[2-4]: ATs OOP in a 3bet Pot

I probably fold to 3bet.
Flop probably x/r
Turn probably donk.

id fold. no FE, no implieds, no odds.

vs a wide 3bettor 4b pre has to be better than calling. make it like $100 pre, 1/3 pot any flop, jam most turns.

Reveal:

I was also debating the 4bet, but ATs just seemed a little too strong for that, I'd rather 4bet something crappier like a low suited Ace. I think this guy has 3bet bluffs, so the difference between A5s and ATs is pretty significant.

And again yea folding is very much my default play here, it's like a 90% thing, I almost never flat 3bets, especially OOP.

I guess if you want to go straight to the decision point, turn is a shove or fold.

Pre I fold, you said so, move on.

Flop I think may be better as a x/r OOP then continue betting the turn with ♣ maybe 6 / 9 / J for SDs? and T because you hit and it favours your range. I guess not A because that's more his range. Thats 21/47 cards which should be enough?

It's a close hand on all 3 streets.

Preflop is close enough to 0EV that you being 1 seat earlier or villain being exploitably passive could nudge it into at least a mixed fold, but I'd pure continue here.

Can't go wrong mixing 4b at these stack depths OOP with any hand that pure continues. It's a majority call with this particular combo though.

Flop is very close between all 4 options: lead flop, x/c, x/f, or x/r. Any reads on villain's bet sizing tendencies would be hugely helpful; a lot of people will overfold to a raise when they use this size.

Turn is close between all non-fold options: lead, x/c or x/jam. If you feel good about your over-aggression read, then I like x/jam because players like that will make a lot of polarity mistakes and end up folding a ton of equity.

I'm personally not sure what to make of villain though. I'm not really reading anything into the showdown you posted. I think they just called because "I haz blockers," which doesn't make him any better or worse than you had him already tagged.

My read is that V is a solid player for six days, then sacrifices his winnings to the poker gods on Sunday by spewing. Why is V there? Usually a straddle means: to have fun. These Vs are super hard to read. You think they’re playing ABC and then you see a hand where they are FOS.

I like raiseannounced analysis that all streets are in fact close. I just think that if the ev is close to 0, you should wait for a better spot. Maybe also focus on weaker opponents. Fold pre.

I like the turn shove as played, but I also think you fold two other times in this hand. Like adonson said, you don’t need to mix it up out of position against this guy with this kind of hand. And if you do, why aren’t you 4betting?

pre i guess is ok, is close ev wise so its really going to come down to who is going to play better post. if you never really play w him i don't think you need to really defend most of the 0ev continues if you're unfamiliar with what to do post as it's really really easy to both under realize and make large mistakes oop in 3b pots imo. flop is ok, turn is good. meh. think you will gain extremely little by posting hands vs midstakes regs on this forum as opposed to alternative study methods (passive aggressive submersible strikes again!).

i would not have a leading range on either street as i dont think it makes sense range wise and will weaken your checking ranges in a spot you are already at a polarity deficit. id imagine people would respond poorly vs it one way or another but i dont have a strong opinion on how they would react and would try to just stay in standard node

fwiw q6 looks good in solver world.

I know I'm supposed to mostly gloss over this but I actually care a ton about what kind of time investment is useful vs. not useful for poker (theoretically way more important as anything about this particular hand!). So actually I'm just gonna take this comment extremely seriously instead.

So, uh... why? To explain my current model, right now my relevant high-level thoughts about poker are something like

• the game is actually pretty easy compared to basically any other competitive discipline
  
• ... but, this is balanced by most people spending almost all their improvement time inefficiently (& playing itself not really helping you improve)
  
• because of this, it's extremely important to make sure there is an actual connection between what you study and your play; most time "improving" (such as looking at random GTO charts you can never memorize anyway) will do nothing because you will forget about it before you'll ever get to apply it
  
• Even if you can remember it, GTO is only marginally useful for live play; it's much more important to understand general player tendencies and play accordingly
  

Based on the above, my usefulness ranking for things-you-could-do-with-the-stated-goal-of-actually-improving-your-game is something like (from worst to best):

... and that's why I spend about 80% of my time-trying-to-improve nowadays on the last two (mostly w/ a weekly hand-review session but also somewhat through this forum). The last two items here are so different because anything that arises from a hand you played automatically has a connection to your play, so you're not gonna study it and then forget it (at least I don't, I tend to remember past lessons about my hands), unlike imho most of "studying".

So I don't get why you think [making this submission is not useful] at all. I guess I didn't really misplay this hand? But I've submitted maybe a dozen hands so far and this was the only one I didn't misplay, and I don't know ahead of time. Usually I select the hands that I misplayed the most from any one week. (I don't explicitly select for that, but I select for "how much uncertainty do i have about the play here", which is pretty similar.) In this case uncertainty was high because I almost never flat 3bets as I mentioned, so it's just a situation I'm not in a lot (low SPR pot but not as the aggressor).

Maybe you just think GTO is a lot more important? But if so, why? To the extent that I've played a solver line here, it's not because I've studied the solver, it's because I know what line I would fold against the most (& I thought this villain was strong enough that this is relevant) so I played that line hoping to get folds from someone else. And that seems to me like a more relevant reason then whether it's GTO approved or not.

Also I think the most interesting hand I've ever play... (even though tbqh I don't think the responses I got were very good, but what I took away from it was still upsteam of posting it. In fact other responses were very useful because they were evidence that the line would in fact work in practice against most winning players.)

And the other thing that makes me feel somewhat weird about this advice from you is that you kinda mocked my treatment of a previous hand, and specifically time usage, for what I thought was in the last category on my list (taking a deep dive into a particular aspect of your hands), and even in retrospect I took a tangible lesson from this that idk how else I could have gotten it. You mentioned GTO but idk how a solver would have helped there since I know Villain doesn't study GTO, so his decisions aren't anchored to the solver, so I wouldn't put much stock into anything a solver says. So ordinarily I wouldn't be super inclined to listen to you, except I feel since then you've clearly made some of the best comments on my threads, so actually I probably should listen to you? Idk.

To engage with this seriously, I think this is wrong and probably what sub is getting at here.

Particularly where the case is HU against a player where you have no specific reads or presume them decent (i.e. can't obviously bucket them as a fish who majorly deviates from equilibrium in a specific, exploitable direction), GTO play provides a solid, unexploitable baseline. It's what you need to play when you're completely in the dark because it's never -EV overall (or more -EV than rake) to play GTO no matter what your opponent does.

I'm not necessarily advocating for "memorizing GTO" but if you're trying to look over a hand where you were HU against a decent player, the obvious starting point is going to be asking what GTO play would be. Learn heuristics for what solver play looks like, and note cases where the solver does counter-intuitive things that you can't imagine a human opponent doing at the table because that's where exploits are available even against better opponents.

I'm a Hungry Horse aficionado and I was amused by Marc's recent segment on playing GTO. He likes to mock GTO play and being a "solver nerd" but ironically (and as he acknowledges) you see that most of his strategy is GTO, just focusing on the right pieces of it rather than "properly randomizing" or making sure you know which blocking and unblocking suits are part of a mix, etc. Like I think Marc knows, the most useful part of studying GTO is understanding where and when it's very difficult for humans to follow - and in fact he has lots of asides in his videos where he will pull up Pio and justify strategies by showing the GTO response to a node-locked deviation of realistic player tendencies.

I hope you guys aren't clamoring to get replaced by AI in your day jobs as much as you're encouraging OP to study GTO instead of coming to the forum for advice!

Keep at it OP. I study with solvers all the time but I find in depth discussions of hands most valuable as well.

Using solvers for study is more analogous to using the worldwide web (Google, Wikipedia, etc) for your day job.

Using AI to replace human inputs in your day job is more analogous to botting (or maybe real-time assistance), which I assume everyone is against.

The number I have in my head is 5 posters, so now I'm worried I don't make the cut :'(

I have three thoughts on the study question:

1) I think poker study is a lot like physical activity where the most effective method is the one you actually do, and the one you actually do is probably something you have some intrinsic interest in. Doing x might be worth 1.25 times as much as doing y, but if you can sustain 10 times as much focus when working on y, then you're going to get 8 times as much gain from doing that.

I take more of an Atomic Habits approach to stuff I hate: just so long as I hit some very attainable milestone with that stuff every month, then I can patch up my weaknesses 1% at a time.

2) I think there's a bit of a study paradox where you kind of have to do a lot of study to even find out where valuable study time can go.

20% of my study has definitely accounted for at least 80% of my gains, but I don't think I could have realistically predicted where that 20% of study would have been without the other 80% of study.

3) I agree that there's pretty useless forms of study

The list of useful things I can think of off the top of my head:

Developing heuristics
Pure GTO study (charts, equilibrium sims, etc)
Nodelocking (opponents)
Nodelocking (yourself)
MDA study
Database review (individual opponents)
Database review (yourself)
Drilling
Your own personal math, spreadsheets, scripting, HUDs, etc

If you can be a guru at any of the above, you'll be a winner, and then you can keep polishing up the stuff you hate doing to improve 1bb/100 at a time.

HH study with the right audience (like a study group or your coach) can probably be on par with the above. It's probably especially valuable in the early stages when there's such a huge universe of unknown-unknowns that starting with spots that happened in the real world where you are lost will help ground you. You might get diminishing returns after that as you find yourself only flagging hands that are close so there's low return on studying the spot.

HH study on here is definitely somewhere higher than railbirding vlogs or whatever, but also definitely lower than all of the above.

Then again, posting on here is literally how you got to the point that you're reading this list, so this goes back to #2.

A lot of egos showed up on this thread, but I did find this useful.

Grunch:

PRE - did you consider 4B'ing?

I'm serious. If you think he's capable of 3B'ing wide, ATs might make it into our light 4B range, especially when we're starting out at a stack depth that makes it harder for him to set-mine profitably.

Like, if we 4B to 115 off 438, he's not getting the correct odds to set-mine.

Otherwise, we end up playing a middling suited ace out of position. We may flop a 2nd best TP or a draw, but either way it'll be hard for us to realize our equity.

FLOP - did you consider check-raising?

I'm serious. This isn't his board when he 3B's, and our hand has some decent properties that will allow us to barrel across a lot of turns. Any club or any 6, 9, J, Q or K will increase our equity.

TURN - I wouldn't raise now. And I think it's okay to fold. We're drawing, and don't really know how many of our outs are clean.

i think there's value in the forum for some things, this just isn't a spot where i think you're going to gain much by talking to people who don't look at solvers / play vs good opponents lol. you have guys applying the socratic method or trying to deflect the question bc they dunno the answers or how to find them.

maybe im fish but ive found the most bang for buck in terms of time in videos bc its easier to copy other peoples hueristics than use my own (though have kissed so many frogs to find my metaphorical princes). in terms of my own studying ive spent by far the most time in nl looking at solver solutions and oddly the most time in plo drilling. i very rarely nodelock anything and have spent embarassingly little time with mda solutions but i also don't really play open games so idk how relevant they would be

i hope the uncertainty eats at you from the inside every day

final follow up maybe.

i think either trying to solve toy games on your own or probably much more accessibly for most people, watching people do toy games and extrapolating takeaways from that (i think the most public facing ones are some old sauce videos on rio) are useful in general. but alas most people dont want to lose the magic of some old school wild west game of out thinking their opponents. has also helped me substantially to watch some live tell content (the actual content been kind of whatever, but just putting more intention on timing and live tells has helped winrate quite a bit). another big one has been playing other poker games in terms of understanding poker is a large strategy game (it also helps all poker games are more or less the same, this was a seiver quote that i didn't understand when i initially heard it but have started to wrap my head around lately) and to not take things nearly as personally. tbh some podcasts / interviews have helped but usually not super efficient method of improving

theres also really no way to get better quickly than to find whatever zoom pool is open to your market - gg, stars, acr, and just grind 100k hands in a month. there is honor and dignity in banging your head against the small / microstakes wall and unless you're hopeless you should be able to figure things out relatively quickly

bro, what? 😰

new premise for the shining sequel just dropped

I can’t imagine having to play catchup because I only got 3k hands in yesterday

ask yourself how bad do i want it (a mental illness)

It is a real question! And one I genuinely care about a lot. I mean the question of how much GTO to use (and to study) will matter for pretty much every day I play for the rest of my career.

You didn't address the crux of what I said though. I agree that I could just look at it in a solver, and that this would go much quicker. The thing I'm doubting is how much I can take away from that. My view right now is still "very little".

My point estimate on the 97o hand where I built the sim was that flatting is 80$ higher EV than 3betting. If you now look at a solver and it says that flatting is actually 40$ worse than 3betting, then my point estimate... would probably be unchanged. Maybe I would adjust it down by 10. But wouldn't make much of a difference. Villain in that hand was a very weak player, so I just don't see why you think GTO is a useful anchoring point. Whereas after my sim, I did ajust my point estimate down (by about 100$).

Like, okay, let's get theoretical about this. GTO does two different things:

- calculate equity (as in, figure out how often a certain range wins on a certain board). <- This is exactly the same as Equilab also does, and you could even do it by hand exactly (although it would be tedious)
- backtrack this information to find actions at equilibrium, including with equilibrium ranges. <- this is completely impractical to do with equilab, or any other tool that isn't a solver

In case of the 97o hand, mostly we were holding specific actions for hero fixed (although there was some ambiguity about River play for the flatting line). That means the main difference between analyzing with a solver and analyzing with a simulation is that the solver will use equilibrium strategy for Villain on the River, whereas the simulation used a hard-coded strategy for Villain.

So I think the key question is, how much difference does this make? And my guess is that it turns the result from almost zero practical relevance to extremely high practical relevance. Because Villain has never looked at a GTO chart in their live, so there really is no reason to assume that GTO is a relevant anchor for them, at all. There is no reason to think that "give them an equilbirum strategy and then adjust from that strategy" will work better than "just give them a range that never considers equilibrium in the first place". And I think there's also no reason (or at least I don't see it) to think that "just give them an equilibrium range and then don't adjust" is anywhere close to actual result given their actual strategy. I have no idea whether GTO!EV it will miss the actual EV by 20, 50, 100, or 200. None of that would surprise me. So idk what update I'm supposed to take away from the solver EV.

I guess you can look at how the solver plays and if the solver plays reasonably similar to villain in that particular spot, then you can conclude that the EV probably won't be off by all that much. I think this was mostly the case in the 63 hand. So maybe because using a solver is so little time investment, it would probably would have made sense to check it first before doing anything just in case. But if the strategy is very different from what GTO says, which i think will be the case most of the time, then yeah I'm back at thinking the result just isn't that informative.

I mean, think about it this way. Am I going to be perfect at figuring out how Villain plays? Of course not. But will I be better at it than GTO, which doesn't try to do this at all? It would be really strange if I wasn't. Like again, I just don't see why you expect the GTO strategy to be a reasonable estimate if the player has never studied GTO and GTO is famously unintuitive.

And this is the central situation against most people. It's not like in other sport where you eventually want to get so good that you only want to play against strong players. In poker, you never want to play against strong players. I'm hoping that in five years I'm still playing mostly against people who have never looked at a solver in their life.

I'm super interested in possible disagreements here, I just don't think you've really explained them yet. Most of the time I talk about GTO with people they talk about it as if it being a useful anchor point is trivial / doesn't need to be justified.

(Gonna post the rest separately because I think there's several strands in this conversation, but "how useful is GTO" is probably the most important one, so will try to keep them separate. )