Running It Twice: The Math and EV

Running it twice deals the remaining board twice, each run-out settled for half the pot. The single most misunderstood fact about it: it does not change your expected value one cent. A 70% favorite still averages 70% of the money. What running it twice does change is variance — it smooths the wild swing of a big all-in into two smaller, independent results.

What “running it twice” means

After the money goes in, instead of dealing one final run-out, both players agree to deal it twice. Say you’re all-in on the turn. The dealer:

1. Burns and deals a river card — this decides the first half of the pot.
2. Returns that river to the deck, burns, and deals a different river — this decides the second half.

Each board is worth 50% of the pot. You can win both, split, or lose both. On the flop with two cards to come, each run-out deals two fresh cards from the remaining deck.

The EV proof: nothing changes

Let your equity be p (your true chance of winning) and the pot be P. Expected value is just probability times payoff.

Running it once:

EV = p × P

Running it twice (two independent run-outs, each for half the pot, each with the same equity p):

EV = (p × P/2) + (p × P/2) = p × P

Identical. Your equity p is the same on both boards because both are drawn from the same remaining deck, so each half-pot returns the same average as one full pot would. Adding two equal halves just reassembles the whole.

Where the variance actually drops

If the EV is identical, the benefit is entirely in the spread of outcomes. Consider a 70% favorite in a $1,000 pot, all-in on the turn:

Method	Possible results	Chance	Your payout
Once	Win	70%	$1,000
Once	Lose	30%	$0
Twice	Win both	49%	$1,000
Twice	Win one, lose one	42%	$500
Twice	Lose both	9%	$0

Run once, and it’s all-or-nothing: 70% shot at the whole grand. Run twice, and a middle outcome appears — 42% of the time you get exactly $500. The chance of the total wipeout drops from 30% to 9%. Same $700 average, but the extreme results are far less likely.

That compression is the entire point. Over a career the average is all that matters, but on any single big pot, reducing variance protects your bankroll from a brutal one-time swing.

A quick equity check

The two run-outs aren’t perfectly independent — the first board removes cards the second can’t use — but the effect on your average is nil, and on realistic all-ins the variance reduction is close to what the simple binomial above shows. What matters is that your equity on each half is the same, so the halves average out to your full-pot expectation.

• As a big favorite (say 80%+): you rarely lose both, so running it twice barely changes your near-certain win but shaves the ugly 20% disaster.
• As a coin flip (~50%): the split outcome becomes the most common single result, turning a nerve-racking 50/50 into a near-guaranteed chop of roughly half.
• As an underdog: same logic in reverse — you’re still expected to lose, just with less chance of the miracle full double-up.

When to run it twice

Because the EV is unchanged, the decision is about comfort, not profit:

• Run it twice when you want to lower the swing of a big pot and you’re playing at stakes where variance matters to you emotionally or for postflop planning reasons.
• Decline only if you simply prefer the gamble, or if house/game rules make it awkward — never because you think it changes your edge.

The takeaway

Running it twice is the rare poker decision with a proven answer: your expected value is exactly the same either way. The only lever it moves is variance, splitting one all-or-nothing gamble into two smaller ones and making the middle outcome likely. Understand it as a variance tool, not an edge, and slot it beside the rest of the poker odds and math you use every session.