Nobody calls a coinflip in poker what it actually is. 88 against AK isn't a coin toss — it's 56/44, and that gap is where the money in the hand lives.
What a coinflip really is
The standard matchup is a medium pair against two big cards: 88 against AK. The table calls it a flip, everyone snap-calls, nobody does the math. Here are the real equity numbers — how often each hand wins — for the spots players lump together:
- 88 against AK offsuit: 56 / 44 for the pair (with AK suited it drops to 52 / 48).
- AK against QQ: 43 / 57. Not a flip. That is three-to-four.
- AQ against JJ: 43 / 57. Just as bad, and snap-called just as fast.
- AK against 22: 47 / 53. The small pair is still ahead.
- 76 suited against AK: 42 / 58.
Those twelve points of equity between a 56/44 and a 44/56 are the whole business. Calling both of them a flip is how money leaves quietly: you take one because you are ahead, and the other because it sounds the same.
In cash games the flip matters less than you think
In a cash game your chips are worth exactly what they say, so an all-in at 50% is break-even before rake and losing after it, while 56% leaves a thin edge. The useful question is not "do I take the flip?" but "do I have a bigger edge playing three streets against this player than I do playing five cards?"
Against the guy who calls 70% of flops, putting your stack in at 56/44 trades your big edge — his postflop mistakes — for a small one. Against a regular who plays as well as you do, that same 56/44 is one of the best deals available all session, because postflop you take nothing off him.
In tournaments the flip is an ICM decision, not an equity one
Tournament chips are not worth a fixed amount: doubling your stack does not double your expected prize, but busting takes it to zero. That is ICM, and it has an uncomfortable consequence — a hand that wins chips in the long run can still lose money.
- As the short stack, your chips are cheap and the blinds are eating you anyway: take it. The right frame there is push or fold, not "I'll wait for a better one".
- When the pot already holds dead money — antes, blinds, an earlier call — you don't need 50% for the shove to pay. Sometimes 40% is enough.
- On the bubble with an average stack, folding is usually right: you have a lot to lose and little to gain while the short stacks knock each other out.
- With a comfortable stack and the best seat at the table, a 52% flip is a bad trade. You are selling your postflop edge at the price of five cards.
What does not work is refusing flips on principle. The player who "doesn't play flips" ends up passive, lets the blinds do his opponents' work, and shoves three levels later with half the chips.
The way players count flips
At 55%, a run of three flips ends in three losses once every eleven runs. That's not a historic stretch of bad luck — that's Tuesday. And because we remember what hurt, the mental record always says we lose more than our share. That's the job of a session log and of understanding variance with numbers instead of memory.
If you have never checked the numbers on your regular matchups, run them through an equity calculator — the ones we use are listed on our tools page — and memorize five. Those five will cover most of the preflop all-ins you play this month.
The lesson
A flip is not an act of faith or a beat you take quietly. It's a price decision: real equity, dead money in the pot, and what a chip is worth at that exact point in the tournament. When all three line up, take it without thinking. When they don't, folding isn't cowardice, it's arithmetic. And when the numbers are level, the pot goes to whoever knows which player has more to lose.
Close all-ins are what the community argues about most, and the answer nearly always depends on the stacks. If you want yours reviewed by players who take it seriously, I want in on the club.