Tournament strategy
MTT Strategy in Poker: What Changes at Each Stack Depth
MTT strategy is how you change your play in a multi-table tournament as the blinds rise and your stack gets shorter in big blinds. Deep, most of the money goes in after the flop. Under about 15 big blinds you shove or fold, and near a pay jump ICM raises the price of calling an all-in.
Two of our lessons put a pocket pair on the cutoff against an under-the-gun raise. With 150 big blinds behind, you call with nines and hope to flop a set. With 15, you move all in with sixes. Same seat, same kind of hand. The stack decides it.
The shove ranges below come from our push/fold solve, and the ICM numbers from the engine behind our bubble factor calculator. If you're new to tournaments, start with what an MTT is.
MTT strategy by stack size
M is your stack divided by what one orbit of blinds and antes costs. Harrington's zones turn it into colours: green from M 20, then yellow, orange and red, and dead under M 1. The figures assume nine players and a big blind ante, the format the WSOP and EPT use, where an orbit costs 2.5 big blinds. The M-ratio calculator works out your own stack, and the M to big blinds tables cover other ante formats.
| Stack | What to do |
|---|---|
| 60 BB and upM 24+, green | Most chips go in after the flop. Position matters, and a small pair can call a raise to flop a set. |
| 40 to 60 BBM 16 to 24, yellow, green | Every steal is worth more with an ante. Open wider from late position and defend your blinds. |
| 20 to 40 BBM 8 to 16, orange, yellow | Stop flatting late-position opens and three-bet all in over them. |
| 10 to 20 BBM 4 to 8, red, orange | Raise near 20. By about 15, open-shove from late position. |
| Under 10 BBM under 4, red, dead | Push or fold from every seat. |
60 big blinds and deeper: play after the flop
Early levels play like a cash game. Stacks are deep enough that most of the money goes in after the flop, so position and which hands you play matter most.
A pocket pair flops a set or better 11.8% of the time, about one flop in 8.5. Call a 3 big blind raise to hit one and you need to win about 22.5 big blinds, on average, on the flops where you do. At 150 big blinds against a tight opener that's easy. At 15 it can't happen: the most you can win is about 17.5, his 15 plus the blinds and ante.
40 to 60 big blinds: stealing pays
Q8 suited from the cutoff is a fold with no ante and an open with one. Once the ante is in, a steal wins 2.5 big blinds instead of 1.5. A 2.2 big blind open needs everyone to fold 59% of the time to break even with no ante, and 47% with a big blind ante. With 40 to 60 big blinds those pickups are a real share of your stack. The same ante cuts your M by 40%, which the M to big blinds article works through.
20 to 40 big blinds: stop calling, start re-stealing
A flat call now costs too much of your stack, and a three-bet all in starts to work. When the button opens to 2.2 and you have 25 in the big blind, there are 4.7 big blinds in the middle, 2.7 of them other people's, and a shove takes them whenever the button folds. Against someone opening every orbit, that's often. In the lessons you re-steal from the big blind twice: a three-bet with A9 suited at 45 big blinds, and an all-in with A8 suited at 24.
Our push/fold solve only offers shove or fold, so it can't show a raise or a call. It still says how strong a hand has to be to put 25 big blinds in first: 30% of hands from the button and 6% under the gun.
10 to 20 big blinds: raise or shove
At 20 big blinds a 2.2 big blind open is 11% of your stack, and you can still fold it to a shove. A PokerNews solver look at ace-king prefers a normal raise from 17 to 20 big blinds and has the open-shove taking over around 14. At 10 there's no flop worth playing for, and the lessons jam A-T from the cutoff.
Here is how wide each seat shoves when it's folded to you, in a solve that only allows shove or fold. Read the 15 and 25 big blind columns as how wide a shove can go, not proof that it beats a raise.
| Seat | 25 BB | 15 BB | 10 BB | 8 BB | 5 BB |
|---|---|---|---|---|---|
| Under the gun | 6% | 10% | 15% | 17% | 38% |
| Cutoff | 19% | 28% | 34% | 37% | 53% |
| Button | 30% | 36% | 43% | 48% | 60% |
| Small blind | 47% | 63% | 72% | 76% | 83% |
Shove percentages by seat has every stack from 1 to 25 big blinds and the hands behind each number.
Under 10 big blinds: push or fold from every seat
Under the gun more than doubles, from 15% at 10 big blinds to 38% at 5. By 5 the big blind calls a button shove with 78% of hands, because the price is that good. Any two cards becomes a shove under the gun at 1.5 big blinds.
Below 2.5 big blinds you're under M 1, Harrington's dead zone. Nobody folds to a stack that small. In our dead zone lesson you ship Q5 suited and hope to win the showdown. Part nine drills every zone, and the push/fold chart gives the exact hands at any stack.
How deep real tournaments are at the bubble and final table
The 2025 WSOP Main Event gave everyone 60,000 chips at 100/200 with a 200 big blind ante: 300 big blinds, with two-hour levels. The money bubble burst on Day 4 at 4,000/8,000. With 584 million chips among the last 1,461 players, that's about 50 big blinds each. The final table started at 800,000/1,600,000 with an average of 40.6 big blinds. The chip leader had 68, the shortest stack 15, and four of the nine sat between 20 and 40.
The Main Event is slow. Its levels last two hours, against one hour in the WSOP's $1,500 Millionaire Maker. Faster events reach the money and the final table with less, which pushes more of the decisions that decide a tournament into the 10 to 25 big blind range.
How the bubble and final table change the numbers
Near a pay jump the chips you lose cost more than the chips you win help, which the MTT explainer covers, and ICM worked by hand shows where the numbers come from. The bubble factor puts a number on it: the equity you need to call an all-in, which the bubble factor chart lays out.
A field of hundreds is too big for an exact ICM count, but a sit-and-go has the same pay jumps on a small scale. Say it pays 50/30/20 and four players are left with equal stacks. Calling an all-in for your stack needs 65.2% equity, where counting chips you'd need 50%. One elimination later, with three left and everyone paid, it drops to 57.1%.
At the 2025 Main Event final table, the chip leader had 18.5% of the chips, and ICM put that stack at 14.7% of the $31.05 million left to pay. The shortest stack had 4.0% of the chips and 6.9% of the money. The second stack, 58 big blinds, needed 64.8% equity to call an all-in from the leader. With nine equal stacks it would be 61.0%. Heads-up it falls to 50.0%, because only one pay jump is left and chips are money again.
Part ten plays both sides of it. Near a pay jump you fold A-K to a covering all-in, and on the bubble with an M of 2 you shove A8 suited from the hijack anyway. Your own stacks go into the ICM calculator.
MTT strategy at a short table
The shove numbers above are for a nine-handed table, counted by the players left to act behind you. At six-max that changes only the early seats: read the lojack row of the shove table as under the gun. The button and blinds play the same ranges, because they have the same players behind them and the same 2.5 big blinds to win. The blinds do come round faster, which effective M accounts for. The short-handed lessons are in part eleven.
How these numbers were made
- Shove and call percentages: our chip-EV Nash push/fold solve, nine-handed, blinds 0.5 and 1 with a 1 big blind ante, one caller at most, shove or fold only, no ICM.
- M and zones: the same code as the M-ratio calculator, with a big blind ante.
- ICM and bubble factors: the exact Malmuth-Harville engine behind the ICM and bubble factor calculators.
- Main Event figures: entries and places paid, starting stack and levels, the Day 4 restart level, final table chip counts and payouts. The bubble average is our calculation from the total chips in play.