What counts as a short stack in a poker tournament?

Anything under about 25 big blinds, where postflop play stops being your main lever. Below 25bb a single preflop decision commits a large share of your stack, and below 10bb almost every hand you play goes in before the flop.

At what stack size should you switch to push or fold?

Around 10 big blinds and below. At that depth a min-raise leaves you no room to fold to a jam, so open-shoving captures the same folds while removing every awkward spot after the flop. Between 10bb and 18bb you shove from late position but can still raise-fold from early seats.

How wide should you open-shove 10 big blinds from the button?

Very wide. Any pair, any ace, any two broadway cards and the better suited connectors are all comfortable jams with only the blinds left to get through. From UTG at a full table the same stack shoves a small fraction of that: strong aces, big broadway and pairs.

Does ICM make you shove tighter or wider?

Both, depending on who covers whom. ICM tightens your calling range sharply because busting costs more equity than the chips you win are worth. It also lets you shove wider into opponents under the same pressure, since they have to fold hands that would call for chips.

Why do antes make short stacks shove more hands?

Antes inflate the pot you are attacking without increasing what you risk. With a big blind ante the preflop pot is 2.5bb instead of 1.5bb, so a successful 10bb shove wins two thirds more, and the break-even fold frequency drops enough to add a real chunk of hands.

Short Stack MTT Strategy Under 25 Big Blinds

Under 25 big blinds most of poker disappears. You are not floating flops or planning three-street bluffs — you are choosing between an open-shove, a jam over someone else's raise, and a fold, and your stack decides which of the three is even on the menu. The thresholds are sharp: 25bb plays nothing like 15bb, and 15bb plays nothing like 8bb. Learn where the lines sit and a short stack becomes a weapon rather than a countdown.

Your stack picks the move, your hand only qualifies

Deep-stacked poker rewards hand reading. Short-stacked poker rewards arithmetic. With 60bb you can open A♠J♦, take a flop and work the rest out later. With 14bb the same open leaves you 11.8bb behind with no answer to a 3-bet, so the hand has to be worth committing before you touch a chip.

The most expensive habit at these depths is treating a short stack like a deep one with smaller numbers: opening 2.2x at 12bb, folding to the reshove, repeating. Each cycle burns nearly a fifth of your stack without a showdown. Three of them and you are in shove-or-die territory holding the same cards you started with.

Choose the action your stack allows, then ask which hands qualify for it. Picking the hand first is how short stacks die.

That order matters because the qualifying hands swing enormously with depth. 8♠7♠ is a fine 12bb reshove over a button open and an awful 20bb call of the same raise. The hand did not change; the action available to it did.

The stack ladder

Measure everything in effective stacks — the smaller of your stack and the one you are up against — and count it before the blinds hit you, not after.

Effective stack	Default mode	What you give up
Under 10bb	Open-shove or fold	Nothing worth keeping
10-15bb	Shove from late position, raise-fold from early	Most postflop play
15-18bb	Open small, reshove over other opens	Multi-street planning
18-25bb	Full raise-fold, 3-bet jam as your main aggression	Deep-stack manoeuvring

The bands overlap in practice, and the right one shifts with table dynamics. A 16bb stack at a table of callers behaves like 12bb, because your fold equity is worse and your shoves get looked up. A 12bb stack against three nits who have all folded twelve hands in a row behaves like 16bb.

Under 10 big blinds: shove or fold

At single-digit stacks, raising smaller than all-in achieves nothing you want. You cannot fold to a jam without setting fire to a third of your chips, so a raise is a shove with extra steps and less folding equity.

Open-shoving buys two things at once: every fold you would have got from a raise, plus the pot when your hand simply wins at showdown. That combination is why the strategy holds up against opponents who know exactly what you are doing. Even the worst hand keeps real equity once it is called, so fold equity on a shove does not have to be anywhere near total for the jam to profit.

Ranges widen sharply as you move round the table, because each player left to act is another chance to be called:

UTG at a full table: pairs, strong aces, big broadway. A tight opening frame, not a desperate one.
CO and BTN: any pair, any ace, any two broadway cards and the better suited connectors down to 65s.
SB with the BB alone behind: close to half your hands, and wider still against a folder.

10 to 18 big blinds: the reshove band

This is the most profitable zone in tournament poker and the one most players handle worst. You are too short to defend a raise-fold line comfortably, but deep enough that your all-in genuinely hurts the raiser.

When a 40bb button opens to 2.2bb and you jam 15bb from the big blind, they are risking 12.8 more to win a pot of about 18. They need a real hand, and most opening ranges are not made of real hands. Your reshove wins immediately far more often than your cards deserve.

The hands that do this best are not your strongest ones. Suited aces block the calls you fear, small pairs flip against everything that continues, and suited broadway keeps enough equity when called. Hands like KQo and AJo are fine reshoves in position but get dominated by exactly the range that calls, so they belong at the edges rather than the core.

ICM is just the start

Multiway neural solves for tournaments — final table spots solved for your exact stacks, then drilled until they're instinct.

Try Sharkling free

18 to 25 big blinds: raising back into play

Above 18bb the raise-fold returns as a real option. Open to 2.2x, and when a 3-bet arrives you can pass without having committed a crippling share of your stack. You also regain some postflop leverage: a continuation bet on the flop sets up a river shove rather than being one.

What you should not do is start playing 22bb like 50bb. Multi-street bluffing needs a stack that can apply three barrels of pressure, and yours cannot. Your value comes from being the player who can jam over anyone's raise, which means keeping ranges tight enough that the jam is credible and wide enough that it is frequent.

Against opens from short stacks who cover you, the 3-bet jam replaces the flat call almost entirely. Calling with 20bb creates a bloated pot you can neither bet nor abandon.

ICM bends every threshold

Everything above prices chips as chips. In a tournament they are not. Chips you win are worth less per unit than chips you lose, because stack size converts to prize money on a curve that flattens as you climb.

The practical consequence is asymmetric. Your calling ranges tighten hard — hands that are a clear chip-EV call become losses in real money — while your shoving ranges can widen against opponents who are under the same pressure. A big stack jamming into two medium stacks near a pay jump is exploiting exactly this.

Run the numbers rather than guessing. An ICM equity calculator turns your stack, the payout structure and the remaining players into an actual dollar figure, and the difference between chip equity and real equity near a pay jump is usually far bigger than it feels at the table. The same maths drives short stack bubble play, where a 12bb stack with three players until the money should be stealing relentlessly and calling almost nothing.

Antes make every shove better

Once antes are in play, the pot you are attacking grows while the stack you risk does not. With a big blind ante, the preflop pot is 2.5bb instead of 1.5bb, so a 10bb shove that takes it down uncontested wins two thirds more than it used to.

That changes the break-even fold frequency enough to matter. Shoving ranges get meaningfully wider the moment antes appear, and defending ranges widen alongside them because the price to call has improved too. The full picture of how antes change your ranges is worth working through, since most players adjust far too little and keep folding hands that became profitable jams two levels ago.

Turning it into muscle memory

Short-stack play is the most learnable part of the game because the decision tree is tiny. There are three actions and one variable, and the correct answers are stable enough to memorise.

Learn the thresholds by position, not as a single chart. UTG at 12bb and BTN at 12bb are different games.
Count effective stacks before the action reaches you. Deciding at 15bb and acting at 13bb is a different hand.
Drill the spots you get wrong, especially reshoves from the blinds — that is where most of the edge lives.
Check ICM separately whenever a pay jump is close, because the chip-EV answer will mislead you.

Start from a solid set of push or fold charts, then practise the reshove band until it needs no thought. The player who jams the right hands from the small blind at 11bb, every single time, quietly outperforms the one who plays beautiful postflop poker and gets short-stacked badly.