Short Stack Poker: The Push/Fold Math You’re Missing

Short stacking gets filed under “desperation,” and that filing is the whole problem. A player with twelve big blinds doesn’t have a bad hand to play; they have a different game to play, one with fewer decisions in it and much less room to be wrong. The decisions are so few, in fact, that the correct strategy can be written down in full. Nobody can hand you a complete solution for a 100 big blind cash game. For a 9 big blind button, somebody already did, and it fits on one page.

What makes that page useless to most players isn’t that it’s hard to memorize. It’s that the arithmetic underneath it runs against instinct in three specific places, and the instinct wins. Players shove too tight when they still have fold equity, too loose when they don’t, and they call all-ins with hands they’d never have pushed themselves. Then there’s the bubble, where every number on the chart quietly changes and almost nobody recalculates.

The chart doesn’t care how long you’ve been card-dead. The chips in the middle before you act are the entire reason the play works, and those chips are there whether your last twenty hands were playable or not.

The orbit clock that tells you when the game has already changed

Push/fold isn’t a mood. It’s a stack depth. Here’s the clock that gets you there.

In a typical modern tournament structure, with a big blind ante — one player posts an extra blind-sized ante for the whole table, which has been standard in most live and online events since the late 2010s — you pay 2.5 big blinds per orbit. That’s half a blind in the small, one in the big, one for the ante. At nine-handed, an orbit is nine hands.

So a 15 big blind stack is six orbits from zero. A 10 big blind stack is four. If you’re sitting at 12 and planning to “wait for a spot,” understand that waiting costs 2.5 every nine hands, and the levels are going up underneath you. The player who folds a 12 big blind button and then shoves 7 from the cutoff two orbits later made the same decision twice and got a worse price on it the second time.

Why the big blind calls so much wider than you expect

Start with the price the big blind is getting, because that price is what sets your shoving range.

You open-shove 10 big blinds from the button. The pot already holds 1.5 blinds of dead money, plus 1 more if there’s a big blind ante, so call it 2.5. The big blind has posted 1 of that. To call, they put in 9 more to win the 12.5 sitting there. That’s 9 into a pot of 21.5, which means they need to win about 42% of the time to break even.

Forty-two percent is nothing. Ace-four offsuit has more than that against a wide button range. So does king-nine, jack-ten suited, pocket deuces. The big blind is not calling you light because they’re loose. They’re calling because the math tells them to, and if you shove 10 blinds expecting folds every time, you’ve misread the situation badly.

Here’s the flip side, and it’s the number people skip. You don’t need folds every time. You need them often enough.

A ten big blind shove, worked out in chips

Blinds 500/1,000 with a 1,000 big blind ante. You have 10,000 on the button. Folded to you. You shove.

Say the small and big blind together fold 65% of the time. You collect the 2,500 that’s already out there. That’s 0.65 × 2,500 = +1,625 chips.

The other 35% of the time you get called, and let’s say you average 38% equity against the calling range. The pot when called is roughly 21,500: your 10,000, the big blind’s 10,000, which already includes the blind they posted, plus the small blind’s 500 and the 1,000 ante. Your share of that is about 8,170. You put in 10,000. So you lose about 1,830 chips in those branches, and 0.35 × 1,830 = -640.

Net: about +1,000 chips on a shove that feels like a coin flip you didn’t want. That’s a 10% gain on your stack, risked on a hand you’d muck without a thought at 60 blinds deep.

Now run the same hand with a passive alternative. You fold. You still pay 2,500 over the next orbit. The shove was worth roughly 1,000 and the fold was worth zero, and the gap between them compounds every orbit you decline it. Small pairs are the clearest case of a hand whose whole character flips with depth, since a hand like pocket fives goes from a set-mining speculation to a premium shove without changing a single card.

The bubble arithmetic that makes a winning shove a losing one

Chips won and money won are not the same thing, and the gap is at its widest right before a pay jump. The short version: in a tournament, the chips you win are worth less per chip than the chips you already have, because you can only cash once.

Four players left. Prize pool $10,000, paying $5,000, $3,000, $2,000. You have 8 big blinds, the other three have 30 each. Your chip share is 8.2%, which in pure chip terms is $816 of the pool.

Your actual equity is higher than that. Run the standard tournament equity model, the one that turns stack sizes into dollar expectations by estimating your odds of finishing in each spot, and you’re worth about $1,073. The short stack always holds more than its chip share, because the big stacks can’t all win first.

Now double up. You go to 16, the opponent you beat drops to 22. Your equity becomes about $1,879.

Look at what just happened. You risked $1,073 to gain $806. In chips you doubled. In money you went up 75%. For an all-in with no chance of a fold, you now need to win about 57% of the time just to break even. A hand with 55% equity, a clear snap-call in chips, loses you money.

That’s the whole bubble effect in one calculation. It isn’t a vague instruction to tighten up. It’s a shift in the break-even point from 50% to 57%, and it reverses specific, nameable hands.

The effect runs the other way too. The big stacks in that example are worth less than their chip share, which is exactly why they can shove into you profitably with hands that are behind your calling range. Their risk is cheap. Yours isn’t.

Satellites push this to the limit. If the top ten finishers all win the same seat and eleven players remain, a chip is worth almost nothing to you and your tournament life is worth everything. Folding aces preflop with a comfortable stack is correct there, not clever. It’s the only play that doesn’t cost money.

Why your calling range should be tighter than your shoving range

This is the most common technical error in short stack play, and it comes from treating “shove” and “call a shove” as the same strength question.

They aren’t. When you shove, you win the pot immediately a large share of the time. When you call, that never happens. Every chip you commit has to be justified by the cards alone.

Concretely: king-nine offsuit is a fine 10 big blind button shove. Facing a 10 big blind shove from a tight player under the gun, king-nine offsuit is a fast fold, even getting 42%. The hands that beat you are the hands they have. A nine kicker against a range of aces, kings, ace-king and middle pairs is drawing to a miracle.

The reverse error exists too. People fold ace-three in the big blind for 9 more into 12.5 because it “looks weak.” Against a late-position shover, it’s a clear call. Weakness is relative to the range you’re facing, not to how the hand looks in your hand.

When the chart is wrong

Equilibrium shoving charts assume you’re playing a competent opponent and that your only options are all-in or fold. Both assumptions break.

Above about 15 big blinds, shoving stops dominating. You have room to raise small and fold to a reraise, which keeps your weakest hands alive instead of flipping them into the pot. The reshove, where you move all-in over someone else’s open, is a different calculation again, because their raise added dead money and your effective price improved.

Against a big blind who folds too much, widen hard. Against the player who calls off with any two broadway cards, shove your top hands and fold the junk, because your fold equity is the thing that made the junk profitable and they’ve taken it away.

And at 4 big blinds or less, stop thinking entirely. The big blind is getting better than 2 to 1 and calling with most of their range. You’re shoving for pure equity. Any ace, any pair, any two cards with a face in them, and from the button, honestly, close to anything. Waiting for a better hand at that depth isn’t patience. It’s paying 2.5 blinds an orbit for the privilege of being in the same spot with less ammunition.