ICM Without Myths: Typical Bubble and Final Table Mistakes in MTTs

ICM is not a magic spell and it is not “tight equals correct”. In 2026 most regulars already know the basic idea: chips are not worth the same once payouts matter. The real edge comes from applying that idea in messy, real tournament spots—different stack depths, antes, re-entry fields, and opponents who react emotionally on the bubble or freeze at the final table. This article breaks down the most common errors I see and, more importantly, how to fix them with practical checks you can use while you are playing. :contentReference[oaicite:0]{index=0}

1) What ICM really changes on the bubble (and what it does not)

ICM converts your tournament stack into a share of the remaining prize pool, assuming everyone has equal skill from this point and plays perfectly. That assumption is never fully true, but the model is still useful because it captures one key fact: losing your stack removes your chance to cash, while gaining chips does not increase your cashing probability in a straight line. This is why “chip EV” decisions can become expensive right before the money.

The biggest mental shift is understanding risk premium. Near the bubble you often need more equity to call an all-in than you would in a cash game or early levels, because the downside (busting) is priced higher than the upside (winning chips). Risk premium is not the same for every stack: short stacks can take thinner gambles, big stacks can pressure, and medium stacks often pay the highest price for mistakes.

What ICM does not change is the value of position, table dynamics, and who is actually capable of applying pressure. ICM does not automatically mean “fold everything”. If your table is over-folding because everyone wants a min-cash, aggression can be worth more than usual—especially with antes and big blind pressure. The trick is choosing the right targets and the right bet sizes for your stack.

Bubble myth: “Just fold into the money and you can’t go wrong”

This myth ruins more tournaments than bad beats. If you fold every marginal spot as a medium stack, you often arrive in the money with a stack that cannot threaten anyone. Your equity might feel “safe”, but you have traded away your ability to win meaningful pots without showdown. In modern MTT structures, blinds and antes do not wait for you to finish laddering.

A better approach is to separate “calling risk” from “attacking risk”. Calling off your stack against a covering stack is usually where ICM punishes you most. Stealing blinds, three-betting small, and opening hands that play well post-flop often carry less tournament risk while still increasing your stack. If people are clearly trying to coast, you should be the one collecting the dead money—within a disciplined plan.

Practical in-game check: if you are about to call an all-in, ask, “If I fold, do I still have a workable stack in the next orbit?” If the answer is yes, calling needs to be strong. If the answer is no (you will be forced all-in soon), your threshold drops. That single question prevents a lot of “I called because I was tilted by the bubble” punts.

2) Typical bubble mistakes by stack size (and simple fixes)

Short stacks (roughly 5–12 big blinds) often make the opposite mistake: they over-wait. They pass up profitable shoves because they hope two other players bust first. Sometimes that works, but in large 2026 fields with plenty of late-registration survivors, bubbles can last longer than players expect. Every orbit you lose fold equity, and once you are down to a few blinds you cannot apply pressure—you can only pray.

Medium stacks (roughly 13–30 big blinds) are the classic bubble victims. They can open and steal, but they hate calling off. Many medium stacks react by tightening everything, including their steals, and then they bleed into the danger zone. The fix is uncomfortable but clear: open enough hands to keep your stack healthy, but avoid calling off versus stacks that cover you unless you are truly at the top of your range.

Big stacks (30+ big blinds) often misunderstand their advantage. They either bully mindlessly and get trapped, or they become scared of “losing the chip lead” and miss easy steals. The correct big-stack edge is selective pressure: attack players who cannot call wide (medium stacks), isolate short stacks with hands that dominate their shoving range, and avoid ego battles against the only stacks that can hurt you.

Bubble leak: calling too wide in the blinds because “pot odds”

On the bubble, pot odds alone are a trap. When you call an all-in from the big blind, you are not just investing chips—you are investing tournament life. Even if the price looks tempting, your calling range usually needs to be tighter than your shoving range, especially when the raiser covers you or when multiple stacks behind can still bust before you.

Common example: you defend a marginal ace or a weak suited king versus a short-stack jam because it is “only a few blinds more”. If you lose, you are out. If you win, you are often still not a monster stack—you have simply moved from “comfortable” to “comfortable plus”. That is the classic ICM asymmetry: the downside is sharper than the upside.

Fix: build a default “bubble call discipline” rule. If you are covered, your calling range should lean towards hands that are less dominated and perform well versus tight jamming ranges—strong pairs, strong aces, and broadways that are not easily crushed. If you are the covering stack, you can call wider, but still avoid turning the bubble into a coin-flip festival without a clear reason.

3) Final table ICM: pay jumps, pressure points, and deal awareness

The final table is where ICM becomes visible, because pay jumps are large and everyone feels them. But the biggest mistakes are still the same two categories: players call off too lightly when they are covered, and players pass up profitable aggression when they have leverage. In 2026, with deeper structures in many live series and faster late stages online, the best players switch gears constantly rather than locking into “tight mode”.

At the final table, stack distribution matters more than the average stack. If there are two micro stacks about to hit the blinds, medium stacks have extra incentive to avoid all-in calls that could bust them before the short stacks go. Big stacks can exploit that by opening more buttons and cut-offs, applying three-bet pressure, and forcing opponents into uncomfortable risk premiums.

Another overlooked factor is post-flop discipline. Many players assume ICM is only about pre-flop shoves. In reality, the biggest equity losses often come from bloating pots with hands that cannot handle pressure, then feeling “priced in” and stacking off. Final table pots should be built with a plan: what happens if you face a shove, a check-raise, or a turn barrel that commits stacks?

Final table mistakes: punting as a medium stack and misusing “laddering”

The most expensive final table punt is the medium stack that takes a high-variance line against the only stack that covers them. It often happens with hands that look pretty—Ace-Jack, King-Queen, medium pairs—where the player wants to “take a stand”. If you are covered, you are usually the one who should avoid thin stacks-off spots, because your bust-out cost includes multiple pay jumps.

Laddering itself is not wrong; mindless laddering is. If you fold every spot waiting for someone else to bust, you hand the big stacks a licence to steal. The correct laddering mindset is targeted: avoid marginal calls that risk your tournament life when other stacks are likely to bust soon, but keep fighting for blinds and antes so your stack does not collapse. You are balancing survival equity and growth equity, not choosing only one.

Deal talk also needs realism. An ICM chop can be sensible when skill edges are small, stacks are shallow, or the pay jumps are extreme. But do not treat a deal as a moral victory or a failure. Treat it like a decision: compare the offered numbers to what your position is worth under an ICM model, then adjust for skill edge, structure, and fatigue. If you are tired and playing worse, that is part of the real calculation.