Every equity number on our poker odds hub begins the same way: someone counted outs. It is the most basic skill in the arithmetic of the game, and the one most often done sloppily, because counting to nine feels too easy to get wrong. It is easy to get wrong. This page covers the count, the corrections, and the shortcut that turns a count into a percentage while the action is still on you.
What an out is
An out is an unseen card that, if dealt, makes your hand the likely best one. Hold two hearts on a flop with two hearts, and any of the remaining nine hearts completes your flush: 9 outs. Hold an open-ended straight draw, say 9-8 on a 7-6-2 board, and any five or any ten gets you there: 8 outs. A gutshot needs one exact rank, four cards: 4 outs.
The word "unseen" is doing work in that definition. You count from the cards you cannot see, which on the flop is 47: fifty-two minus your two hole cards minus the three on board. Folded hands do not enter the count. Yes, some of your hearts are probably in the muck, but you have no way to know which cards are, so the honest denominator is everything hidden from you. The math forgives ignorance as long as it is evenly distributed.
Counting clean
The count goes wrong in one direction: too high. Players tally every card that improves their hand and skip the question of whether it improves someone else's more. An out that helps your opponent beat your improved hand is a dirty out, and dirty outs get discounted.
The classic case is a straight draw on a two-tone board. Your 8 open-ender looks lovely until you notice that two of those 8 cards put a third flush card on the table. If anyone is drawing to that flush, hitting your straight on those cards may cost you a stack rather than win one. Count 6, and 6 is generous.
Same discipline with overcards. Ace-king against a probable pair has 6 outs in theory, but if the ace pairs the board where an opponent holds two pair, or your king makes someone's ace-king-queen run out, the real number is smaller. Against a flopped set, your overcards are close to worthless. When in doubt, knock off an out or two. An estimate that is honestly low beats a precise number built on wishful counting, and the discount rarely changes a clear decision anyway. It changes the close ones, which are the ones that pay for everything else.
The full chart
Turn column is the chance the next card helps (outs of 47 unseen). River column is the same for the last card (outs of 46). By-river is the chance at least one of the two arrives, which is the number for all-in situations on the flop.
| Outs | Turn | River | By river |
|---|---|---|---|
| 1 | 2.1% | 2.2% | 4.3% |
| 2 (pair to a set) | 4.3% | 4.3% | 8.4% |
| 3 | 6.4% | 6.5% | 12.5% |
| 4 (gutshot) | 8.5% | 8.7% | 16.5% |
| 5 (pair to two pair or trips) | 10.6% | 10.9% | 20.4% |
| 6 (two overcards) | 12.8% | 13.0% | 24.1% |
| 7 | 14.9% | 15.2% | 27.8% |
| 8 (open-ender) | 17.0% | 17.4% | 31.5% |
| 9 (flush draw) | 19.1% | 19.6% | 35.0% |
| 10 | 21.3% | 21.7% | 38.4% |
| 11 (open-ender + one overcard) | 23.4% | 23.9% | 41.7% |
| 12 (flush draw + gutshot) | 25.5% | 26.1% | 45.0% |
| 13 | 27.7% | 28.3% | 48.1% |
| 14 | 29.8% | 30.4% | 51.2% |
| 15 (flush draw + open-ender) | 31.9% | 32.6% | 54.1% |
The row worth staring at is 15. Past fourteen outs, your "draw" is a favorite over a bare made hand by the river, which reframes the whole street: the player with top pair is the one drawing thin against you across two cards. This is why strong players play their monster draws fast. They are betting a favorite, and getting called by hands that think they are ahead.
The rule of 2 and 4
Nobody divides by 47 mid-hand. Instead: with one card to come, multiply your outs by 2. With two cards to come, multiply by 4. Nine outs on the turn is about 18%, true value 19.6%. Nine outs on the flop, seeing both cards, is about 36%, true value 35.0%. Close enough to bet money on, which is the point.
Know the error margins, though. The times-2 half runs slightly low, since each out is really worth about 2.2 points, so add a point when a decision sits right on the line. The times-4 half is trustworthy through 8 outs and then drifts high, because it double-counts the runouts where both cards would have helped: at 15 outs it says 60% against a true 54.1%. The standard patch is to subtract the excess over 8. Fifteen outs: 60 minus 7 is 53. Twelve outs: 48 minus 4 is 44, against a true 45.0%. Patched, the rule stays within about a point across the whole chart, which is more accuracy than any live read you will ever make.
From equity to a decision
Equity by itself decides nothing. A gutshot at 8.7% on the turn is neither a call nor a fold until you know what the call costs: against a half-pot bet demanding 25%, it is a clear fold, and against a tiny probe laying 8:1 it can continue. The comparison between what you hold and what you are charged is the subject of the pot odds page, and the two pages are really one method split in half. Count, multiply, compare. If the middle step is the one that wobbles, wobble here first, where it is free.
Drill it
Counting outs is a perception skill as much as a math skill, and perception trains fast under repetition. The odds trainer deals draw after draw and grades your count and your estimate against the exact figures from the chart above. When the counts come without effort, add the showdown trainer to sharpen the other half of the judgment: knowing which improved hand actually wins the pot. Both cost nothing. The mistakes you make there are the cheapest poker education on the market.