Modern Short-Stack Play (Part 2)
For years, poker books, articles and training videos taught us that we can’t flat a whole lot with shallow stacks because we don’t get to realize our equity that often. We still miss the flop around two thirds of the time, and when we don’t hit, we can’t really do anything because we are too shallow to try check-raise bluffing and so on.
The education material always advocated that whenever we have a strong enough hand with a shallow stack, we should just go all-in. That way we’ll always see all five cards and won’t get blown off our equity. While there’s a lot of merit to this strategy, I don’t believe it to be anywhere near optimal from one position specifically: the big blind.
Part 2: Defending the Big Blind with a Very Short Stack
In the first part of this article series, we discussed defending the big blind against a late position open with a resteal stack (around 15-20 big blinds). But where do you draw the line between a resteal stack and a “shove and pray” (meaning that you’re so short that you’ll always get called) stack? The answer, if you ask me, is almost never. I don’t believe there’s really any stack above ~3BB that you need to treat as an all-in or fold stack.
How you should play the big blind against a single open when you have less than 10 big blinds is another great example of how blinded we all were for years. I can remember many occasions from my own career back in the day when I tagged someone as a fish for playing a hand like this:
Poker Tournament on the Internet
Hero is dealt on the Button
Hero raises to 2.2 BB, Big Blind (8BB stack) calls
(Pot: 6 BB) Flop
Big Blind checks, Hero bets 2 BB, Big Blind raises all-in, Hero calls
Big Blind shows and holds.
After the hand, I’d type a note on the villain in question that said something along the lines of: “******** idiot fish who flats 6-5s from the big blind with ****ing 8BB!!!!”. Turns out it was me who was the idiot the whole time. Villain’s flat makes sense for a couple of reasons:
- Being shallow-stacked actually allows Villain to realize more equity because they will see all 5 cards more often.
- It will be easy for Villain to play his hand post-flop with such a shallow stack.
When you have something like 30 big blinds, flatting a hand like 6-5 from the big blind often becomes a bit difficult. So often the flop comes like K-J-5, and you call a c-bet because you have a pair. If the turn is virtually any card besides a five or a six and your opponent bets again, you have to fold. You didn’t get to realize your equity, because you were too deep to see five cards.
But what if you had just five big blinds to start the hand? You’d always get to see all five cards when you hit something by getting it all-in with any piece (you’ll be priced in if you hit almost any piece.) At the same time, you can save your last few BBs by getting away when the flop is A-K-J.
Playing Post-Flop 7.5 Big Blinds Deep
There’s not much room for post-flop maneuvering when your stack gets this short, but there’s still some. Let’s dive into a hand:
Poker Tournament, Blinds 500/1,000/100. Effective Stacks 7,500 (Villain covers)
Hero is dealt in the Big Blind
Villain raises to 2,200 on the Button, Small Blind folds, Hero ???
Let’s estimate Villain’s Button opening range to be 37.3%. It’s best to estimate a slightly tighter opening range than usual because of our short stack. You can give him almost any halfway reasonable range, though, and it won’t change what I think is by far the best option here.
Our three options are:
- Call (we need 20.7% raw equity)
- Go all-in (we have 39.5% against villain’s range, and approximately 0 fold equity)
First, I’d recommend finding out which is better between shoving and folding. If we assume that villain never folds, which is pretty reasonable given that we’re only 7.5 big blinds deep, HoldemResources calculator tells us that shoving makes a tiny profit:
However, I believe that we can make a much bigger profit by calling. Let’s look at a few scenarios on different flop types (Note: After our call, the pot on the flop is 5,800 and we have 5,300 behind):
- Flop #1:
We’ve completely whiffed the flop. We check, our opponent bets and we fold. There are very few hands that we are beating, and even a hand like J-T is a pretty big favorite against us. Whenever our opponent has an ace or a king, we’re basically dead.
We only have 8.33% equity against our opponent’s opening range, so folding isn’t painful at all. We clearly had the odds to call pre-flop with our hand, and now that we whiffed we can muck our hand without thinking twice. Had we gone all-in before the flop, our opponent would’ve called and we’d find ourselves more or less dead on the flop for the last of our chips. In this scenario we saved our last 5,300 chips by not shoving pre-flop.
- Flop #2:
We’ve flopped huge for a 5 BB stack. We could shove ourselves, or check and put the rest of our chips in against a bet of any size.
I don’t think it even matters a whole lot which one we choose. We are 70.75% against our opponent’s range. Sometimes we’ll run into a monster like JJ, but we’d have run into them by going all-in pre-flop anyway. By lead-shoving the flop we’ll accomplish one of two things:
- We’ll force the Button to fold many hands with decent equity against us, like QJ.
- We’ll induce the Button to call with many worse hands, like AK.
It’s almost irrelevant to us what he does with something like K-Qo. If he calls, he only has 22.8% against us, so we’re a big favorite to double. If he folds, then cool – we just won a big pot without showdown and now it’s our opponent who didn’t get to realize his equity!
Also, if we decide to check we may induce a bet/fold from out opponent because it’s a brick flop that people c-bet virtually 100 percent of the time (especially in MTTs, where people generally c-bet too much). When we shove over the bet, he can either fold out his equity or call with a hand that’s behind. That’s a mistake he never would’ve had a chance to make had we gone all-in before the flop.
- Flop #3:
This is probably as close to a hairy-looking spot as it gets with these stacks. We have 34.71% against villain’s opening range. If we check and he goes all-in, we can actually call. We need 32.3% equity, so it’s quite close.
However, being out of position with stacks this short is actually favorable because you get to make the first move and there’s nothing your opponent can do about it. This allows us to play the hand even more profitably.
By going all-in here, the ball is in your opponent’s court and it’s now him who’s facing a tough decision with a lot of his range. I wouldn’t expect anyone to really fold a pair, but even if he always makes the correct call with every 3-3 and A-2 type hand we still aren’t doing all that bad. This is our equity against some of his hands that actually have paired up already:
vs. A-2o: 41.5%
vs. 3-3: 44.2%
vs. T-9s: 22.8%
vs. A-A: 23.3%
Obviously, if we always run into top pair or better we aren’t doing particularly great. But the good news is that it’s hard to make pairs in Texas Hold’em. Remember that, he, too, will have missed the flop most of the time! To be able to figure out if donk-shoving the flop is profitable or not, we need to figure out the following:
- What’s our equity when called?
- How often will he call?
To figure out all that, all we need to do is a simple Flopzilla/Equilab simulation. This is the precise opening range I gave our opponent:
And this is how often he will call on the flop (Note: this is never an exact science, but I wanted to use an example here that’s actually pretty bad for us):
I have the button calling our jam with:
- Every pair (22+, A2o, A5o, A9o, A2s, A5s, A9s, T9o, J9o, Q9o, K9o, 98s, T9s, J9s, Q9s, K9s, K5s, K2s)
- Every gutshot with an overcard (A3o, A4o, A3s, A4s)
- AT+, even without a backdoor flush draw
- Every backdoor flush draw with overcards (from KhTh/KcTc/KdTd down to JhTh/JcTc/JdTd – the dark blue means that he’s calling ¾ of those, only folding the spade combos that don’t have a backdoor flush draw)
I basically only have him folding the flop with stuff that’s completely whiffed, such as K-4s, J-To, and A-7o.
All in all, he’s calling our shove 63.9% of the time. Now, let’s calculate our equity against his continuing range:
We have 32.4% equity. Remember, all we have is a gutshot, yet we still actually have pretty reasonable equity! So, by shoving all-in for 5,300 chips into the 5,800 pot we:
1) Win the pot 36.1% of the time immediately when our opponent folds
2) Still win the pot 32.4% of the time when we do get called.
It’s time to compare our options. If we check-fold the flop, we’re obviously left with 5,300 chips (our remaining stack). But, if we go all-in:
36.1% of the time end up with 11,100 (pot + our stack)
63.9% of the time end up with 5,316 (16,400 total pot * 0.324)
=7,367 chips on average.
Pretty sweet deal, eh? We only flopped a gutshot with no overcards, yet we still profited 0.84 big blinds on the hand (had we folded before the flop, we’d have been left with 6,500)! Remember how much that pre-flop shove would’ve made? Yep, 0.04 big blinds. I think we’re onto something here.
We’ve flopped pretty well and will almost certainly be playing for our stack here, granted we aren’t fist-pumping about it. I think we could either lead-shove or check and let our opponent bet first. If he goes all-in, we call, if he bets anything less than all-in, we still push.
Which of the options is better is actually a fairly interesting question, and depends a lot on how often your opponent c-bets. Without doing the math, I think I actually lean towards checking because that gives our opponent more opportunities to make a mistake. But just to illustrate one last point, let’s look at if we just always donk-shoved this flop. On this texture, our opponent has a pretty easy decision with most of his hands: If he has a pair of eights or better, he’s never folding, and same goes for a flush draw. I’m also giving him credit for being an absolute clairvoyant where he folds every single worse pair and calls with every single better one (and 87s for a chop). This is, essentially, as bleak as it can possibly get for us.
That’s 38.5% of his opening range. In this extremely nasty scenario where he basically soulreads our hole cards, he has a whopping 80.7% equity against us. But let’s add everything up together and see some magic happen:
61.5% of the time he folds, and we end up with 11,100 chips
38.5% of the time he calls, and we end up with 3,165 chips (16,400*0,193)
= 8045 chips on average.
We had our opponent soulread our hand, yet still we made over 1.5 big blinds on average! (And in many cases I’d say that checking could actually make us even a bit more.)
Why Being Out of Position With Super-Shallow Stacks is Great
I almost hope that none of you will actually follow the advice in this article, because dealing with super-short-stack flats and donk-shoves is miserable and annoying.
There are few scenarios more tilting than when you get dealt A-K on the button, open into a short stack ready to take the last of their chips, and then he ****ing flats and lead-shoves a T-8-5 flop. But that’s exactly why it works so well. You see the power of having position being discussed to no end in poker articles, but when you’re super-short, being out of position is actually a big advantage.
When you only have something like 1x pot-sized bet behind, whoever gets to pull the trigger first always gets to put the other one in a nasty spot with a lot of their range. Yet at the same time you get to play your own hand to perfection. You could even just play a very silly strategy of always donk-shoving when you have a pair+ or a draw and check-folding the rest, and there wouldn’t be a damn thing your opponent could do about it. He’d frequently be forced to fold boatloads of equity and a bunch of actually best hands on various textures.
In reality, you can exploit your opponents even more by playing a bit more unbalanced – trapping with your monsters, check-shoving instead of donking against people who c-bet too much, etc. It’s also noteworthy that as funny as it sounds, people just aren’t used to playing 7BB deep post-flop poker. You frequently see people shoving ridiculous hands that have virtually no equity when you check it to them with top pair.
How much exactly they should shove in these scenarios depends on a lot of factors, but the answer is by no means 100 percent, and we only allow our opponent to make those mistakes when we flat pre-flop instead of going all-in.
Ask yourself this question: You have 50 big blinds and open the button with A-4o into a big blind with a stack of 7 big blinds. He calls, and you have 1x pot-sized-bet behind on a flop of T-8-2. Your opponent checks. What do you do?
The answer is: I don’t know. You should mayyybe just shove, because you just have 1xPSB behind anyway, who cares about 5 more big blinds when you have so many. You might have the best hand and you really want to make him fold something like 7-6 that actually has a lot of equity against you. But then again, if the big blind has trapped with, say, T-9o, you’re drawing almost dead. Also, if the big blind had donk-shoved into us, would you call?
As you can see, the big blind has just made our life absolute hell by flatting. An important side note is also that since people virtually always open A-x hands over other non-pair hands (for obvious reasons), it actually makes it pretty hard for them to smash most flops without an ace in them. For example, ~38% of the button opening range we used before has an ace in it. All those A-x hands just really hate their life with awkward stack to pot ratios on two thirds of the flops (especially when you have 1.5x pot or so behind).
How Wide Should We Flat From the Big Blind?
Alright, so we’ve now concluded that we can flat a lot of hands in the big blind, because we can (simplifying it slightly) just play fit or fold after the flop. Should we just flat all of our hands, then? Of course not.
While I don’t by any means hate sometimes also flatting a hand like A-A for deception, I believe we should still just push the vast majority of our hands that are easy pushes. Let’s look at that shoving chart from the previous spot again:
As you can see, shoving a hand like 2-2 or A-4o makes over 1.1 big blinds in our 7.5BB deep example. I would always just jam these, because that’s a huge equity advantage, and I’d hate my life on virtually every flop we didn’t bink. But when you look at those smaller edges like J-8s or K-6o, as well as some losing shoves like 6-5s or Q-8o, I’m pretty convinced that flatting is much better.
I don’t know where exactly to draw the line, but unless you’re somehow really screwing things up post-flop, I’d say that in this example you could easily flat at least every hand that’s between -0.25 and +0.25BB and do better than you’d do by just shoving the +EV ones and folding the rest. I wouldn’t be shocked if you could even flat hands as weak as 8-5s.
More Absurdity: The Really, Really (Really) Short Stacks
The shallower we get, the less likely it becomes that our opponent ever folds either pre or post once he’s opened. Thus, with a stack like 4 big blinds it becomes pretty much impossible to push him off hands post-flop, but that just works in our favor in another way. Since our opponent is never folding regardless, we can actually call and fold on some flops and save money as opposed to just automatically moving all-in with every hand that’s a profitable shove.
Here’s a funny example. Our opponent min-raises on the button at 500/1000/100, and we have 4,000 total in the big blind (3,000 after posting the blind). We can shove almost every hand profitably before the flop (I’m using that same 37.3% opening range again):
But if we just call, do we have to get in every single flop? The answer is no.
On the flop, the pot is 5,400 chips, and when your opponent goes all-in, we’d need to call another 2,000 to win a pot of 9,400 total. That means we need 21.2% equity call. If we look at a hand like 3-2 of spades, we can now happily fold on a flop like QdJdTd, where we only have 8.55% equity against our opponent’s range. As long as we never fold a pair or a draw, it’s basically impossible for us to make a mistake, but we can still save chips on a lot of flops.
It’s still noteworthy that since we can never know for sure how wide our opponent opens, we should still always get it in as long as we flop at least something. The only way to make flatting worse than shoving into 0 fold equity is if we start making wrong folds after the flop ourselves. But since we were getting it in before the flop anyway with our, say, 4 big blinds, it’s never a mistake to push something like 8-7 on J-T-2 and find ourselves a big underdog (in this scenario, we have 25.8%, way more than we’d need). And the more our opponent opens, the less we can fold ourselves.
Basically, I’d just do this to avoid getting it in in the absolute worst scenarios where we are completely dead against virtually anything. It’s still much better than always getting everything in.
How to Counter Shallow Big Blind Defenders
My prediction is that in a couple of years’ time, we’ll see the standard button raise go up to something like 2.5x the big blind again. This would be standard evolution for the game. A few years ago, everyone was under-defending their big blind, meaning that we could open any two on the button and print infinite money doing so. But now that people are getting better and better (and after reading this article, will start defending more), the only way to make people fold more is by opening to a bigger sizing.
This article may have been a lot to digest and possibly a little overwhelming, so I want to finish it with a few quick pointers to summarize:
- When you’re super-short in the big blind, it’s often better to flat and fold the worst flops than to just shove, even if your spot is an any two shove.
- When you have something like 6-10 big blinds, you should flat the big blind wide, because you get to realize more equity than if you were deeper.
- It’s hard to make pairs in hold’em – the good old stop and go actually still works in 2016.
- Being able to make more profit by flatting as opposed to shoving requires you to play correctly after the flop yourself. If you’re unsure, you should err on the side of caution and just shove all hands that are clear shoves, especially those that have a lot of raw equity but don’t connect too well on most flops.
- This is really important: it’s only the really short-stacked spots (~5BB or less), where you’re playing to get in most of your hands and only folding the absolute worst flops. The deeper you get, the more it becomes both a game of pushing your opponent off their equity and selecting your spots carefully. But you also can’t overdo it in either direction, or you just do worse than you would by jamming whatever the calculators suggests. The more big blinds you lose when you jam a gutshot into your opponent’s top pair, the closer it becomes to a mistake. This article is a good start, but I strongly recommend you to use all the available tools out there to do some calculations yourself
The third installment of this series will cover blind vs blind showdowns from the small blind’s point of view — read it here.
As always, I’ll be happy to answer questions in the comments box below, or on Twitter @chuckbasspoker.
Read more from Miikka Anttonen’s Modern Short Stack Play series:
- The Ultimate Guide to Big Blind Defense in MTTs (part 1)
- How to Master Short-Stacked Small Blind Play in Blind vs Blind Battles (part 3)
- Make More Money in Big Blind vs Small Blind Clashes (part 4)