Folding AKs on the bubble?
When ICM beats the chips. Concretely, with numbers.
One of the hardest lessons in tournament poker: a decision that wins chips can cost money. It sounds absurd. It is essential to understand if you play tournaments seriously.
The situation: the bubble of an $11 tournament. You have 20 big blinds, the small blind goes all in. You are in the big blind with AKs. Counted in chips: an easy call. With the bubble factor included? A fold.
Here is why.
The situation
scenario
- Buy-in: $11
- Players: 11 left, top 10 in the money ($20 for place 10, $1,000 for place 1)
- Your stack: 20 big blinds
- The small blind goes all in for 20 big blinds
- Your hand: A♠ K♠
- Hands he does this with (estimate): 99+, AJ+, KQs and suited connectors (~14% of all hands)
Counted in chips
Against those hands AKs wins roughly 47 percent of the time. The expectation of calling, in chips:
Calling: 0.47 × 40 bb - 0.53 × 20 bb = 18.8 - 10.6 = +8.2 bb
Folding: 0 bb
Difference: calling wins +8.2 bb in chips
An instant call then, surely?
Counted in money (ICM)
Now in money, and this is where it gets interesting. ICM (Independent Chip Model) converts your stack into your chance at every place in the prize money.
The bubble factor in this spot (1 place from the money, 11 players, a big jump from 20 to 1,000 dollars) is roughly 1.5.
Chance of winning needed to call profitably under ICM:
Needed in chips = 33%
Needed under ICM = 33% × 1.5 = 50%
Chance of AKs winning against his hands = 47%
Outcome: calling is a loss in money
In money: calling costs you roughly $2.20 in expectation. Folding keeps you in the tournament with 20 big blinds and preserves $2.80 of expected prize money.
net
Folding earns you $0.60 in expected profit compared with calling, while calling is +8.2 big blinds in chips.
Why?
Three reasons why a decision that wins chips can still lose money:
- Busting weighs heavier than doubling. Losing means zero. Doubling does not mean twice your current value, because the prize money is concentrated at the top.
- Others are running a risk too. On the bubble no small stack wants to go out. By taking the risk yourself you clear their path to the prize money, even if you win the pot.
- A stack is worth more than its chips. With 20 big blinds you still have options after the flop: reraising, trying to hit a strong hand cheaply, betting for value. With 0 you have nothing. The value of those options does not grow in proportion with your chips.
3 ICM rules of thumb
The big stack applies pressure, the small stack ducks away
With a big stack on the bubble you may raise: smaller stacks cannot afford to call, because for them losing means being out. With a small stack you do the opposite: throw away the borderline cases, wait for a clearly strong hand, or look for your chances against other small stacks.
What your stack is worth counts, not how many chips it is
In a tournament not all chips are worth the same. Your first chip is worth more than your fiftieth, because whoever loses their last chip is out and gets nothing. ICM works that difference through. The rule of thumb: the closer to the bubble, the bigger the gap between winning in chips and winning in money. With 100 places to go before the money that gap is minimal; with 1 place to go it is enormous.
The bubble factor sets how tight you call
A bubble factor of 1.4 means: you need 40 percent more chance of winning than the chips alone ask for to call profitably. A decision that is neutral in chips (50 percent chance of winning needed) suddenly asks 70 percent under ICM. That explains why AK is sometimes thrown away on the bubble: not because it is a bad hand, but because 70 percent against an all-in range is rarely realistic.
frequently asked questions
Deeper questions.
What is a reasonable bubble factor to work with? +
Common bubble factors per situation: 100 places before the money 1.0 (near enough pure chips). 20 places before it 1.1. 5 places 1.3. Exactly on the bubble 1.4 to 1.6. At the final table 1.5 to 2.0, depending on how the prize money is spread. Learn these by heart, because at the table you do not have a calculator open.
Does ICM apply in cash games too? +
No. In cash games there is no prize distribution: every chip is real money. ICM exists only in tournaments and sit and gos, where your chips are converted into prize money through the payout structure. In cash you always simply count in chips.
How do I learn to apply ICM at the table? +
Build rules of thumb instead of trying to calculate at the table. Examples: throw away AQ against an all in of 20 big blinds on the bubble; go all in yourself with any two cards at 8 big blinds or less; call half as wide around a big jump in the prize money. That is exactly what Pokiq's daily tournament training practises: a feel for ICM through repetition.
What is the biggest ICM mistake on the bubble? +
Calling with AKs or QQ against a small stack's all in, on the assumption that this is an automatic decision. In chips that is often right, but the bubble factor turns it into a fold. Best approach: check the real money expectation with the free ICM calculator at /en/tools/icm-calculator/ before you decide.
Do all pros play this tight on ICM? +
On the bubble: yes. Well known tournament pros such as Lichtenberger, Boeree and Kornuth are explicit about it. At final tables pros sometimes deviate because they are exploiting a specific opponent or table dynamic, but the basis stays: play according to ICM.
What is a good book on ICM? +
"Mathematics of Poker" (Chen and Ankenman) is technical but it is the standard work. "Excelling at No Limit Hold'em" by Jonathan Little has a good ICM chapter. Online, the ICM courses by Steve Paul on Run It Once are strong.
train your feel for ICM
An ICM situation every day.
Pokiq's tournament training puts an ICM situation in front of you every day. Plus a free ICM calculator to work situations through straight away.