The anti HEM, stats, GTO, and balancing thread

[RedJoker]

Originally Posted by jaykay25

GTO play given villain's HEM stats

This is meaningless, there's no such thing as GTO based on HEM stats. GTO is GTO regardless of the opponent you're facing, their stats, what they did in the last 10, 50 or 1,000,000 hands or any type of game flow.

I'm pretty sure I'm talking about a completely different thing to everybody else at this stage, I have no idea why game theory would be grouped together with HEM stats, they couldn't be further apart. I think the issue is just a misunderstanding of game theory and balance.

If OP had said that sometimes HEM stats can be misleading and that once a dynamic has been created then this should supercede HEM stats because your opponent will be adjusting their game to you and that the play you would usually make against their particular stats might be horrible given recent flow then I'd have absolutely no issue with it. In fact, I think it would be a very worthwhile topic. That would have absolutely nothing to do with game theory though.

Originally Posted by jaykay25

lol i just realized ur the guy who made the (awesome) math videos for leggo. Trus me brah theres something to this psychology thing too : D )

Thanks.

I'm not saying there isn't anything to psychology and obviously reads are hugely important. HEM stats are another form of read and are an exploitative tool, just like psychology and reads. If you want to argue that one of these is more important in a given situation then that's fine. However, game theory and balance aren't exploitative tools, you're pretty much comparing apples to oranges and I think that's where the disagreement stems from.

[lewialex]

Thank you all for the feedback, especially Red Joker and JayKay.

After reading Joker's criticisms, I'm starting to see that the thesis of OP was muddled and I didn't get my point across; however, JayKay stole the rebuttal out of my mouth, and he did a very good job clarifying the point I was trying to make.

Not only was I trying to express the value of gathering reads on specific opponents, I was trying to emphasize the importance of 'timing', a term I failed to include in the OP.

[Xeta23]

Originally Posted by lewialex

Not only was I trying to express the value of gathering reads on specific opponents, I was trying to emphasize the importance of 'timing', a term I failed to include in the OP.

I'm not sure if this is news to anyone. I'm also not sure what your point is OP. As far as I know, what any decent thinking player is doing is, trying to figure out where his opponents are exploitable and doing whatever he thinks is the best way to exploit that, while again doing what he knows best to not let others exploit him if they are capable. So basically if one of the sides is significantly worse (or playing worse) than the other, he consistently loses money in the spots (cases, way he plays, whatever) where he is exploitable because either he's not aware he's being exploited or he doesn't know how to adjust correctly OR if both sides know the game and the concept of GTO pretty well and are also playing well both of them get closer and closer to playing optimally (well-balanced, near-optimal bet sizing etc.) as they play. This process obviously includes a lot of guessing (or leveling, which are what you mean by timing). You guys are talking like we have a choice of whether to play GTO or not. And I'm not a game theory expert or even a very good poker player but I'm pretty sure we don't know how to play GTO. Even if we did, as far as I know, humans can't play GTO poker (humans can't even play optimal roshambo even though it has a very trivial solution). So basically, what you said is what people are already doing.

I'm pretty sure that RedJoker is in no way trying to say that timing, psychology or meta doesn't matter. I'm also pretty sure he's not saying the smartest way to play poker is GTO vs. everyone. What I understand from what he said is, even if we're trying to maximally exploit when our opponents are exploitable, we have to have a good understanding of the concept of optimal poker, since being "exploitable" is relative to the optimal. So if we have no idea about what the optimal ranges/frequencies/bet sizes etc. are in given spots, we can't really exploit anything. So yeah, it kinda is a prerequisite to develop an exploitive strategy.

Also, since HEM stats only show how villain played in the past, it can only be a rough guideline to guess at villain's tendencies. It can not show what villain's frequencies are in specific spots no matter what the sample size, since, unless he's using some kind of randomizer and never deviating from his certain strategy he doesn't have "frequencies", he can only have tendencies. And as joker said, has nothing to do with game theory.

These are all afaik obviously. Correct me if I gave any false info.

[lewialex]

I didn't know that saying being balanced or playing GTO can be catastrophic it would trigger so much backlash. But rightfully so, because obviously playing balanced makes your game exploitable, so I definitely misspoke. Let me be more precise.

Let's say you play the following hand with an opponent who is aggressive.

He raises on the button and you call in the big blind with 88.

The flop comes 4c 5s 7d. You check, he cbets, you call.

The turn comes the Jd. You check, he bets, you call.

The river comes 2c. You check, he bets, you call.

He shows down Kd6d, and you win the pot


Now suppose some time passes and then you play the following hand with the same villain, again button versus big blind. He has maintained his aggressive image and again you have Qs10s.

The flop comes 10d 5d 2h. You check, he bets, you call.

The turn comes Kh. You check, he bets, you call.

The river comes 2c. You check, he bets

Now, is he really going to be balanced in this instance? Probably not. His range is going to be very skewed here. He knows you are capable of calling down light when a thousand draws miss, so his range is going to be weighed heavily towards value bets.

Even your read that your opponent is aggressive and capable of barreling off might be misleading given the timing here.

Now, the purpose of my OP was not to share some wisdom that I have. Far from it - I'm looking to learn more about this aspect of poker because there is so much confusion and uncertainty.

Here are some issues that I have that I want to get a good grasp on:

1. Do we know if the player is capable of adjusting his play?

2.If so, would he adjust his range in certain spots after one hand? How quickly/often is he adjusting to your perceived ranges?

3. How can you assign a level to an opponent? For instance, is he going to assume based on the last hand that you are going to adjust right away and therefore not adjust his range at all? Or is he not even going to think that far?

4. How does history in one kind of board texture translate to other board textures? For instance, if you catch a player 3 barelling in good bluff spots, what are is ranges going to look like in spots where bluffing is traditionally bad?

I think that what elite players do best is they are able to get a beat on their opponents. They understand when their opponents are adjusting and readjusting, and anticipate it.

I'm sorry if this is incoherent. I'm a bit tired and my thoughts are a bit muddled.

[RedJoker]

Originally Posted by lewialex

I didn't know that saying being balanced or playing GTO can be catastrophic it would trigger so much backlash. But rightfully so, because obviously playing balanced makes your game exploitable, so I definitely misspoke. Let me be more precise.

Let's say you play the following hand with an opponent who is aggressive.

He raises on the button and you call in the big blind with 88.

The flop comes 4c 5s 7d. You check, he cbets, you call.

The turn comes the Jd. You check, he bets, you call.

The river comes 2c. You check, he bets, you call.

He shows down Kd6d, and you win the pot


Now suppose some time passes and then you play the following hand with the same villain, again button versus big blind. He has maintained his aggressive image and again you have Qs10s.

The flop comes 10d 5d 2h. You check, he bets, you call.

The turn comes Kh. You check, he bets, you call.

The river comes 2c. You check, he bets

Now, is he really going to be balanced in this instance? Probably not. His range is going to be very skewed here. He knows you are capable of calling down light when a thousand draws miss, so his range is going to be weighed heavily towards value bets.

Even your read that your opponent is aggressive and capable of barreling off might be misleading given the timing here.

Now, the purpose of my OP was not to share some wisdom that I have. Far from it - I'm looking to learn more about this aspect of poker because there is so much confusion and uncertainty.

Here are some issues that I have that I want to get a good grasp on:

1. Do we know if the player is capable of adjusting his play?

2.If so, would he adjust his range in certain spots after one hand? How quickly/often is he adjusting to your perceived ranges?

3. How can you assign a level to an opponent? For instance, is he going to assume based on the last hand that you are going to adjust right away and therefore not adjust his range at all? Or is he not even going to think that far?

4. How does history in one kind of board texture translate to other board textures? For instance, if you catch a player 3 barelling in good bluff spots, what are is ranges going to look like in spots where bluffing is traditionally bad?

I think that what elite players do best is they are able to get a beat on their opponents. They understand when their opponents are adjusting and readjusting, and anticipate it.

I'm sorry if this is incoherent. I'm a bit tired and my thoughts are a bit muddled.

Hold on, you see that this guy is capable of bluffing and now you think he's LESS likely to be bluffing? You haven't seen if he's even capable of not bluffing. And we could equally take it a level further than you did, what if he knows that you know that he's seen you call down light, so surely he must be bluffing with 100% of his range in this instance.

All of the issues you list seem like stronger and stronger arguments for game theory and balance. When you do manage to get a grip on these issues for a specific player than of course you should adjust and attempt to maximally exploit him. Or if you have a read on how the general population at a particular stake plays then you could adjust your play to exploit that tendency, you may be exploited by some players but should win more on average. But if you're just guessing about what his range is then you don't really have any expectation to it. In fact, if villain was playing a balanced strategy himself and not skewed in any direction then you'd actually be losing expectation with your approach, no matter which way you guessed.