You pick up his blind, which is about 10% of your stack with no risk at all.
You're usually behind, and you'll still win a good share of the time. The chart has already weighed both of those, so you don't need to agonize over it.
Pokeriad / Learn / Push/fold chart
Updated November 2026
A push/fold chart tells you which hands to move all-in with when your stack is too short to play normally. Slide your stack below and the grid shows what the small blind should shove and what the big blind should call, from 2 to 20 big blinds.
Find your two cards where their row and column cross. Above the diagonal is suited, below it is offsuit. Empty means never.
Once you're down to about 10 big blinds, a normal raise commits so much of your stack that you may as well put it all in. By 15 big blinds you're getting close to that point.
Moving all-in has one thing a small raise doesn't: your opponent can't re-raise you off your hand, so you get to see all five cards every time he calls.
Because the shover can win two ways. He takes the blinds when you fold, and he can still win at showdown when you call. The caller can only win at showdown.
Share of all hands
Look at what happens as stacks get shorter. Below about 4 big blinds the lines cross and calling becomes wider than shoving, because the caller is getting a great price on the last few chips.
The short stack gets wider. Almost everything is worth a shove at 3 big blinds.
Say you shove 10 big blinds from the small blind. If he folds, you win his 1 big blind. If he calls, you win or lose your whole stack depending on the cards.
The chart finds the point where neither player can do better by changing what he does. Your shove has to be good enough against the hands he'd call with, and his call has to be good enough against the hands you'd shove with.
Each one depends on the other, so the only way to get it is to let a computer work it out.
The brighter the square, the deeper it still shoves
Far fewer hands stay bright
It covers one situation: everyone has folded to the small blind, and the two blinds have the same stack.
From earlier seats you need a tighter range, because more players still get to wake up with a big hand behind you. Poker positions shows how fast that adds up.
It's the smaller of the two stacks, because that's the most either player can win or lose in the hand. If you have 30 big blinds and he has 10, you're playing a 10 big blind hand, and that's the number to use on the chart.
A lot. Most tournaments now use a big blind ante, which adds another big blind to the pot before anyone acts, and that's more to win for the same risk. We solved the game again with it.
Stacks in that chart are what's left after the ante goes in. If your tournament has antes, you should be shoving noticeably wider than the grid at the top shows.
We solved the two blinds, and the grid only applies there. From earlier seats you need a tighter range, because more players get to look at their cards.
Those numbers treat each player's call as separate, so they're a rough guide. They're enough to show why a shove that's fine from the small blind can be a mistake from the first seat.
Because tournament chips aren't money. The chips you could lose are worth more than the chips you could win. The usual way to measure this is the Independent Chip Model, which turns stacks into a share of the prize pool.
Say three players are left with equal stacks, and the prizes are $50, $30 and $20. Each stack is worth $33.33. You go all-in against one of them.
+$10.00−$13.33You stand to lose more than you stand to win, so a coin flip isn't enough. You'd need to win this all-in 57% of the time just to break even.
You pick up his blind, which is about 10% of your stack with no risk at all.
You're usually behind, and you'll still win a good share of the time. The chart has already weighed both of those, so you don't need to agonize over it.
What you do need is the discipline to shove with weak-looking hands when the chart says so. Players who wait for a premium hand at 9 big blinds get blinded away.
It's often the best play available to a short stack. Someone opens to 2.2 big blinds from the cutoff and you shove 12 big blinds from the big blind. You risk 11 to win 3.7, so you'd need folds about 75% of the time if you could never win at showdown.
You can also win at showdown, so the real number is well below that. Choose hands that do fine when called: pairs, big aces and suited broadways.
The bubble is the point in a tournament where one more elimination means everyone left gets paid. Going broke near it hurts a medium stack the most, so big stacks can raise a lot of hands and medium stacks should fold more than the chart says.
With 15 to 20 big blinds a small raise can be fine, and good players use it. Under about 10, it's usually a mistake: if anyone shoves over you, you'll have to call with most hands anyway, and you've given them the chance to choose.
Learn it in a few simple pieces. At 10 big blinds, the small blind shoves every pair, every ace, and every king. Queens need a kicker of 7 or better when offsuit.
Calling is much tighter, so learn that second. At 15 big blinds the big blind still calls with every pair and every ace, and then only with the better kings.
Heads-up is the closest match, since it's the same two seats. Cash games rarely get this short, and the rake makes loose calls more expensive, so use the chart there only for the rare very short stack.
The big blind already has 1 big blind in the pot, so calling costs less than it looks. At 10 big blinds you pay 9 more to win the 10 he shoved and your own blind back.
That's the same pot odds sum as any other call.
Twelve spots in three rounds, focused on hands near the edge of the chart.
Rarely. Cash game stacks are usually much deeper, and the rake changes the numbers.
An ace blocks his best calling hands and often wins by itself at showdown. Connected cards need to make something to win.
Putting all your chips in. People also say jam, push or move all-in.
The math doesn't. What changes is your opponents: at low stakes people call too much, so shove fewer bluffs and keep shoving your value hands.
Good players do it sometimes from the small blind. It's harder to play well, so learn shoving first.
We solved the game with our own solver, using fictitious play: each side keeps adjusting to the other over 4,000 rounds at every stack size from 1.5 to 20 big blinds. Each matchup between two hands is measured by dealing 4,000 random boards.
A hand is shown as a shove at a given stack when the solver plays it at least half the time. At 10 big blinds it shoves 57.5% of hands and at 20 big blinds 40.3%.
Send them this page, or save the chart for your next tournament.