Prize Ladder Protocol: Online ICM Calculator Tutorial
June 2, 2026

Tournament chips climb a prize ladder—not a cash slope. Malmuth-Harville ICM math, full 3-handed worked example with verified equity, bubble factor, calculator workflow, deal-check steps, and model limits. No invented solver push/fold charts.
- ICM calculator
- independent chip model
- tournament poker
- bubble factor
- Malmuth-Harville
- final table
- prize pool equity
In cash games, chips sit on a flat floor—1,000 chips = $1,000. In tournaments, chips climb a prize ladder: each rung pays differently, you cannot convert stacks to cash at will, and losing a rung costs more than gaining one helps. An ICM calculator maps your stack to $EV (dollar equity) on that ladder—how much prize money your chips represent right now, given payouts and every remaining stack.
This tutorial teaches the Malmuth-Harville model behind online ICM tools, a full hand-calculated example, bubble-factor decision math, and a step-by-step calculator workflow—without fake solver push/fold percentages.
Why Chips ≠ Dollars on the Ladder
Three forces bend chip value:
| Force | Effect |
|---|---|
| Ceiling | You cannot win more than 1st-place prize |
| Floor | Zero chips = zero additional prize |
| Concavity | Each marginal chip adds less $EV than the last |
Core insight: chips lost hurt more than chips gained help. Going from 0 → 1,000 chips saves your tournament life (huge $EV). Going from 100,000 → 101,000 barely moves your finish distribution.
Sources: Wikipedia ICM / Malmuth-Harville, TryBluff ICM tutorial, Poker Exploit ICM math.
Malmuth-Harville Model (What Calculators Run)
David Harville (1973) published finish-order probability math for racing; Mason Malmuth adapted it for poker (1987). Assumptions:
- Each player's chance of finishing 1st = their stack / total chips.
- Conditional on who finished 1st, 2nd-place probabilities re-normalize over remaining stacks.
- Repeat recursively for all positions.
- $EV_i = Σ (P(finish k) × Prize_k) for each player.
P(1st) for player i:
P_i(1st) = s_i / S where S = total chips in play.
P(2nd) requires summing over every possible 1st-place finisher j:
P_i(2nd) = Σ_{j≠i} P_j(1st) × s_i / (S − s_j)
No single closed-form equation exists for N players—calculators iterate (exact for small N, Monte Carlo for large fields).
Full 3-Handed Example (Hand-Calculated, Verified)
Prize structure: 1st $50, 2nd $30, 3rd $20 (total $100).
Stacks: A = 5,000 (50%), B = 3,000 (30%), C = 2,000 (20%). Total = 10,000 chips.
Step 1 — P(1st)
| Player | P(1st) |
|---|---|
| A | 5,000/10,000 = 0.50 |
| B | 0.30 |
| C | 0.20 |
Step 2 — P(2nd)
A: 0.30 × (5,000/7,000) + 0.20 × (5,000/8,000) = 0.2143 + 0.1250 = 0.3393
B: 0.50 × (3,000/5,000) + 0.20 × (3,000/8,000) = 0.30 + 0.075 = 0.3750
C: 0.50 × (2,000/5,000) + 0.30 × (2,000/7,000) = 0.20 + 0.0857 = 0.2857
Step 3 — P(3rd) = 1 − P(1st) − P(2nd)
| Player | P(3rd) |
|---|---|
| A | 0.1607 |
| B | 0.3250 |
| C | 0.5143 |
Step 4 — $EV (ICM Equity)
A: 0.50×$50 + 0.3393×$30 + 0.1607×$20 = $25.00 + $10.18 + $3.21 = $38.39
B: 0.30×$50 + 0.3750×$30 + 0.3250×$20 = $15.00 + $11.25 + $6.50 = $32.75
C: 0.20×$50 + 0.2857×$30 + 0.5143×$20 = $10.00 + $8.57 + $10.29 = $28.86
Check: $38.39 + $32.75 + $28.86 = $100.00 ✓
| Player | Chip % | Naive chop ($) | ICM $EV | ICM % of pool |
|---|---|---|---|---|
| A (5k) | 50% | $50.00 | $38.39 | 38.4% |
| B (3k) | 30% | $30.00 | $32.75 | 32.8% |
| C (2k) | 20% | $20.00 | $28.86 | 28.9% |
Player A holds half the chips but only 38.4% of prize equity. Player C holds 20% of chips but 28.9% of equity. Chip-chop deals subsidize the chip leader—ICM corrects this.
TryBluff's 3-handed $300 example shows the same compression: chip-leader ICM ~$115 vs naive $150 chop on $300 pool.
Online Calculator Workflow (5 Steps)
Any standard ICM tool (ICMizer-class, free web calculators, or built-in tournament tools) follows the same input pattern:
| Step | Input | Notes |
|---|---|---|
| 1 | Remaining stacks | Every player still in—exclude busted |
| 2 | Prize structure | Exact payouts per place (include deals already locked) |
| 3 | Calculate | Tool runs Malmuth-Harville (or FGS if selected) |
| 4 | Read $EV per player | Compare to deal offers or decision spot |
| 5 | Sensitivity | Adjust stacks ±10% to see $EV swing |
Deal check: proposed payment > your ICM $EV → +$EV deal (lock profit). Payment < ICM → decline or renegotiate.
4-player FT example [TryBluff]: stacks 50%/30%/15%/5% on 40/30/20/10 payout—chip leader's 50% stack ≈ 33% of prize equity; 5% short stack lifted toward ~15%. Middle stacks near chip-proportional; extremes compress hardest.
Bubble Factor — Risk Premium on the Ladder
Bubble factor quantifies ICM asymmetry for a specific call:
BF = |Δ$EV if lose| / Δ$EV if win
Required equity (ICM) = BF / (1 + BF)
Example from Poker Exploit:
- Win: +$2.50 $EV
- Lose: −$5.00 $EV
- BF = 5.00/2.50 = 2.0
- Required equity = 2.0/3.0 = 66.7% (not 50% chip-EV breakeven)
| Situation | Typical BF | Required equity |
|---|---|---|
| Early MTT | 1.0–1.1 | ~50–52% |
| Approaching bubble | 1.5–2.5 | ~60–71% |
| On bubble | 2.0–5.0+ | ~67–83% |
| FT big stack | 1.2–1.5 | ~55–60% |
| FT medium stack | 1.5–3.0 | ~60–75% |
Spot: Bubble, 15 left / 14 paid. Hero BB with A♠Q♥ (~57% chip-EV vs CO shove). BF ≈ 2.3 → required = 2.3/3.3 = 69.7%. 57% < 69.7% → ICM fold despite +2.8 BB chip EV.
Online ICM calculators that support before/after stack simulation compute BF directly; otherwise estimate from $EV change in two manual runs.
When to Open the Calculator
| Situation | Why ICM matters |
|---|---|
| Final-table deal/chop | Verify fair split vs chip-chop trap |
| Money bubble | Quantify bust cost vs chip gain |
| Satellite bubble | Survival > chip accumulation; often fold marginal +chipEV |
| Pay jump decisions (FT ladder) | $EV loss from busting 4th vs locking 3rd |
| Short-stack push/fold review | Compare chip Nash to ICM-adjusted ranges (use dedicated push/fold tools for ranges—not guessed %) |
Early tournament: BF ≈ 1.0; play chip-EV. ICM activates near payouts and short-handed tables.
Calculator Limitations (Know Before You Trust)
| Limitation | Impact |
|---|---|
| Equal skill assumption | Model ignores edge; strong players have higher real $EV |
| No blind structure / future hands | Static snapshot; FGS extensions add simulation |
| Malmuth-Harville tail errors | Poor estimates for extreme long-shot positions in huge fields |
| No table dynamics | Position, aggression, and image not modeled |
| Deal psychology | Calculator gives fair $; negotiation is separate |
ICM answers "what are the chips worth?" not "what line maximizes ROI against this player pool?"
Review Checklist
| Step | Question |
|---|---|
| 1 | All remaining stacks entered correctly? |
| 2 | Prize structure matches actual payouts (taxes/deal overrides)? |
| 3 | Your ICM $EV noted before decision? |
| 4 | For calls: BF or before/after $EV computed? |
| 5 | Chip-EV positive but ICM negative? → fold tier |
| 6 | Deal offer compared to ICM, not chip %? |
Quick Reference
P(1st) = stack / total chips
$EV = Σ P(finish k) × Prize_k
BF = |Δ$EV lose| / Δ$EV win
Required equity = BF / (1 + BF)
Fold EV when chip-EV call positive but equity < required
The prize ladder never pays linearly. Online ICM calculators translate stacks into $EV so you stop treating tournament chips like cash-game chips. Run the numbers before you accept the chop, call the bubble shove, or pass the satellite—then log where your instinct disagreed with the ladder. That log is your MTT edge compounder.