The biggest difference between a tournament and a cash game is not the rules. It is time.
In a cash game you can sit 100 big blinds deep all night. In a tournament the blinds go up every fifteen or twenty minutes, so even if you never lose a hand your depth slides from 100 big blinds to 40, then 20, then 10. Play the same way from start to finish and at some stage you become the easiest target at the table.
- ✓A tournament is a timeline that only goes down, and strategy has to change with stack depth.
- ✓Once antes arrive, a steal needs to work only 47% of the time instead of 59% — the mathematical reason to widen in the middle stages.
- ✓Under 20 big blinds the question shifts from “how do I win this hand” to “when do I get the chips in”.
- ✓Pressure on the bubble is asymmetric: short stacks fear elimination, big stacks apply pressure for free.
- ✓Online, table count and the clock both erode decision quality. Play as many tables as you can genuinely handle.
Early: deep stacks, but no hurry
Most online tournaments start with 100 to 200 big blinds. It is the deepest stage of the event, and the one where the most players head in the wrong direction.
The common misconception is that you need to “establish an aggressive image” early. In reality the blinds are so small relative to your stack that they barely matter, and risking a third of your chips in a big pot to win a 1.5 big blind pot is a poor trade.
What actually matters early is two things.
First, tie your hand selection to position. Tight from early position, wider from late. It is the most reliable adjustment available at depth, and it needs no read on anyone. For how position shapes ranges, see preflop basics.
Second, speculate with the hands that need depth to be worth it. Small pairs and suited connectors like 8♠7♠ earn their value by flopping big and winning a full stack. That payoff only exists while stacks are deep, so the early stage is the one time they are genuinely profitable.
With a caveat: play them in position, and keep the investment small. Calling a preflop raise with a small pair should cost no more than 5% of the effective stack. For how depth drives these decisions, see stack depth.
Middle: antes arrive and stealing gets cheaper
In the middle stages the structure usually adds an ante, and that affects strategy more than most players realise.
Look at the numbers. You open to 2.2 big blinds from late position, trying to steal:
| Situation | Pot on the table | You risk | Success rate needed |
|---|---|---|---|
| No ante | 1.5 | 2.2 | about 59% |
| With a big blind ante (1 BB) | 2.5 | 2.2 | about 47% |
The formula is amount risked ÷ (amount risked + pot available). The ante inflates the pot before the hand starts, so a successful steal wins more and the required success rate drops a full twelve percentage points.
That is why the middle stages call for a wider late-position opening range — not because aggression is a virtue, but because the profitability threshold for the same action genuinely fell. By the same logic, expect others to be stealing from you, and widen your big blind defence rather than folding your way through.
Late: from playing hands to picking moments
Below 20 big blinds, the nature of the game changes.
Under 20, every preflop decision needs an answer to what happens when you get 3-bet. If the answer is that you can only fold, the hand should not have been raised.
Under 10, preflop simplifies to shove or fold. The reasoning is direct: an open-raise leaves less than half your stack behind, so any postflop bet is all in anyway. Do it in one move and collect the full fold equity while you are at it.
Hand values invert at this stage too. Small pairs and suited connectors lose their speculative value, because the “win a whole stack” payoff no longer exists. High cards that hold up inside a calling range — something like A♥J♦ — go up in value instead.
Position matters most here as well. With the same 12 big blinds, shoving from the button means getting through two players; shoving from early position means getting through seven. The success rates are nothing alike.
The bubble: pressure is asymmetric
Close to the money, a chip is no longer worth its face value. Doubling your stack does not double your prize, but busting means collecting nothing at all — and that asymmetry is the core of ICM.
What it produces in practice is this: the same all-in decision costs completely different amounts depending on your stack.
- Short stacks: every decision could be the last, so they over-tighten and wait for someone else to bust first.
- Big stacks: knocking someone out carries no fatal risk, so they can apply constant pressure and force medium stacks to fold.
- Medium stacks: the worst position. Without a big stack’s leverage, and without a short stack’s “I am nearly out anyway”, they get squeezed from both sides.
The practical rule is simple: on the bubble, avoid calling all-ins with marginal hands, but increase how often you apply pressure yourself — particularly against medium stacks visibly waiting to cash. Once the bubble bursts that pressure disappears, and it is important to widen back out.
Three things specific to online play
Match your table count to your processing speed. More tables means more hands per hour, but less time on each decision. If you regularly need the time bank to finish thinking, you are on too many — trading decision quality for volume rarely pays in tournaments, because the decisive spots are concentrated in a handful of hands.
Work out rebuys and add-ons in advance. The add-on at the end of the rebuy period is usually good value, since it is the cheapest chips in the event. But unlimited rebuy formats push your real cost far above the advertised buy-in, so decide your maximum before you register.
Treat event length as a cost. A large online tournament can run eight to ten hours, and the drop in decision quality late on is a real loss. Rather than grinding through it, make schedule length part of how you choose events.
