How Is ICM Calculated in Poker?
ICM is calculated from finish probabilities times payouts. Here's the exact formula, a full three-player worked example, and the calculators pros use.
On this page · 4 sections
ICM is calculated by working out each player’s probability of finishing in every paid position, multiplying those probabilities by the prize for each spot, and adding them up. The starting point is simple: your chance of finishing first equals your stack divided by the total chips in play. From there the model branches through second, third, and so on.
The core formula
The model rests on one assumption: each chip is equally likely to be the last one standing, so your odds of winning equal your share of all chips. Formally:
P(finish 1st) = your stack ÷ total chips
For the next places, you “remove” the winner and repeat the same logic on the remaining stacks. P(you finish 2nd) is the sum, over every opponent who could finish first, of the chance they win times your chance of being top of the leftovers. Your ICM value is then:
Equity = P(1st)×Prize₁ + P(2nd)×Prize₂ + P(3rd)×Prize₃ + …
Worked example: three players left
Three players remain. The prize pool pays $50 / $30 / $20 for 1st / 2nd / 3rd. Stacks:
| Player | Chips | Chip % |
|---|---|---|
| A | 5,000 | 50% |
| B | 3,000 | 30% |
| C | 2,000 | 20% |
Total = 10,000 chips. Let’s calculate Player A’s equity step by step.
Step 1 — first place. A’s stack is 50% of all chips, so:
P(A 1st) = 5,000 ÷ 10,000 = 0.50
Step 2 — second place. A finishes 2nd only if B or C finishes 1st and then A is the larger of the two survivors.
- If B wins (P = 3,000/10,000 = 0.30): remaining are A 5,000 and C 2,000 (total 7,000). A is 2nd with probability 5,000/7,000 ≈ 0.714. Contribution: 0.30 × 0.714 = 0.214.
- If C wins (P = 2,000/10,000 = 0.20): remaining are A 5,000 and B 3,000 (total 8,000). A is 2nd with probability 5,000/8,000 = 0.625. Contribution: 0.20 × 0.625 = 0.125.
P(A 2nd) = 0.214 + 0.125 = 0.339
Step 3 — third place. Probabilities must sum to 1, so:
P(A 3rd) = 1 − 0.50 − 0.339 = 0.161
Step 4 — multiply by prizes:
Equity = 0.50×$50 + 0.339×$30 + 0.161×$20
= $25.00 + $10.17 + $3.22 = $38.39
The full table
Run the same process for B and C and you get every player’s real-money value:
| Player | Chips | Chip % | ICM equity | Equity % |
|---|---|---|---|---|
| A | 5,000 | 50% | $38.39 | 38.4% |
| B | 3,000 | 30% | $32.75 | 32.8% |
| C | 2,000 | 20% | $28.86 | 28.9% |
| Total | 10,000 | 100% | $100.00 | 100% |
The headline result: Player A owns 50% of the chips but only 38% of the money. The big stack’s chips are “overvalued” by raw count, while the short stack (20% of chips, 29% of the money) is “undervalued.” That’s diminishing chip value made concrete — and it’s exactly why a chip leader can’t bully recklessly and a short stack isn’t as doomed as the chip count suggests.
Doing it for real
The branches multiply fast — four players means twenty-four orderings, and full tables are impractical by hand. In practice you:
- Estimate at the table. First-place equity (chips ÷ total) plus a feel for the pay jumps is usually enough for a decision.
- Verify with a calculator. ICM calculators and trainers crunch the exact numbers; many also show the equity swing of a specific all-in, which is what you actually care about.
- Learn the swings, not the digits. What matters is “how much equity do I gain vs. risk,” not the third decimal place.
If the probability-times-payout logic feels familiar, it’s the same expected-value thinking behind poker odds and math. For where these numbers change real decisions, see what ICM is and ICM deal making, or return to the ICM hub.
Frequently asked
What is the ICM formula?
ICM assigns each player a probability of finishing in each paid position, then multiplies those probabilities by the prize for each spot and sums them. The chance of finishing first equals your stack divided by total chips in play.
What is ICM value in poker?
Your ICM value is the real-money equity of your chip stack — your expected share of the remaining prize pool right now if everyone agreed to stop and pay out by the model.
Can you calculate ICM in your head?
Only roughly. The first-place probability (your chips ÷ total chips) is easy, but the second- and third-place branches multiply quickly. Most players estimate at the table and verify with an ICM calculator afterward.
Why doesn't ICM consider skill or position?
ICM is a deliberate simplification — it assumes every chip is equally likely to win, so finish odds depend only on stack size. Skill matters in reality, but ignoring it keeps the model usable and still highly accurate near pay jumps.