

This is very obvious when it comes to preflop play, obviously if we always or never 3bet a certain hand preflop there will be certain boards where we cannot hold the nuts on etc.
However this is because we are not dealing with the option of folding, post flop is a different beast.
See previous argument, we can't fold a hand some of the time if calling (or raising) is +EV, and if calling (or raising) is not +ev then we always fold.

So when it comes to the choice between calling and raising we will have the same hand in both ranges, but there wont be a hand we fold only some of the time but call other times. (assuming no adjustments).


Or am I wrong in thinking GTO play does not contain a hand (range!) which we fold some of the time.

I´m sorry but I am having a really hard time understanding your post. The words seems to mean different things for you and me.

But you say that GTO play wouldn´t include folding a hand that has +EV vs his range and this is completely and utterly wrong. If opponent always bets the river as a bluff but never fold value for example (I use the river as an example because it´s the most easy to understand but it applies to all streets) then the "optimal" play would be to always call. But optimal is not the same as game theory optimal. And a GTO-player would still fold some hands that beats all those bluffs (even though it´s +EV to call!) because GTO has nothing to do with winning the most money and has everything to do with not opening yourself up for counter-exploitation.

sorry I was typing it while being a bit busy with other stuff, but I meant to say, GTO play would never include folding a hand that is +ev with another action vs. a player who plays GTO.
When talking about GTO play we are assuming our opponent is using the same strategy.

Thus my argument is, that we cannot have a hand which we call some of the time and fold some of the time if we are playing GTO.

I also made a few mistakes in my haste I meant to say, 'Just because a strategy is exploitable does not make it optimal'

Or am I wrong in my assumption that given knowledge that calling on the flop or turn with hand A is neutral EV, then we always want to be at least calling with all hands stronger than A?

sorry I was typing it while being a bit busy with other stuff, but I meant to say, GTO play would never include folding a hand that is +ev with another action vs. a player who plays GTO.

On the river (you might be right) but earlier on in the hand you´re not. They both have to mix up their range in such a way that they are not exploitable (it could still mean taking an EV loss) to have a certain hand on a different branch. (which might make the Ev loss back again taking the strategy as a whole)

When talking about GTO play we are assuming our opponent is using the same strategy.

Generally no, but if you so wish.

Thus my argument is, that we cannot have a hand which we call some of the time and fold some of the time if we are playing GTO.

Well if a person needs to peel 0.1% with low cards to bluff a later street (just a rough example) then a single hand let´s say 54o could still represent to many combos and so we would call a certain % of the time with 54o, no?

I also made a few mistakes in my haste I meant to say, 'Just because a strategy is exploitable does not make it optimal'

Of course not. A question: When you say optimal, do you mean GTO or optimal as in Optimal Exploitive Strategy?

On the river (you might be right) but earlier on in the hand you´re not. They both have to mix up their range in such a way that they are not exploitable

But then our play is per definition not GTO since our opponent could play the same strategy but make the play less often and profit from us, if the amount of times he doesn't make the play is low enough we would be unable to tell we are being exploited.
Taking less than the max EV decision only comes into play when choosing between calling and raising, or betting and checking.
When we are talking about the lowest range of our hands and purely making the choice between a fold, a call or a (bluff) raise, then the EV of the action (excluding non pair draw outs), will be based on the hand strength.

Also on a further note I am unsure how far the concept of 'being able to have any hand' actually values in a limit game, Mathematics of poker talks about playing more hands in a no-limit games as stacks gets deeper, in order to be able to hold the nuts on any board.
Thus it is very likely that we are much less concerned about being able to hold a every hand in a given situation.
In fact I wouldn't be surprised if GTO play for a limit game is to take the highest EV action under the assumption that our opponent is playing GTO as well.
In fact the argument is the same, if we take a less than max EV decision, then our opponent can play the same strategy but take the action less often, thus profiting from us.

Of course not. A question: When you say optimal, do you mean GTO or optimal as in Optimal Exploitive Strategy?

Well GTO, I was actually just using the mathematics of pokers definition, where unexplainable, is just a strategy our opponent cannot profit from, and optimal is a strategy that our opponent cannot profit from but will profit more if our opponent is making mistakes.
Folding Q high and calling with J high might be unexplainable, but in that case doing the opposite is optimal.

Well if a person needs to peel 0.1% with low cards to bluff a later street (just a rough example) then a single hand let´s say 54o could still represent to many combos and so we would call a certain % of the time with 54o, no?

Maybe, but I also think it will be a while before the game is solved to that point.

But then our play is per definition not GTO since our opponent could play the same strategy but take said action less often and profit from us, if the amount of times he doesn't take said often is low enough we would be unable to tell we are being exploited.
Taking less than the max EV decision only comes into play when choosing between calling and raising, or betting and checking.
When we are talking about the lowest range of our hands and purely making the choice between a fold, a call or a (bluff) raise, then the EV of the action (excluding non pair draw outs), be based on the hand strength.

Also on a further note I am unsure how far the concept of 'being able to have any hand' actually values in a limit game, Mathematics of poker talks about playing more hands in a no limit games as stacks gets deeper, in order to be able to hold the nuts on any board.
Thus it is very likely that we are much less concerned about being able to hold a every hand in a given situation.
In fact I wouldn't be surprised if GTO play for a limit game is to take the highest EV action under the assumption that our opponent is playing GTO as well.
In fact the argument is the same, if we take a less than max EV decision, then our opponent can play the same strategy but take the action less often, thus profiting from us.

Well GTO, I was actually just suing the mathematics of pokers definition, where unexplainable, is just a strategy our opponent cannot profit from, and optimal is a strategy that our opponent cannot profit from, but will profit more if our opponent is making mistakes.
Folding Q high and calling with J high might be unexplainable, but in that case doing the opposite is optimal.

I don´t know what this discussion is about anymore. It seems like we are talking about some of the same things with different wordings.

The thing of folding a hand that is +EV to call is what I defined earlier as a stupid mistake and as long as you (or a bot in this case Sonia) avoids those you could play pretty unbalanced and still go +/-0 against a GTO player. But there are other decision where it´s not about possibly making a stupid mistake (and one of those is calling or folding with a bluffcatcher on the river). That is why I said playing exploitively against a GTO player could still lead to plays that are not GTO (since they could be counter exploited if only your opponent had the ability to adjust).

WTF is unexplainable? or do you mean unexploitable?

Maybe, but I also think it will be a while before the game is solved to that point.

I thought you wanted to know why a bot might chose a certain action some % of the time and another action some % of the time holding the same hand.

ninja you reference mathematics of poker in your post but I agree with darkhorse that some of your assumptions and definitions just don't make sense. For example GTO play makes no assumption whatsoever about the opponent's strategy. I think you are confusing yourself at this point and should re-read some of the parts of MoP where the authors try to explain exactly how the terms are strictly defined in game theory and how a GTO strategy differs from an exploitative one.

So I´ve tried out my thoughts from http://www.deucescracked.com/forums/1-Mid-High-Stakes-Limit-Hold-em/23537-Thoughts-and-Analysis

I´ve started to check/back flops as a valuecheck (folding some hands on later streets) whenever the board and villains range suggests that 1. he will check/raise a lot 2. I will put in a difficult spot when he check/raises (should I call down or not) 3. He doesn´t fold that many hands and the hands that would fold doesn´t have that much equity anyways.

I´m not asking anyone to agree with my just thought it would lead to some discussion..

5/10 Limit Holdem
6 players

Pre-flop: (1.4 SB, 6 players) Hero is CO T A
UTG raises, 1 fold, Hero 3-bets, 3 folds, UTG calls

Flop: 3 2 3 (7.4 SB, 2 players)
UTG checks, Hero checks

Turn: 6 (3.7 BB, 2 players)
UTG checks, Hero checks

River: Q (3.7 BB, 2 players)
UTG bets, Hero folds

----

15/30 Limit Holdem
4 players

Pre-flop: (1.7 SB, 4 players) Hero is CO A J
Hero raises, 2 folds, BB calls

Flop: 8 T K (4.7 SB, 2 players)
BB checks, Hero checks

Turn: 9 (2.3 BB, 2 players)
BB bets, Hero calls

River: 8 (4.3 BB, 2 players)
BB bets, Hero calls

Final Pot: 6.3 BB
BB shows: J Q
Hero shows: A J

----

5/10 Limit Holdem
4 players

Pre-flop: (1.4 SB, 4 players) Hero is CO Q K
Hero raises, 1 fold, SB calls, 1 fold

Flop: 8 J 8 (5.0 SB, 2 players)
SB checks, Hero checks

Turn: 9 (2.5 BB, 2 players)
SB bets, Hero calls

River: 8 (4.5 BB, 2 players)
SB checks, Hero checks

Final Pot: 4.5 BB
SB shows: 8 7
Hero shows: Q K

Hand 1- This is a super easy valuebet, i dont think you get c/r here very often fwiw. Call riv as played

Hand 2- I dont hate it, id probably prefer to make life easier and bet the flop and have an easier turn decision. If he is c/r at a high % i can see checking here on the flop but you rhand kinda looks like what it is.

Hand 3- I probably just valuebet here also, i mean it depends on what villain's tendancies. The pot is small, if you think you can induce bluffs then i dont mind checking

I think there are better boards than the above to be checking back SD'able hands. It all comes down to balance, but TBH, i dont know i would waste too mich of my time playing against a villain where i need to be altering my game to this degree. Id find a better game.

The money in HULHE still comes from fierce valuebetting.

Give me a 579dd flop and i like a check back with KcTc for instance against all but the more tighter/straighfwd villains but the above hands i like betting the flop with all of them

I think people are making to much out of the difference between the theoretical concept "GTO" and our everyday use of "optimal". With optimal we mean a strategy that it seems very hard to exploit. This is all we can talk about. We dont know, and the guys who build the bots dont know what GTO is cause poker isnt solved yet and it seems that it is unsolvable for more then 2 players.

Also, u cant discuss hands indivdually. If an aggressive player like Bryce suddenly started to play a few hands exactly like Sonia, it would probably not make his game more optimal, it would just be pure misplay when put into his overall strategy. U cant say that if Sonia is checking a certain hand on a certain flop it got to be optimal to do so, not unless u play all ur other hands like Sonia.

Hand 1- This is a super easy valuebet, i dont think you get c/r here very often fwiw. Call riv as played

People, will you please stop referring to hand 1 as a valuebet. It´s a semibluff in position (with a hand that has SD value in the hopes of staying in control of the betting on later streets). I think we need to not be sloppy with the word use valuebet for us to anaylyze hands better. I see people say valuebet all the time when there is nothing value about it whatsoever.