ICM in Poker Tournaments: When to Fold a Good Hand Near the Bubble

Last Updated on October 12, 2026 by Bala Kumar

Early on Day 4 of the 2026 WSOP Main Event, 1,389 players had chips but only 1,382 would get paid. Play went hand-for-hand almost immediately. In the first round, four short stacks went all-in and all four survived. In the second round, six players were all-in at once, and the bubble burst. Former world champion Chris Moneymaker was one of the casualties.

Every one of those players faced the same question: is this hand good enough to risk my tournament right now?

That question sits at the heart of ICM, the Independent Chip Model. ICM is the reason a hand that’s an easy call in a cash game can be a clear fold near the money. This guide explains how it works in plain English, using real numbers from the 2026 Main Event and calculations we ran ourselves.

What Is ICM in Poker?

ICM (Independent Chip Model) converts your tournament chips into real-money value. It looks at every player’s stack and the payout structure, then estimates what share of the prize pool each stack is worth.

The key idea is simple: in a tournament, chips you win are worth less than chips you lose.

In a cash game, a $100 chip is always worth $100. In a tournament it isn’t. If you double your stack, you don’t double your prize money, because you can only win first place once. But if you lose every chip, your tournament is over, and you give up every future payout jump you might have reached.

This imbalance is called ICM pressure. It’s strongest:

• right before the money bubble bursts
• at final tables, where pay jumps get big
• whenever a short stack at the table might bust before you

Real Data: What ICM Did to the 2026 WSOP Main Event Final Table

Here’s where ICM gets eye-opening. The 2026 Main Event final table started with 553 million chips in play and $30.25 million in prize money for the last nine places, from $1,000,000 for ninth up to $10,000,000 for first.

We ran the stacks through a standard ICM calculation (the Malmuth-Harville model, which most ICM tools use):

Source: final-table chip counts and payouts as reported by Bleacher Report. ICM values calculated by PokerClubGames.

Two things jump out:

1. Jumalon had over a third of the chips, but ICM valued his stack at about $6.2 million, not $10.6 million. Big stacks are always worth less than their chip share, because they can’t win more than first-place money.
2. The shortest stack, Evagorou, was worth $2.19 million, almost double his raw chip share. He was already guaranteed $1 million, and every player who busted before him added to his payout.

This is exactly why short stacks are worth more than they look, and why risking them carelessly is so costly. Saaskilahti started the final table seventh in chips and finished second for $6 million.

How ICM Changes the Hands You Can Call

Here’s the practical part. When someone shoves all-in and you’re considering a call, ICM raises the equity you need. Equity means your chance of winning the hand.

Without ICM, calling an all-in for even stacks needs about 50% equity. With ICM, you need more, sometimes a lot more.

Example 1: The final table call

Imagine chip leader Jumalon (194M) shoves into Greg Mueller (48.5M) at the start of the final table. If Mueller calls, he’s risking his whole stack. For simplicity, we’ll ignore blinds and antes.

• Chip-EV breakeven (ignoring ICM): 50%
• ICM breakeven: 63.8%

Mueller needs nearly 64% equity just to break even. For perspective, even pocket jacks against ace-king are only about 57%. In other words, Mueller should fold some very strong hands here.

The pressure varies by seat:

Interestingly, the second-biggest stack faces the most pressure. Hammoud had the most money locked in and the most to lose by busting in ninth place.

Now reverse the spot. If Mueller shoves and Jumalon considers calling, Jumalon needs only 52.5%, because losing doesn’t knock him out. This is why big stacks can attack medium stacks so aggressively at final tables.

Example 2: The classic four-left bubble

Let’s bring it down to a typical online tournament or sit-and-go. Four players remain and three get paid: $500 / $300 / $200. Stacks are 4,000, 3,000, 2,000 and 1,000 chips, with 300 in blinds and antes in the pot.

The chip leader (4,000) shoves. How much equity do the others need to call?

Look at that first row. The 3,000-chip player needs over 70% equity to call. That means folding hands like pocket tens, ace-king and ace-queen against most realistic shoving ranges. It feels painful, but folding is right because:

• If you fold, the 1,000-chip short stack is very likely to bust first and put you in the money.
• If you call and lose, you leave with $0, while the short stack you were waiting on gets paid.

That’s the core lesson of bubble play: don’t race the short stack to the exit.

Want to run your own spots? Use our ICM Calculator: enter the stacks and payouts, and it shows each player’s real-money equity instantly.

When You Should Fold a Good Hand Near the Bubble

Use this checklist. The more of these that apply, the tighter you should call:

1. A shorter stack is still at the table. If someone is likely to bust before you, your fold has real value.
2. The shover covers you. If losing knocks you out, ICM pressure is at its highest.
3. You’re a medium stack. Medium stacks feel the most pressure. They have enough chips to survive, but not enough to absorb a loss.
4. The pay jump is large. A move from “nothing” to “min-cash”, or from 9th to 8th at a big final table, raises the cost of busting.
5. Your hand is good but not dominant. Hands like AQ, AJ and medium pairs suffer most. Premium hands like AA and KK are still almost always calls.

When ICM Says Go Ahead and Play

ICM isn’t just about folding. It also tells you when to be aggressive.

• You’re the big stack. Opponents must fold more often, so your shoves and raises win more pots without a fight. Pressure medium stacks, not the short stack, who has little to lose by calling.
• You’re the short stack. Your chips are worth more than they look, but you can’t wait forever, because the blinds will eat your stack. Choose spots to shove first rather than calling.
• The payouts are flat. In tournaments with small pay jumps, ICM matters less, and you can play closer to chip EV.

Why Moneymaker’s Bust Wasn’t an ICM Mistake

Back to the Main Event bubble. Moneymaker called all-in with sevens full of eights and ran into aces full. That’s a cooler, not an ICM error. A full house is near the very top of any reasonable range, and no ICM model would fold it.

The bubble also shows how small the margins are. The players who busted shared the first two $15,000 payouts, effectively getting their $10,000 buy-ins back. Everyone who survived locked in a payout. Over the next day, that field shrank from 1,382 players to 533.

ICM helps you make good folds, but it never asks you to fold monsters.

Common ICM Mistakes

• Calling too light as a medium stack. This is the most expensive error in tournament poker. Your stack’s real-money value is higher than its chip count suggests.
• Big stacks being passive. If you have chips, the bubble is your best time to accumulate more. Opponents are scared to call.
• Applying ICM too early. On Day 1 of a big field, far from the money, play mostly by chip EV. ICM pressure builds as the money approaches.
• Forgetting future play. Basic ICM doesn’t account for skill or future hands. A strong player may choose to pass up a marginal call to keep playing hands where they have an edge.

Quick-Reference Summary

1. Chips lost are worth more than chips won in tournaments.
2. Near the bubble, medium stacks need far more equity to call (often 60–70%+ instead of 50%).
3. Big stacks should apply pressure; short stacks should shove rather than call.
4. Check the table: if a shorter stack might bust first, folding often earns you money.
5. Premium hands are still calls. ICM trims your calling range; it doesn’t erase it.

FAQs