Push or Fold Calculator
Shove or fold, worked out from your own numbers rather than a memorised chart. The arithmetic is exact; the equity is simulated and says so.
One hand
The whole chart
169 hands, each simulated separately — it takes a few seconds and the borderline cells stay borderline however long you leave it.
What is computed and what is sampled
- Equity against a range is sampled, not enumerated. One hand against one hand is already 1,712,304 possible boards, and against a whole range it is that many times over — so the figure is a simulation, and the error bound is shown beside it rather than hidden.
- The sampler is checked where exact answers are cheap. With a flop out there are only 990 turn-and-river combinations, and with four cards out only 44 rivers; the tests enumerate those completely and require the sampler to agree. That exercises the dealing, the evaluation and the tie-splitting all at once.
- Ties split the pot rather than going to whoever was listed first. Over a large sample that difference is small; on any individual hand it is the whole result.
- The push-or-fold arithmetic itself is exact. Folding costs the small blind, shoving wins the dead money when they fold, and turns into a race for both stacks when it does not. Only the equity fed into it is a sampled number.
- None of this prices a bubble, a pay jump, or the fact that your opponent is a person who has been watching you. It is chip expected value against a stated calling frequency, which is a narrower claim than a push-or-fold chart usually makes for itself.
How to use
- Set your stack, the ante, and the range you think they call with.
- Check one hand, or build the whole 169-hand chart.
- Read the equity you need against the equity you have.
Frequently asked questions
Why not just use a published push/fold chart?
Because a chart is the answer to one specific question — a particular stack, a particular calling range, a particular ante — and it does not tell you which. Change what you think they call with and the chart changes underneath you. This works it out from the numbers you actually put in.
What is exact here and what is not?
The push-or-fold arithmetic is exact: folding costs the small blind, shoving wins the dead money when they fold, and turns into a race for both stacks when it does not. The EQUITY going into that is a simulation, because one hand against one hand is already 1,712,304 possible boards. The sampling error is stated rather than hidden.
How do you know the simulation is right?
It is checked against exact answers wherever they are cheap. With a flop out there are only 990 turn-and-river combinations left, and with four cards out only 44 rivers; both are enumerated completely in the tests and the sampler is required to agree with them.
Why are some cells marked too close to call?
Because they are. At a few hundred samples per hand there is a couple of points of noise, and a hand whose edge is a hundredth of a big blind cannot be separated from that. Marking it as borderline is honest; giving it a confident verdict would not be.
Why does a shorter stack shove wider?
Because the blinds are worth relatively more. At five big blinds the 1.5 you pick up uncontested is nearly a third of your stack; at thirty it is a twentieth. The equity you need when called falls accordingly, and the tool shows that figure alongside the equity you actually have.
What do antes change?
They add dead money, so a shove that gets through is worth more and the equity it needs when called goes down. That is why shoving ranges widen once antes come in — not because the hands got better, but because the prize for winning uncontested got bigger.
Does this account for the bubble?
No. It is chip expected value against a stated calling frequency, which is a narrower claim than a push-or-fold chart usually makes for itself. Near a pay jump a chip you win is worth less than a chip you lose, and nothing here prices that.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.