Board-pairing turn with SPR 1
Posted by DirtyD
Posted by
DirtyD
posted in
Mid Stakes
Board-pairing turn with SPR 1
BN: $115.90
SB: $216.53
BB: $197.00 (Hero)
UTG: $261.18
Having a hard time wrapping my head around this turn spot. I make a somewhat light pre-flop call (let's skip past that, please) and flop top pair + wrap. I think the flop check-raise is pretty standard, but I'm very uncertain on the turn. BB should have a large range advantage, but it's not clear to me that betting accomplishes much with any part of our range. The board has become very static, so very often we either A) have a boat that needs no protection B) are up against a boat and drawing dead. Blocker effects are very important here: if we have the boat-y cards it's much more likely our opponent has the draw-y cards, and vice versa. I think this probably points towards a check or a small bet over a shove. So...
Do you have a betting range on this turn? If so, what size? Does it make any sense to have a shoving range?
Loading 12 Comments...
I think a small bet (say 25-35% of pot) makes sense here. At least from what I see, players tend to call check-raises fairly wide with draws, so you should still get enough folds with a small bet. You may know your player pool better, for more precise judgement.
If you were somehow guaranteed an opponent who just checks it down, then you'd prefer not to bet. I don't think shoving makes sense with enough of the range to make it meaningful.
Why is betting better than check-calling?
Because you get to choose the bet size? Because you get to bluff when betting but can't bluff when calling?
It's not a bad idea to check-call either though. I think either way it's going to be very dependent on opponent tendencies.
What I'm thinking is, with our actual hand, we're in a way ahead/way behind situation, because we're drawing dead against boats, but need very little protection against draws. We'll also have many boats, which are in a similar situation but even more extreme. Blocker effects are extremely important here: if we have a boat, it's much more likely our opponent has a draw, and if we have the draw, it's much more likely they have the boat. So I think much of our range wants to check: boats because they don't need protection and can't get action with a bet, and wraps for fear of running into boats. I'm sort of talking myself into checking, but I'm still confused by the whole situation.
You can be way behind for sure, but I'm not sure you have a way of being way ahead. If draws fold to a bet, they should be folding our a decent chunk of equity. You could argue that they wouldn't fold, but that's really hard to know without a range context.
One thing to note is that the strategic difference between a check and a small bet may not be all that large.
Another thing: you say blocker effects are extremely important. Have you actually quantified it? If you haven't, you may actually be surprised by how blocker effects work in PLO sometimes. I can't say exactly how it will play out in this situation because it depends on the range context on the turn, but it's not immediately obvious to me that blockers should be the reason for why you check this hand. (if at all)
If you're really interested in clarity here, then I do think you need to dig deeper. If you can come up with turn starting ranges for both players, then it's easier to say meaningful things about strategies and range interaction. If you also want to define turn action ranges, then that's even better. Given starting ranges, I'll help from my end with what I can.
Personally I do think it might be worthwhile to spend some time on a spot like this, because it should give you valuable information on how you want to construct x/r ranges on the flop.
For example, we have 82.5% turn equity against QJT9, which is a very strong draw on the flop. Of course there's some distance between 82 and 100, but I'd still count that as way ahead. Our exact hand is somewhat unusual in that it's sort of a "medium" combo hand, but a lot of our range will be more polarized towards boats or draws without much showdown value.
Blocker effects:
Just copy and pasted that from my notes, not sure why I chose K986, but you can see there's a huge difference in how often button has a boat depending on our holding. I expect the difference will be even pronounced if we give ourselves, say, KK32. (I would 3bet most, but not all, KK at this stack depth).
The ranges I was using for the flop action were
K8+,T96+,K+:T9+,8+:965+ for BB
and
965+,K7+,K+:J9+ for BU
These definitely aren't perfect. For one thing, with my small flop check-raise size I am saying I have some bluffs, but the range I used there is basically all strong value/semi-bluff, no real bluffs. So I should probably be including a few bluff combos.
It's also fair to point out that with our exact hand, check-potting is probably a very good and simple way to play the hand. But the turn spot I got myself into is still interesting.
Why things work out the way they do can be somewhat deep depending on how experienced you are with applied game theory. But let's give it a go.
Let's say you were to play a strategy where you check your entire range, and BN uses only shoves or x/b. This should be your worst option, as many of the hands that you could use as bluffs would now have to fold. Specifically your KT96 would have to fold in this strategy.
If you were to choose a strategy where you also get to pot lead some of your hands, you'll do much better. Now some of your hands that would otherwise have to x/f, will get to bluff and do much better. This strategy should capture a fair bit more of the pot, and your hand KT96 could fit quite well into a shoving range.
So far straightforward? Now the trickier part is why a small bet strategy does even better. Intuitively, the way it works is that your range has become too strong on this turn. You'll do well with your shoves, but that lets the BN get away with folding correctly. With a smaller bet size, he's forced to continue more often, and yet also has to fold some because your range is so strong. However, when he continues, he'll still be behind, and as such you get him to put in even more $$ bad into the pot. I think your KT96 hand can be played as a b/c or a b/f, but less likely as a x/r actually.
If you want traditional terms, the small bet has some parts blocking bet in it, some parts protection bet, some parts value bet and some parts bluff. But the old school terminology used for holdem isn't really adequate to explain what's going on in PLO, so I generally prefer not to use those terms myself.
Maybe a picture helps explain it better?
I've used $FI50 for BN and 65%!$3B8O for BB as pf ranges. This picture tells me that you need to have a betting strategy, and that you should be able to capture quite a bit more than your fair share of the pot.
It also tells me that something probably went wrong on an earlier street. You're supposed to be strong here, but I don't know if you're supposed to be this strong. And knowing that BN folds 2/3 of his c-bets to a x/r (?) you should probably want another x/r strategy on the flop, no?
Anyway, if my calculations are correct, then you should be able to find a strategy that captures most of the pot here. The best bet size appears to be something like 24 or so.
Thanks for the detailed reply. If I understand correctly, you're saying
check range < check/shove < bet small
How did you arrive at your conclusions here? Logic, software, experience with similar spots, something else?
No, that's not what I'm saying. Not quite that simple I'm afraid. If your ordering refers to strategies, then it's too simple. And if it refers to how to play your actual hand, then I specifically believe that x/r would be less likely with your actual hand.
But looking at just one single hand doesn't really allow for a lot of clarity. Then you could probably argue for just about any way to play the hand, but still not really know anything more in the end. How to play your actual hand depends strongly on the strategic context that it's in. It depends on opponent tendencies as well, which is something we haven't even mentioned yet.
I mix of all I suppose. Maybe the easiest way is to calculate EVs for a few different strategies for both players. I did this partially for a few strategies, and started seeing indications in this direction. I stopped early because you probably want to adjust the flop play before going deeper into calculations. But I recognize the same things from other situations, both real poker situations and solved toy games. I also set up a couple of toy examples that resemble this situation, that I then solved. But I also have some experience looking at equity distributions and seeing how strategies should play out, so just the plot actually says a lot as well.
Some of this relates to what we do in Elite PLO Fundamentals (strategy construction, applied game theory), and the Advanced course (range construction). Most of the tools and methods I've developed are for those courses.
The best way to gain more clarity here could be for you to calculate EVs for some of the strategies that I've suggested, and see how the values change when you change ranges and parameters.
Thanks, I will take a shot at calculating some equities. Was that range vs. range equity graph you posted from Odds Oracle?
As far as calculating EVs of strategies - it's quite easy to calculate EV of simple strategies, like BB shoves range on turn, but somewhat more difficult and requiring more assumptions to calculate EVs of complex strategies over multiple streets, right? (because we'll have to make assumptions about how each side plays their range)
No, but you can get something similar with the graph function in OO.
Right. That would indeed become more of an issue with higher SPR and more streets left to play. For this example I made a quick approximation that overestimates BN equities when play goes to the river -- but it still doesn't seem like that becomes a significant part of the strategy. But check for yourself with your own assumptions. Esp. with different flop ranges I expect checking back, or calling a small lead bet could become more commonly used actions.
There's some work involved even for one street when allowing for one round of raises, but not too bad. You need to define four subranges for each player, the frequencies of those ranges, plus five range vs range equity matchups. If you want to calculate the next street as well you double that for every card dealt, but you can usually cluster cards into just a few groups.
You could also make an argument for multiple bet sizes for BB, but I don't think that would change EVs all that much. BN would still only want one bet size I expect.
Be the first to add a comment