Small Blind Play 11-14bb Deep: Raise, Openshove, Fold, or Limp?
In the last article, we described the best approach above 8bb from the small blind as making use of the ability to limp, minraise/call, and minraise/fold. Let’s go into a bit more depth about those decisions and talk about how you can use data to maximize your expectation at this stack depth.
When we are 12bb deep, I estimate that a random opponent will call an openshove with around the range of [22+, A2+, K8o+, K6s+, QJo, Q9s+]. Certainly, there are some opponents who will call with K7o, K5s, QTo, and Q8s, if not even wider. Similarly, there are opponents who would fold K8o, K6s, QJo, and Q9s for 12bb, if not even tighter. Still, from 25,000 games of experience, this is my best guess at a standard villain calling range at this stack depth. When looking at openshoving, the question NASH asks is "how deep can I jam in equilibrium so that going all-in is better than folding?”. However, with the vast majority of hands that we might consider going all-in with 11-14bb deep, folding is rarely an option. When we have T7s, Q8o, J9s, 98o, 54s, and those type of borderline hands, if we decide not to go all-in, limping or minraising will likely be what we choose instead – not folding. Because of that, looking at a NASH chart can be a pretty poor approximation above 8bb.
NASH tells us that jamming is better than folding with 87o up to 14.7bb deep in shove-or-fold equilibrium that does not allow the small blind to limp or minraise – but does this mean we should actually go all-in with 87o that deep? The most important piece of missing information has to do with our expectation from minraising and limping. Luckily, after all those games I bragged about having under my belt in the last paragraph, I have developed a strong sample of results to lean on. Historically, my expectation from limping and minraising marginal hands around 12bb deep has been much stronger than folding – in general, it lies between +0.1bb/hand and -0.2bb/hand for most holdings, with folding the small blind resulting in -0.5bb/hand. While an average does not encapsulate all of the different opponent tendencies we could be playing against, especially given that if I am at all decent at poker, my sample should be biased towards opponents where making this play has better expectation than usual, it does illustrate the point of how bad a comparison the expectation of folding is and give us a different benchmark to evaluate the EV of going all-in against.
So when do we jam and when do we do something else? To help us in answering this question, I have made a table. It shows the expectation of openshoving with each hand 12bb deep against this calling range. 0EV means that if we start the hand with 12bb, our expectation is to end the hand with 12bb. -0.5bb means that if we start the hand with 12bb, our expectation is to end the hand with 11.5bb. Since that is what happens when we fold, -0.5bb is the same thing as openfolding in this framework. It also gives some example recommendations of what to do with these hands in light of this information.
Let’s do a quick Q&A:
1. Wow! This looks so cool. Is it a new equilibrium? Is it balanced?
Some aspects of our range are definitely balanced – our minraising range, for example, despite having all of our monsters, also has a ton of junk in it. Similarly, while the openjamming range has a lot of suited connectors in it, it also has a good amount of Ax and pocket pairs. The biggest thing that makes this not close to an equilibrium is our limping range – it's all stuff that's folding to a shove. That's obviously highly exploitable. However, remember one of the most important axioms of poker: If you want to exploit your opponent, you have to be exploitable. Maximum money comes from exploitable strategies. Force your opponent to adjust and start jamming over your limps wide, and then re-adjust in response to that. Notably, though, this strategy only calls for limping once every 6 buttons or so – your opponent is unlikely to feel like he's getting limped to death.
2. Why are some hands openshoves and other hands limps or minraise/calls, despite having the exact same expectation when jamming?
54s and K2s are very different hands than J7s and Q9o, despite having the same expectation when going all-in. The latter duo plays much better in limped pots, whereas K2s and 54s don't have the luxury of that alternative. It all goes back to the main point: We should compare the expectation of plays against the expectation of the next best option. That's why looking to NASH for most of these decisions can lead you astray.
3. Are you sure about these ranges as standard? I would play a certain hand differently readless. Am I wrong?
No, you're not categorically wrong. I made this table six months ago and have changed my mind about a lot of hands since then. There is a lot of wiggle room for a lot of these hands – if you want to limp JTo or minraise/fold 95s, I would not argue too strenuously with you (in fact, I would probably agree on both cases now). However, if you think openjamming K3o, Q2s, or 97o is best readless 12bb deep, there are strong reasons to believe it is not (despite NASH’s advice to jam). There are actually always right and wrong answers, but we should care most about the ones that are clearly right and clearly wrong. For example, if you think limping with 53s is best readless instead of investing the extra big blind in hopes of a fold, I would strongly disagree with that conclusion.
4. Are you sure the openjamming range isn't too weak? It seems like there are a lot of weak hands in that range and no big hands. Can't our opponent exploit us by calling wider?
It is a little weak – the big blind is correct to open up his calling range wider in response to this jamming range. That is actually what you will see a lot of at the nosebleeds – people call openshoves pretty wide 11-14bb deep because of the ranges they represent. If you feel like your opponent has picked up on how much you are exploiting his calling range and starts to open up, there is one pretty easy adjustment you can make: Start openjamming your strong Ax hands as well. Those hands have rather mediocre inducing value especially considering that the big blind’s expanded playing range will rarely be dominated, and the value from openjamming goes way up if your opponent is calling jams wide.
5. It seems like a bunch of small differences. Are you sure this even really matters that much? Aren't there more important things to be worrying about?
Keep in mind that NASH is -EV against the NASH calling range from this stack depth. People should still have positive expectation using most strategies, but I think my expectation using this type of approach has been exceptionally good. In my $100,000 month, my EV adjusted winnings from the small blind 10-15bb deep were $27,000. Lifetime, my EV has been 8bb/100 over a huge sample. I am not sure what numbers other players are accomplishing, but I think those make for some pretty strong proof that the expectation from having ranges like this can be really good.
It all goes back to the Ed Miller article about meaningful errors – sure, it can be hard to get motivated about improving in certain situations by 0.2 big blinds. However, because these situations constantly come up in endgame play, your leaks get magnified over time, and you lose a lot of money by not playing optimally.
6. How does your approach change as stack depths get shorter?
The expectation of your marginal hands goes up significantly and your opponent will jam wider over limps, so more hands become openshoves. You can and should still limp hands that have mediocre jamming expectation but play decently postflop, such as Q5o, T7o, and J7o 10bb deep. In general, people are willing to get it in wider from the big blind, so minraise/folding becomes less attractive as well, although I think you absolutely should have a minraise/folding range against most opponents 10bb deep. Getting flatted becomes an even worse proposition with more Ax hands, so those hands can become openshoves as well. Basically, the ranges squeeze tighter and start to more closely approximate the shove or fold equilibrium that is +EV for the small blind 8bb deep or less.
7. How does your approach change as stack depths get deeper?
Against opponents who 3-bet jam appropriately wide, 15bb can be a great stack depth for limping, given how much the big blind has to risk to jam over them. More hands on the openshove list become limps or raise/folds. The better the hand flops, the more likely it is to be a limp. T7s, for example, is a great hand to limp 15bb deep. Some hands that were strong enough to raise/call, like K7o, aren't strong enough against most opponents anymore, and become limps or minraise/folds. Low Ax and weak pocket pairs are still great jamming hands even 15bb deep. Against players who are not willing to reshove wide, it is important to keep pounding your opponent with minraises and not get too limp happy. Just because it feels less fun to minraise/fold 98s does not mean that you should not do it if your opponent is 3-bet jamming infrequently.
Throughout all this advice, keep your attention on the foundations that we’ve previously outlined. There are no charts to follow to maximize your expectation – doing so will depend on how well you react to the tendencies of your opponents and how willing you are to think through games rather than autopilot them. All that said, there is a massive opportunity for significant edge in the endgame if you make the most of the options of minraising, openshoving, folding, and limping.
