ICM Postflop in Tournaments
In short: Under ICM, chips you lose are worth more money than chips you win, so postflop near the bubble and at the final table the aggressor gains: they bluff more, the player under bust-out pressure folds more than MDF suggests, and solvers prefer smaller bet sizes and lower-variance lines (downward drift). How strong the effect is depends on the tournament stage, who covers whom and the payout structure. Playing pure chipEV in these spots loses real money.
ChipEV vs ICM: Basics
- ChipEV — the goal is to maximize the chip count.
- ICM (Independent Chip Model) — the goal is to maximize $EV, the monetary value of the chips.
- Even perfect ChipEV play can lose a large share of a buy-in in 3-bet pots on the bubble or at the final table (one course example: up to 30%; the source was not verified).
- Chips are not money but a means of "survival" in a tournament.
Risk Premium and Bubble Factor
Definitions
- Risk Premium (sometimes called "reward for risk") = the extra equity you need under ICM compared with ChipEV to call or get the money in (ICM required equity − ChipEV required equity).
- Bubble Factor = $EV you risk losing / $EV you can win in the same all-in.
| Situation | Bubble Factor |
|---|---|
| ChipEV (no payout pressure) | 1.0 |
| Tournament bubble | 1.5–2.0+ |
How ICM Pressure Changes Through a Tournament
- ICM starts to matter noticeably at roughly half the field remaining (in an HRC simulation of a 200-player tournament; switching to ICM only at 25% already cost ROI). A reliable guide: when the players left are about three times the number of paid places. The pressure keeps rising until the bubble.
- Maximum pressure: the money bubble and the final table.
- Pressure decreases towards heads-up (HU).
Postflop Adjustments for ICM
Key Principles
- Bluff more as the aggressor: under ICM the betting player gains an advantage.
- Fold more as the defender under bust-out pressure.
- Smaller sizings give the opponent fewer incentives to call.
- Lower-variance lines keep $EV more stable.
Example Mini-Game (3 players, 2 tickets)
| Scenario | ChipEV | ICM | MDF (call %) |
|---|---|---|---|
| Player in position (bluff %) | ~45% | ~64% | — |
| Player out of position | ~17% | ~10% | ~16.7% / ~10.3% |
Illustrative numbers from a course example; the original source was not found — read the table for the direction of the effect.
Conclusion of the mini-game: under ICM the aggressor in position can bluff more, and the opponent out of position calls less.
EV Leakage in Tournaments
- EV gained ≠ EV lost.
- Example: you bust a player and win $698 of $EV, but $514 leaks to the rest of the field — 42% of the $1,212 the busted player lost ($698 + $514 = $1,212).
- Leakage increases with stack imbalance.
- The covering player has a "risk advantage" (see "Risk Advantage in Tournament Poker").
Downward Drift: Smaller Bets Under ICM
- Concept: under ICM solvers choose smaller bets to reduce variance.
- Example: a ⅓-pot bet is more common in ICM spots than in ChipEV.
- Even a shove can be a lower-variance option than a small 3-bet — with short stacks, where a small 3-bet already commits a large part of the stack and leaves you guessing postflop.
GTO Wizard Example: K87ss Board
- Comparing ICM vs ChipEV strategies on a K87 two-tone (spades) board (positions and stacks are not given in the source).
- ICM: dominated by ⅓-pot bets.
- ChipEV: a balance between ⅓-pot and ⅔-pot.
- A GTO solution under ICM accounts for stack coverage, tournament phase and payout structure.
Further Study
- ICM visualization (video).
- 5 tips from Oston Obrezkov (video).
- Theory of Risk Advantage — Hunt (video).
- Auto-generator of reward-for-risk graphs for different tournaments.
Conclusions
- ICM pressure affects frequencies, sizings, aggression and passivity.
- Ignoring ICM postflop means losing $EV.
- The closer to the bubble or the final table, the more important ICM adjustments become.
Tags: #ICM #MTT #postflop #EV #EVleak #bluff #variance
Drill it in the trainer
Theory sticks when your hands do the reps: range quizzes, an AI coach and reviews of your own hands — free start on Telegram.
Train in the bot