superT

Just finished my coaching package with Qing. He does a great job of improving your poker IQ by distilling the reasons for why we play spots a certain way into simple key concepts that you'll quickly notice apply to many other situations as well. The lessons were well structured with no fluff, and the vibe was efficient but also nicely relaxed and positive. Qing was also very available and thorough in answering messages between sessions, both to clarify what we'd discussed but also to share his thoughts on other theory questions. Overall I'm very pleased with the coaching and definitely recommend working with him.

Thanks for the answer. Bluffing AK5x/KQ5x type of hands makes sense. I guess you could put it that the prime blockers add EV to betting more than the bottom pair helps checking, which is basically what you said anyway.

How far do we take it though? Are hands like AK33/AQ77/AKQ8/KQ87 interested in bluffing as well, or does having slightly higher SDV and/or blocking draws make these hands lean towards checking? (though these combos might not bet the turn a lot I guess)

Thanks for the clarification!

It seems that the unpaired hands you look into bet quite often. However, our total river betting freq is still only 48%, so it's not a spot where we fire away super wide overall. Does this mean that we mostly use the unpaired combos as bluffs and lean towards giving up/realizing the small SDV when we hold a PP or a pair to the board, or do weak combos that have a pair bluff often as well?

In the first sim when you look at how flush draw blockers affect river bluffing: are you sure you don't have river buckets set to small in the simulation? In the viewer all combos are grey with no spades highlighted. If that's the case, the results aren't meaningful re that question since FD blockers "don't exist" in the sim on the river.

Done as well, hope I'm not too late!

With no fold equity you would need about 48% eq to raise/gii on the turn. Against a continuing range you seem to have somewhere around 45-50% eq hot and cold, but effectively less than that since he's not jamming everything he continues with but will instead flat and play river with some hands which ofc reduces your EV by quite a lot.

This means that when you raise and he doesn't fold, you're at best breaking even and at worst losing a decent amount relative to the pot (SPR is so high that stacking off with a couple % too little eq is pretty expensive), so the money you'll make on the raise comes from fold equity. Like you said it seems unlikely that you'll have very high feq after he pots it. Haven't looked at the numbers very hard but with meh fold eq raising seems marginally +EV but not a spectacular play. Calling on the other hand seems intuitively very +EV so I think that's indeed the best line.

"I check river because I will have TT and JJ here a lot and I want to check some hands to stop him from having an easy vbet with all his Qx hands."

And then you check/fold? :D

In the 9653cc hand (37:40) you say you're broking on every turn except 7 and 2. I ran a few PJ simulations and it seems that we also have a pretty clear fold on any A/K/Q/J turn except clubs (and Ac is still kinda close). We need a bit over 35% eq to stack off OTT, but we don't ever really have more than 30% on those cards even against overly wide ranges. Calling flop to dodge bad turns still seems to be clearly better than shoving, though, so wp :)

Well in hand 2, assuming that each player has a bluffcatcher 15% of the time, someone will have a bluffcatcher about 39% of the time. Add to that the nutted hands everyone also has in their ranges, as well as some air floats/raises, and I doubt the field will fold the required 50% of the time to bluff for pot without equity. Also it's not like MP's range hits this board hard either, so if he starts betting all of his air, the field will in theory counteradjust and attack his wide and weak betting range with thinner bluffcatches and more raises/floats.

In general though it's ofc possible to have such a range advantage that villain(s) just can't defend enough, for example a tightish 4bet range catching a dry A-high flop.

Some quick thoughts:

1) Having to defend just the bluffcatching part of your range only applies if villain doesn't gain anything from making you fold air (e.g. OTR if villain can just check back and always win vs air). This is not the case here, since if villain checks back his own air, he will sometimes lose the pot to your air when you bluff him, so getting you to fold the flop is preferable for him.

2) I don't like saying the players are "responsible" for defending. Everyone will continue when it's profitable for them, and if they can't do it often enough to stop villain from betting 100% then too bad. That said, different players will be able to defend a different amount of hands. Being close to the bettor is bad since you also have to worry about the guys left to act. If it was say BTN betting into SB+BB, SB would have to play tighter than BB for sure. In these examples the guy first to act has the benefit of having position on the bettor. Not sure how that affects things other than it obv lets you play some amount of extra hands.

Tough spot for sure. Defending oop on this board seems very difficult because of how hard our UTG open/call-range misses it. I don't think we can have a flop raising range unless it's an exploit, since it's so hard for us to have something worth value raising. Not like villain nails this flop either obviously, but a sizable part of his range is AAxx which is quite a monster here (if our range is say $fi15!$4b4, pocket AAAA has 64% equity vs our range otf!).

Leading doesn't make much sense either, so I think we defend the board by check/calling and going from there. This hand def seems worth continuing. We have 46% against a range of $3b6i (and a lot of people prob cbet all of their air here, which might even be fine given how hard we miss?), and having the gs + good backdoor draws is important since we ofc want to be able to hit those draws as well to be less capped on later streets.

Turn/river play is obv gonna be shitty with our range kind of face up as mostly overpairs with some low card stuff in the mix. It's kinda close to polarized ranges in NL I think, though the equities shift more based on turn cards. Intuitively with this hand I think we want to continue mainly when we turn a decent draw and usually give up on blank turns. Would guess we want to start with a check on turn as well since I don't see many cards improving our range hard enough to warrant a lead (though a small HSNL type protection/"f your air hands"-lead could make sense on some cards maybe? this hand might fit well into such a range to realize the gs + set outs + w/e cheaply?).

FWIW I'm really not sure at all about most of the above. Just my initial thoughts, and I could easily have gotten something wrong. Very interested in hearing some other perspectives as well.

Thanks for the input. You could def be right, and another thing in favor of betting is that it avoids awkward river spots if it checks through. The flush cards aren't the only ones where we can get some value though, we still have the GS too as well as nice outs to top two. I'll try to do some rough math tomorrow and see what shows up.

I think you guys are focusing way too much on the times we need to make a decision for stacks on the flop. What I mean is that CO won't have a huge hand super often. If we bet and get raised, yeah that's a sticky spot, but most of the time he will not raise and instead folds so we win/gamble with SB, or CO calls and we can bluff him off later or hit one of our 10000 outs and go from there.

Similarly, if he has a weaker hand and we check, he's probably checking behind quite often. Then they both get a free card to maybe pick up 2pair/flush draw/spike their nut gutshot, and now we're sandwiched ott in a smaller pot where bluffing will probably be harder. Of course we'll still have a really good hand after it checks through but I don't see too much merit in letting that happen.

I feel like I've misread the hand or missed something really obvious. Don't we have the second nut flush draw and a low wrap? How could we check/fold this otf?

Not sure about bet/folding either, think getting it in will be a really minor mistake at best, and if villain ever raise/folds here, bet/jamming should be clearly +EV. If not, we can still maybe call the raise though that might get tricky/not be smart. In any case, I don't think CO will raise us very often at all with SB behind ready to punt the small stack with most any piece of the board, so getting raised isn't something we should be overly concerned about.

There's $2.85 in the pot when it comes back to you ($1.20 from villain and SB, $0.10 from BB and the $0.35 you put in already). Going by your assumptions, when you 4bet, 50% of the time they both fold and you make +$2.85 (compared to folding to the 3bet).

In the other 50% you get the rest of your stack ($9.80) in vs villain for a $21.60 pot where you have 37.7% equity vs PPT top 4%. In that case you get 0.377*$21.6 = $8.143 back from the pot, for a net "gain" of $8.143 - $9.80 = -$1.657 (equity in the pot minus how much you had to put in).

The total EV is then like you said 50%*+$2.85 + 50%*-$1.657 = +$0.60.

The number probably isn't super close to reality as I don't think the assumptions are very realistic, but that's how you do the math :) You're right that adding in villain flatting and SB coming in for the gamble makes it more complicated/tedious, but the idea remains the same.

Any reads on villains?

I might fold this preflop unless the players in position are nitty. On the flop with a monster draw I'd start betting for sure to make them fold whatever. I don't think CO will stab a ton of weak hands on such a wet flop 3way, so not betting yourself means the flop checks through often, letting them pot control/see free cards with hands you could make them fold. Could maybe check/raise if we have a read that CO likes to stab a lot, but check/calling is too weak with such a big hand imo. As to 2), no you should not fold this to a single bet even if he pots it.

On the turn, no way you can ever fold. You are flipping against top set so you definitely have enough equity to continue. If you think he's bet/folding somewhat often here, check/raising turn seems nice to push him off a hand that's doing well vs us, but even if you think he's really strong you can call for sure.

Yeah, that came into my mind as well when I thought about the hand later. The thing is, there are a LOT of non-blank turns, and on those we basically always make the second nuts and villain will have many nuts in his range. That's not a nice situation to be in deep, though I don't expect people to go crazy in these games so we can show it down/play a smaller pot quite often.

Also, while we can put pressure on stuff like Acc or two pair nicely on blank turns, our equity drops quite a lot against the strong parts of his range. I don't think we can raise/gii on a blank turn with an SPR of ~3, so we either semibluff with our blockers and very tainted outs, or call again letting him realise equity and putting us into a marginal river spot nearly always.

Turn float tells how often we bet when PFR cbets flop, we call IP and he checks to us on turn.

Probe bet turn (or something similar) is the % at which we lead the turn after PFR checks back on the flop. Similarly I've understood river probe bet to be when we check/call a flop cbet and then lead river after PFR checks back turn.

Yes, but what I'm disagreeing with here is that EV(call) is necessarily underestimated. I'm not saying it's obviously overestimated either, but having money left to play after calling doesn't only work in our favor via implied odds. It can also benefit villain by allowing him to push us off of decent equity OTT, especially when it goes HU (or get another valuebet in when we have to call it off with low but adequate equity). There's something like 20 turn cards where we have equity in the 21-31% range which we might be forced to fold esp if we were still a bit deeper so villain could bet full pot OTT ("might" because the sidepot could make the better ones still a call). Obviously calling pot OTF to fold say 30% equity OTT is worse than just calling the flop pot bet and realising full equity without further betting. At least to me it's not clear how these effects sum up.

That said, this is mostly nitpicking the model. In reality I can easily see calling being the best here (even though I said fold earlier, dunno, I'm a nit) since according to your numbers it's at least unlikely to be terrible, and BTN seems like a weaker player based on his stack so he could do something funny to gamble in a big pot. In that sense I agree that calling is undervalued in the model and tbh the spazz/mistake factor might easily be the most important variable here and def works in our favor.

I might have misunderstood you earlier. If the question you're looking to answer is roughly "if we decide to call/shove, do we want more or less people in the pot?", and we're comparing the EVs of the same action with different amounts of players in the pot, I have no objections. In this case I agree that the errors at every step are roughly similar and that the results should be useful despite the simplifications.

However, I did get the feeling that you were also comparing EV(call) to EV(jam). I don't agree the model is very meaningful for that, as the errors in each of those figures are definitely not similar. EV(jam) should be pretty much spot on, whereas for calling the model effectively assumes that the stacks are ~70bb shallower, which is obviously a major difference at reasonable stack depths. A quick look at the HU case for example suggests that in reality we have to make quite a few of very close folds/BE-ish calls OTT (assuming villain always jams turn), whereas in the model we can't be forced out of the pot/never have to put in more money behind. I think ignoring this is too big of a simplification for the results to be realistic/comparable.

Again, if I got the idea of your post wrong, my bad and sorry for the ramble :)

- superT

Thank you for sharing this, very well written and concise.

However, I don't really agree with the assumptions. Ignoring future betting when calculating the EV of calling leads to results that can easily be way off, since in reality there is another $70ish left behind that is going in on very nearly any turn (or on the flop if someone ships ofc). The math you did is basically comparing whether we prefer to have a smaller or bigger stack when we gii on the flop, and it should be kind of obvious that we prefer the smaller one since we have odds to go with a smaller stack, but never really have an equity edge so that there'd be value in getting even more money in.

To get the actual EV of calling with stacks left to play we'd need to figure out the EV and probability of different ways the hand plays out after the initial call (we get jammed on otf, we gii ott, turn pairs and we fold etc.). No idea what that amounts to but there could easily be a somewhat significant difference since calling to fold turn is really expensive, and if something like QJxx comes along we only win 1/2 of the pot with a sizable chunk of our outs, which is worse the more there is left to play for. Calling could still easily be the best play, but I'm not sure we can conclude it from this analysis alone.

My intuition said gaaaaamble, but looking at it more closely we need ~42% to get it in assuming it goes HU against the bettor. Against 30%:(KT,KK,TT,88,AQJ,QJ9) we only have about 38%, and I kinda doubt others cram it in with hands that do poorly against us often enough to make stacking off good.

I wonder if we could call though. That way we can fold on board pairs OTT, maybe have weaker hands come along, or get a multiway gamble which should be alright when drawing to a lot of nut outs. The weaker hands that might come along make a split often when we hit though, which is horrible and probably means that we rather want them to fold.

TLDR: I think it's a fold.

Yeah, I get what you mean and we're basically on the same page. Just wanted to point out that even bet/snapfolding is so profitable here that getting raised doesn't matter too much. Checking has to have some definite merits of its own to surpass betting with this much FEQ, even if getting raised puts us in a toughish spot. Basically we might have such a good bluff spot that we can almost give up the marginal value our hand has and use it more profitably as a pure bluff instead. Not saying that we 100% do but intuitively I feel that this is actually the case.

Very nice analysis, thanks for sharing!

However, I think there's a logical mistake when you say "whether to bet or not in the first place, since its so close...". The decision that is close is what to do after villain happens to ch/raise, and this does not imply that betting itself is a close decision.

If villain folds 80% of the time to our $3.4 bet (could go smaller if he really can't continue often, just pulled numbers from the discussion), the EV of bet/folding is ~+$3 even if villain never flats the bet. In a $4.8 pot an EV of $3 is huge, 62.5% of the pot. FWIW, against a $fi15 range our hand only has 54.6% equity. Thus for checking to be better than betting, we'd actually need to do significantly better than realize 100% of our equity after checking back. At least intuitively that sounds very unlikely without rock solid reads, since like you said our hand will be pretty marginal and hard to play on many runouts.

If we're not planning to barrel the guy off Axxx, shouldn't we check on the flop already? Like you said, otf he's pretty much calling Axxx and folding the rest. Obviously we'd rather save the bet against Axxx if we're not getting him off the hand later on, and the hands that fold are in very bad shape against our hand. We can continue on a ton of turns after checkback, and a passive fish isn't too likely to put us in many tough spots later on.

Thanks for the comments guys, seems like everyone is on the same page and I wasn't too off either. As played I did cbet and pot a blank turn, just noticed afterwards that x/raise looks nice as well and wanted to hear if people have a strong opinion either way. Also thanks for the video recommendation midori, going to have to check those out.