Home/ Blog/ ICM in Poker: Why Chips Are Not Money
Strategy 27 August 2026 · 8 min

ICM in Poker: Why Chips Are Not Money

In a tournament a chip you win is worth less than a chip you lose. That single asymmetry explains most of the folds that look too tight near a bubble and are not.

In a cash game a chip is a dollar. Win ten, you are up ten. Tournaments do not work that way, and the difference is not a technicality.

Prize money is fixed and shared out by finishing position, so stacking off to double up gains you less than busting costs you. The chips you win are worth less each than the chips you already have. ICM, the Independent Chip Model, is the arithmetic that puts a number on that.

You do not need to calculate it at the table. You need to know which spots it changes, and there are fewer of them than the internet suggests.

Where the asymmetry comes from

Take a satellite with ten identical seats and eleven players left, all with equal stacks. Ten of them win the same prize. Doubling your stack barely improves your position, because you already had a ten in eleven chance and the extra chips buy you almost nothing. Busting takes everything.

Regular tournaments are a softer version of the same shape. The pay jumps near the bubble and near a final table are large relative to the chips at stake, so survival carries real value that chip EV alone does not see.

That is the whole idea. Every practical ICM rule follows from it.

Where ICM actually changes your play

Four situations, and outside them ICM mostly does not matter.

  • Approaching the money bubble, especially with a medium stack.
  • Approaching a pay jump at a final table.
  • Satellites, where the effect is at its most extreme.
  • Facing an all-in from a player who covers you.

Early in a tournament with a hundred players still in, ICM is close to irrelevant and you should play much as you would in a cash game. Play chip EV, accumulate, and stop worrying about a bubble that is four hours away. Applying bubble logic in level three is a leak in the opposite direction, and it is common among players who have just learned the concept.

The stack sizes that matter

ICM pressure lands unevenly, and knowing where you sit tells you most of what to do.

A big stack near the bubble should attack relentlessly. Everybody else is trying to survive, so your steals face far more folds than the maths alone justifies. This is the most profitable position in a tournament and the one most players underuse.

A medium stack is where the real discipline lives. You have something to lose and you are not the one applying pressure. Hands that are clear calls in a cash game become folds here, because busting costs you a pay jump you were on course for.

A short stack is under the least ICM pressure. You are close to busting anyway, so survival has less value and you can shove wider than the medium stacks can call. Short stacks that play scared waste the one advantage they have.

The mistake almost everyone makes

Calling too wide when covered, near a pay jump.

A player shoves and you hold a hand with slightly more than the equity you would need in a cash game. Chip EV says call. ICM often says fold, because busting costs a pay jump and the chips you gain do not repay it.

The size of the adjustment surprises people. Near a meaningful bubble, hands that need roughly 40% equity to call in chip terms can require well over half in ICM terms. That turns a lot of marginal calls into folds, and it is why good tournament players look strangely tight in exactly those spots.

The mirror image is also a leak: folding too much as the big stack, where ICM is barely working against you and everyone else is folding too much.

A worked hand

Nine players left, eight get paid, so one more bust ends the bubble. Blinds 5k/10k. We have 200k. The chip leader on our left has 600k, and there are two short stacks under 60k at other tables.

The cutoff, covering us with 350k, shoves for 24 big blinds. It folds to us in the big blind with AJo.

Against a shoving range that wide, AJo has decent equity and in a cash game this is a comfortable call. Here it is a fold, and not a close one. Two short stacks are about to bust regardless, so waiting is worth real money. Calling risks our whole tournament to win chips we do not urgently need, against a player who covers us.

Change one detail and the answer flips. If we were the short stack at 60k rather than 200k, we would be calling, because waiting no longer buys us anything and the blinds will take us out before those short stacks bust.

What this looks like in our clubs

Tournaments run daily in the club apps we operate in, and the ICM mistakes we see are not subtle. Medium stacks call off near bubbles with hands that were never worth their tournament life, and big stacks sit quietly instead of attacking a table full of people trying not to bust.

Club-app tournament fields also skew more recreational than a big room's, which pushes the adjustment further. In a pool where the average player has never heard of ICM, the big stack's steals face even more folds than the model predicts, and the medium stack's fold becomes even more clearly correct.

If you play tournaments in our clubs, that gap is where the money is, and it needs no software to exploit.

Frequently asked questions

What does ICM mean in poker?

Independent Chip Model. It converts a chip stack into a share of the remaining prize pool, which is what lets you compare decisions in money rather than in chips.

When does ICM matter most?

Near the money bubble, near pay jumps at a final table, and in satellites, where it is at its strongest because every qualifying seat pays the same. Early in a tournament it barely applies.

Should I fold pocket aces because of ICM?

Essentially never. ICM tightens marginal calling ranges, not premium hands. If a rule you have learned makes you fold aces outside an extreme satellite endgame, you have applied it wrongly.

Does ICM apply in cash games?

No. A chip is worth its face value in a cash game, so chip EV and money EV are the same thing. That is why cash-game calling ranges do not transfer to tournament bubbles.

How do I practise ICM without doing maths at the table?

Study the spots away from the tables until the shapes are familiar: covered medium stack near a pay jump means fold more, big stack near a bubble means attack more. You are learning patterns, not calculating equity while a clock runs.

Where to start

Next tournament, notice one thing: are you the covering stack or the covered one. That single question resolves most bubble decisions before you look at your cards.

For which of our clubs run the tournament schedules worth playing, message @supremaunion. If you are still choosing where to play, the tournament schedules differ a lot between clubs on the same app.

18+. Play within your means.

Read more on the blog

All posts →