Run it twice in poker means dealing the remaining board twice once two players are all-in: half the pot is decided by the first runout, half by the second. It takes two words to ask for and it starts an argument at every table. Here are the mechanics and the numbers, because most players ask for it for the wrong reason.
How it works
Two players are all-in and no action is left. Before the remaining cards go out, either player proposes running it twice. If the other agrees — and if the house allows it — the dealer runs the rest of the board once, records that result, then deals a second board from fresh cards in the same deck, never repeating a card that already came out.
The pot is split in half. Each half goes to the winner of its board. Win both and you take it all. Win one and lose one and both players get their money back. Three things that matter:
- It exists in cash games only. Not in tournaments, and not out of stubbornness: tournament chips cannot be cut in half when an elimination is on the line. ICM would break.
- It needs both players to agree. Online it is usually a checkbox in the table options and only fires if both have it on.
- House rules vary, and at some tables it changes how rake is taken. Ask before the all-in, not after.
It is normally offered from the flop onward. Some rooms and clubs allow it on preflop all-ins too, and PLO tables often run it three times.
The numbers: what changes and what doesn't
This is the whole argument. You are all-in on the flop with a flush draw: nine outs, two cards to come, roughly 35% equity.
Running it once: you win the whole pot 35% of the time and nothing 65% of the time.
Running it twice, rounded: you win both boards about 12% of the time, split one each about 46%, and lose both about 42%. Do the math: 12% of the full pot plus 46% of half the pot is the same 35%.
That is the line to remember: run it twice does not touch your EV, only your variance. The expectation of that all-in is identical. What drops, roughly by half, is the variance of that single hand. Same expected result, delivered in a narrower spread of outcomes.
Which is why there is no edge in accepting or declining. Not as the favorite, not as the underdog. If someone at your table claims you only ask when you are behind, they are making it up. It does not matter who proposes it — the math is the same on both sides.
When it is worth asking
If it does not change expectation, what is it for? Risk management. It earns its place when the size of that pot is large relative to your bankroll, not when it stings.
Where I ask for it:
- The pot has grown past what a normal stack looks like at that table and you are playing a tight bankroll in buy-ins. Less variance per hand means less risk of having to move down.
- You want to keep playing that table for another two hours. A pot lost in one card takes you off your A-game; a split result barely dents it.
- You play a high-variance format (PLO, very aggressive tables) where coinflip-sized all-ins are routine.
Where it adds nothing: stacks that are small relative to your roll, or asking because you hope the card will change. The card does not know how many times it is going to be dealt.
The mindset mistake
The real reason most players ask is not risk. It is not wanting to feel the bad beat. Used that way, run it twice is a painkiller, not a decision.
The problem is that it fixes nothing. You still need the only skill that counts when a big pot comes down to two cards: letting it go and playing the next hand well. If running it twice helps you reach hand 200 of the session intact, use it. If you use it to avoid looking variance in the face, the problem is still there and it comes back the first time someone declines.
Regs keep it simple: ask based on bankroll, accept instantly when asked, and keep a written stop-loss for the days when neither card arrives.
The takeaway
Run it twice is a risk tool, not an edge tool. There is no version of poker where running it twice makes you more money long term — it just smooths the road to that long term. Deciding by bankroll is judgment. Deciding out of fear is superstition with a rulebook.
At your table, is it standard or frowned upon? Tell us in the community — this argument never ends. Join for free.