Updated on 21 Feb 26 by
Barry carter
Poker Expert

Poker Basics - ICM Final Table Deals

We look at several ways you can chop a final table, starting with the gold standard, the ICM deal.

Most people do not think about final table deal making strategy until they actually find themselves at a final table. There are lots of ways to deal at a final table, the most common of which the ICM deal, which is where each player's stack is valued according to the Independent Chip Model (ICM). 

Most players these days know ICM as a formula for playing end game tournament strategy, but it was first devised under a different name as a deal making method by Mason Malmuth and David Harville. It was designed to estimate the chances of each player finishing in each finishing position based on their stack size (skill is not included in the calculation). 

It's best to explain the ICM deal with an example, here is the final four players at a final table and a deal has been proposed. The potential prize is what they would win if they finish in the position they currently find themselves in:

Player Chips Potential Prize ICM Value
Amy 100,000 $200 $140.84
Paul 70,000 $120 $123.80
Sophie 50,000 $80 $108.96
Tim 20,000 $50 $76.40

In this example Amy is the chip leader and if she maintains her position will win $200, but a bad run of cards could see her bust next for $50. There are lots of reasons to make a final table deal and avoiding variance is the most common reason most final tables end up in a handshake deal. 

In this example Amy is on course to win $200 but with the ICM deal she would get $140.84. That is perhaps a lot less than a chip leader might expect. Likewise Tim is the short stack in 4th place but has an ICM value of $76.40 which is almost the 3rd place prize. 

ICM deals do not factor in skill and the first time players see them they are often surprised. They highlight a fundamental rule of ICM which is that the more chips you have the less each one is worth. Typically ICM deals will see the chip leader have to give up the most in potential prizes and the short stack gets the biggest boost compared to what they would get for maintaining their position. 

This is the nature of deals, the chip leader has to give up the most potential equity in favour of securing a big pay day. The short stack has to be incentivised to stop playing as they are essentially freerolling after the bubble has burst. 

To demonstrate why deals are structured this way, this is the percentage chance of each player finishing in each position, according to ICM:

  1st 2nd 3rd 4th
Amy 41.7% 31.9% 20% 6.4%
Paul 29.2% 31.2% 27.5% 12.2%
Sophie 20.8% 25.4% 33.2% 20.6%
Tim 8.3% 11.6% 19.3% 60.8%

Tim comes last 60.8% of the time, but 39.2% of the time he manages 3rd or better, which would get him $80+. This is why he gets to lock up an amount close to that 3rd place prize. Amy wins 41.7% of the time but gets a place lower than her ICM value 58.3% of the time. Most of the time Amy gets less than her ICM payout, so locking up $140.84 risk free is actually a favourable outcome. 

People who do not understand ICM will often reject a deal when they have a big stack and snap the hand off of anyone who offers them close to ICM as a short stack. It's probably more important for deal making to understand these finishing position distributions than it is to understand the actual ICM value of your stack (you can do them yourself at PrimeDope).

Next time we will look at an alternative form of final table deal known as the chip chop.

Related Resources

poker money

The truth of variance

Our Twitch coach LemOn36 with a guide on variance for cash games

poker money free

MTT variance

Collin Moshman explains the ins and outs of tournament variance

Poker Expert

Barry Carter is the editor of PokerStrategy.com and the co-author of The Mental Game of Poker 1 & 2, Poker Satellite Strategy, PKO Poker Strategy, Endgame Poker Strategy, GTO Poker Simplified, Mystery Bounty Poker Strategy and Beyond GTO. In 2025, he won the Global Poker Awards for Best Book and Twitter Personality of the Year.