On July 16, 2025, the WSOP Main Event resumed with four players left, and Kenny Hallaert had about eight big blinds. On the first hand he moved all in with ace-queen. Michael Mizrachi called from the big blind with king-jack and caught a jack on the turn. Hallaert went out in fourth for $3,000,000. Nobody who studies short-stack play thinks he made a mistake. At eight big blinds, ace-queen is a shove from almost any seat, and he was a clear favorite when the chips went in.
That hand sums up the short stack’s reality. You make the right move, you are ahead, and you still need to win the race. What you can control is how many good spots you take before the blinds take your stack away.
The basics of push/fold play, and why deep stacks should fold so much early, are covered in big stack vs short stack strategy for the early levels. This piece starts where that one stops: the middle and late stages, when antes are in, most of the field is short and the money is close.
What is a short stack, late in the event
A short stack is measured in big blinds, never in chips. Fifty thousand chips is huge at 200/400 and desperate at 5,000/10,000. Late in a tournament the useful bands are tighter than early on. Around 20 to 25 big blinds you still have a raise-fold game. Around 12 to 20 you can open small, shove over other players’ raises and pick up pots without seeing a flop. Under about 10 you are in push-or-fold territory, and every orbit costs a real share of what you have.
The other number that matters is the effective stack: the smaller of your stack and your opponent’s. A 15-big-blind player facing a 60-big-blind opener is playing a 15-big-blind hand. Against a covered short stack, the big stack’s size is mostly irrelevant.
Why the ante changes the maths
Most major tours now use a big blind ante, in which the player in the big blind posts an ante equal to the big blind for the whole table. Before any cards are dealt the pot holds 2.5 big blinds: half a blind, one blind and one ante.
That dead money is the short stack’s best friend.
Take a 10-big-blind shove from the button. You risk 10 to win 2.5. If every opponent folded with no thought for your cards, you would need them to fold four times in five to break even on the steal alone. They do not need to fold that often, because you have equity when called: even a hand like king-nine offsuit often has around a third of the pot against the strong hands that call. Add the dead money and the fold equity together and a wide range of shoves becomes profitable from late position.
You will notice that strong players at 10 to 15 big blinds rarely limp and rarely open-raise small and fold. Each of those actions gives away the dead money without a fight.
The re-shove: the short stack’s best weapon
The most profitable play in the 12-to-25-big-blind range is not the open shove. It is the re-shove over an opener. Late-position players open wide when the antes are in. When a cutoff or button player raises to 2 or 2.2 big blinds, the pot is already 4.5 to 5 big blinds before you act. Moving all in for 18 puts them to a decision with most of their range, and most of their range has to fold.
The hands that work best here are not always the strongest ones. Ace-five suited, small pairs and suited broadways do well because they have reasonable equity when called and block some of the hands that call. Pure trash does not, and premium pairs are often better flatted or shoved for value against a range that will call. Software such as HoldemResources Calculator and ICMIZER, and the range work described in the guide to poker solvers for beginners, have made these re-shove ranges well known among regulars. That also means that against a strong field, opens from late position get tighter once short stacks sit behind.
Position matters more when you are short
A short stack under the gun is in a bad spot: eight players behind, any one of whom can wake up with a hand. On the button only the blinds remain. For the same stack, the button shoving range is several times wider than the under-the-gun range. If you are getting short with the blinds approaching, it often pays to look for one good spot from late position before you are forced to post.
The bubble and the pay jump
Near the money, chips change value. Losing your stack on the bubble costs you the min-cash; doubling up does not double your prize equity. The Independent Chip Model (ICM) puts numbers on this, and the practical result for short stacks is two rules that pull in opposite directions.
When you are the one shoving, the bubble helps you. Medium stacks who can cash comfortably will fold hands they would call with at any other time, so your shoves get through more often. When you are the one calling, it hurts you. Calling a big stack’s shove with a marginal hand risks your entire tournament for a small gain, especially when another short stack is close to busting.
The same logic returns at every big pay jump and above all at final tables, where the gap between places can be larger than the whole min-cash. That is when a short stack can sometimes fold into a higher payout while others clash. It is also why a double-up late in an event, the stack jump from 8 to 17 big blinds, changes your options far more than the chips suggest: it moves you from shove-or-fold back into the re-shove game.
Common short-stack leaks
The first is waiting too long. A player at 12 big blinds who folds three orbits waiting for a premium hand reaches the big blind with 7 and no fold equity left. The second is the small open with nothing behind it: raising to 2 big blinds with 9 left and then folding to a shove. The third is calling off light against a covered opponent near the bubble.
A fourth leak shows up in poker series with re-entry. Players short on Day 1 sometimes gamble wildly because they can buy back in. That changes the maths for them, but it does not make a bad call good, and once registration closes the option disappears.
Hallaert’s hand from 2025 is the right model for the end of this list. Eight big blinds, ace-queen, all in on the first hand of the day. He did everything a short stack can do and still finished fourth, a result that paid $3,000,000 and remains the best of his career.