Hi skates,
In your video you distinguish between two 3bet shoving strategies around 20ish BB effective stacks. (Actually, at which stacks do you recommend starting to 3bet shove vs different players?)
Anyway, Type 1 guys mostly play push/fold oop at these stacks and rarely flat oop. Thus, they're shoving sort of a wide value range. You say that 2p2 seems to recommend this strategy. Otoh, Type 2 guys polarize their shoving range and flat oop w/ hands that play better postflop (say, QTs).
I'm a Type 1 guy, mostly because the math is easier to do, I've done it, and so I'm sure I can play that strategy more or less perfectly. I'm pretty sure I play well against it, too. The edges are pretty small either way if both players are doin it right. However, I'm not sure if there are greater edges to be had by keeping a flatting range at these stacks. I'd imagine that broadway hands, especially, do great flatted oop vs a SB opening range but pretty medicore when incorporated in a 3bet shipping range.
What do you think about whether or not to have a significant oop flatting range? Have you done much math on it? How much does it depend on specific stacks? What villain tendencies sway the decision one way or the other? Finally, what hands in particular do you tend to work into your flatting range, and what consequences does that have for your 3betting strategy?
Thanks!
Will
Okay. So, in short, I think you can get away with not having a significant OOP flatting range between 10 and 25 BB at all stakes up until 550+ (or if you get into wars with regs, perhaps a little sooner than that). Against a "standard" villain and some winning regs, the equity difference between playing a mixed strategy and a jam/fold one can be pretty small, but that edge quickly diminishes against stronger regs as they adjust (off-hand, if you play this way, I'm probably going to raise ~85% of my buttons because I don't think you're going to jam wide enough, and I'm going to be raise/calling a heck of a lot to boot).
I have probably done more math on this subject than anyone else, and I wrote a program to solve the equilibriums for this game (similar to the jam/fold game, I have the small blind either folds, raise/folds, or raise/calls and the big blind either folds to an open or jams over an open), and I varied it across a number of open raise sizes. I also solved equilibriums for fixing the open percentage and solving from there (useful if you're the BB and you're playing someone who never limps). I did this a little bit before my first big upswing in 2008 (maybe in April or so?) and was satisfied with it then, but I've since opened up my game and very few opponents I play now are so linear so the simplified game isn't as effective.
It is very difficult to calculate equity for mixed strategies because there's no good function to discuss post-flop playability. I've tried functions that give the SB an adjusted edge based on the two respective hand strengths, but nothing looks how I want it to. Someone needs to write a function that discusses a hand's playability in and out of position with varying stack sizes, because right now there's no effective way to do it. I'm too busy working on other things to really want to sit down and work it out when the edge it would provide over my approximations is minimal. I will say though, I have inputted someone's PF stats into a program before to come up with a maximally exploitive strategy.
If I'm in the BB, I virtually never play jam/fold over their opens anymore. The only time I would is if villain is opening to 3x or more, and even then, if that's their standard, I might merge a jamming range and polarize a flatting range (I know that sounds backwards, but I'd be flatting with intent to donk or c/r, so having low cards and AA in range is nice since you win on low card boards, and most people think you flat Kx/Qx so you win on those boards too).
My 3betting strat is totally dependent on villain's opening frequency and whether or not villain limps. If they limp alot, the raises are polar, so I'm raising pretty much only for value and flatting the rest. If they raise a lot and limp sometimes, the limps are polar, so I'm jamming polar and flatting the rest. To be fair though, I really just jam a ton. If they play raise/fold pf, I'm jamming polar, flatting the rest. As for specific hands... I call enough hands to have a distribution that works because I don't want to have to forfeit low card boards. I'm not that happy writing this but against most people I actually have less Tx/Jx in my range than 6x/7x/8x because I'm already able to win on Tx/Jx boards (if villain knows I jam polar, then they think my flatting range smacks up Tx/Jx, so they give up easily). I'm more inclined to jam T9s than 87s strictly for postflop playability reasons with respect to my entire range.
Thanks for the question, ask more if I'm unclear or there's more you want to talk about.
Sick response. Thanks Skates!
So, I also wrote a program a couple weeks ago to solve that same game. Basically, it's made me start 3bet shoving a lot wider, but it's also made me realize that the edges aren't huge in that game, so there is probably some profit to be had by thinking about an oop flatting range.
To that end, I've also thought about using some sort of weighted equity calculation to include a BB flatting range into the approximate game, but I haven't actually tried anything. I think the best way would be to just do it empirically. Filter your database for BB defenses and write down %won for each hand. Might have to do some sort of averaging/smoothing over "nearby" hands if you don't have a large enough sample, but it should work. We would ofc have to empirically estimate the total pot size for each hand, also, to do any equity calculations.
The only thing is that, I think our maximally exploitative flatting range will probably deviate than the equilibrium one more than the preflop ranges will since people play worse/more varied postflop. And it seems like the good ranges might change a lot depending on opponent. As an extreme example, if villain is v passive postflop and good about checking down his air, then A6o plays pretty well in a flatting range whereas you probably just want to ship and forget versus most guys. In addition to changes in hands' equity due to changes in playability due to opponent style, we'd also have to think about changes in average postflop potsized due to opponent style, and these corrections will probably be large.
I think I'll probably keep playing jam/fold in this spot, for a while at least, in order to get a better feel for different opponent tendencies. I especially hope to nail down opponents' 3bet shove-calling ranges. I think that will have as much effect as anything on how we play oop.
I do have a couple remaining questions regarding how you actually construct your flatting ranges versus different guys (last 2 paragraphs of your response), but I'll think about those more and post later.
On a slightly related note, I've noticed from railing livb that, at these stacks, he does 3bet shove a lot, but he also 3bets 3x as well as, say, 2.3x kind of a lot. Do you have any idea why he might do this or how he might construct those ranges?
Thanks again for your high-content response.
Will
I have an idea of what livb is doing there having played him probably about 50 times. I'm not gonna talk about it though. You could try asking him, he's actually pretty friendly if you're respectful.
Also yea you make a really good point. The postflop playability of each hand changes with respect to villain's postflop frequencies and stack depth. You're right that against someone who cbets infrequently, A6o plays pretty darn well, so it'd be hard to write something as a be-all end-all.