How to calculate outs, equity, pot odds, and implied odds
A handful of numbers are the mechanical backbone of every drawing decision in poker. Get comfortable running them and a huge share of "should I call?" spots stop being guesses. We'll walk through one real hand start to finish, then break down why each number matters on its own — and finish with the assumption that quietly breaks all of them.
YOUR HAND
A♠K♠
BOARD
Q♠7♠2♦
You hold A♠K♠. The flop comes Q♠7♠2♦ — a flush draw. Pot is 10bb. Villain bets 15bb into you. Call or fold?
1. Outs
An out is a card left in the deck that turns your hand into (probably) the best hand. Counting them starts with identifying your draw, then counting how many cards of the suit or rank you need are still unseen.
outs = (cards of your suit/rank in a full deck) − (cards of that suit/rank you can already see)
There are 13 spades in a deck. You can see four of them: A♠, K♠, Q♠, 7♠ — two in your hand, two on the board. 13 − 4 = 9 outs. Any of the 9 remaining spades makes you the nut flush.
Clean vs. dirty outs: your Ace and King also outrank the whole board, so pairing either would likely give you top pair too. Real players often don't count that — a made pair is easy to lose to two pair, a set, or a made straight, and the same card that pairs you can pair the board into something a villain hit as well. Most of this guide assumes your outs are clean; section 7 is about the spots where they are not.
Why it matters: every other number below is built directly on top of this one. Miscount your outs and equity, pot odds, and implied odds are all wrong too — it's worth being careful here before anything else.
2. Equity (the rule of 2, and when to use 4)
Equity is your percentage chance of making your hand — but you have to be careful which question you are asking, because there are two of them and they have very different answers.
next card = outs × 2 (the card a bet buys) by the river = outs × 4 (flop, both cards — only if the money is already in)
With 9 outs: the turn is a spade 9 ÷ 47 = 19.1% of the time, and the rule of 2 gets you to 18%. By the river, counting both cards, it is 35%, and the rule of 4 gets you to 36%.
This is the mistake that costs the most money. The 35% is only a price you can pay in one go when you are all-in — when both cards are guaranteed to come for a single payment. If your opponent still has chips behind, calling their flop bet buys you the turn, and nothing else. They can bet again, and often will, exactly when you miss. So the number to hold against a flop price is 19.1%, not 35%.
Why it matters: equity is the number you compare against the price you're being asked to pay, and the two have to describe the same thing. One street's price against two streets' equity flatters every draw you will ever hold.
The outs and pot odds charts list the exact number beside the shortcut for every out count from 1 to 15, if you'd rather look it up than approximate.
3. Pot odds (the immediate price)
Pot odds tell you the minimum equity you need for calling to be profitable right now, based purely on the current pot and the size of the bet — no assumptions about future streets.
pot odds needed = call amount ÷ (pot after their bet + your call)
Pot was 10bb, villain bet 15bb, so the pot sits at 25bb before you act. If you call 15bb, the final pot is 40bb. 15 ÷ 40 = 37.5% — that's the equity you'd need for an immediately profitable call.
Calling buys you the turn, so the equity to weigh against it is 19.1% — against a 37.5% requirement. That is not close. On the immediate price alone this call is a long way short.
Careful with the by-the-river number here. It is tempting to hold 35% against 37.5% and call it marginal. It isn't: you would be paying a one-street price and counting two streets of outs. Do the algebra with the card you are actually buying and a nine-out draw is priced out by anything above about 0.31× pot — which is to say almost every real bet. That is why draws lean on implied odds rather than pot odds, and why the next section exists at all.
Why it matters: this is the single fastest gut-check in poker. Equity above the required number means call, full stop, no further math needed. Equity below it doesn't automatically mean fold — it means check the next number.
4. Implied odds
Implied odds account for the extra chips you can reasonably expect to win later if you hit, on top of what's already in the pot. It's what turns a borderline "no" on the immediate price into a real "yes."
breakeven total = (1 − equity) × call ÷ equity implied odds needed = breakeven total − (pot after their bet)
Using the 19.1% that one card actually gives you: breakeven total = (0.809 × 15) ÷ 0.191 ≈ 63.3bb in the pot by the time you'd see a showdown. The pot holds 25bb before your call, so you are short by about 38bb.
That is the real ask, and it is a demanding one: you need to extract roughly 38 more big blinds from this opponent on the turn and river when the flush comes in. Against a 100bb stack that is most of what they have left, and they will not pay it every time. Against an overbet this size, the honest read is that the call is bad unless you are deep and confident of getting paid.
If you ran this with the by-the-river 35% you would get a breakeven of 27.9bb and conclude you needed only about 3bb more — an easy yes. That is the same error as before, wearing a different hat: the 15bb you are paying buys one card, so the equity in this formula has to be one card's worth too. The gap between "I need 3bb" and "I need 38bb" is the whole reason this distinction matters.
Why it matters: implied odds are why good drawing players make calls that look thin on paper, and why the same call turns bad in a short-stacked or face-up spot where you're unlikely to get paid when you hit. Discount them honestly — assuming you'll always get max value is the most common way this number gets misused.
A different hand
Everything above assumes you're the one facing a bet. Flip the situation around — you hold the draw and act first — and there's a fifth number worth knowing.
YOUR HAND
J♥T♥
BOARD
9♥8♣2♥
You hold J♥T♥. The flop comes 9♥8♣2♥ — a flush draw plus an open-ender. Pot is 10bb, nobody's bet yet. You're deciding whether to bet 6bb as a semi-bluff, and you estimate villain folds to it about 40% of the time.
5. Semi-bluff EV
Betting a draw gives you two ways to win: villain folds right now, or they call and you still have your normal equity. The full value of the bet has to account for both.
EV if called = equity × (pot + bet) − (1 − equity) × bet semi-bluff EV = fold% × pot + (1 − fold%) × (EV if called)
This hand has 15 outs (flush + open-ender). If they call, you see the turn — not the river, because they can bet again — so the equity in this formula is 15 ÷ 47 = 31.9%, not the 54% this draw reaches by the river. EV if called = 0.319 × 16 − 0.681 × 6 ≈ +1.0bb: still profitable when called, but only just, and far from the +5.9bb the by-the-river number would have promised. Blending in the 40% fold equity: 0.4 × 10 + 0.6 × 1.0 ≈ +4.6bb. Positive, so betting still beats checking — and note how much of that now comes from the folds rather than the draw.
The catch: fold% isn't handed to you in a real hand — you have to estimate it from bet sizing, board texture, and reads, the same range-reading skill behind any bluff-catching decision. That's a fuzzier, experience-built skill, not something with a single correct number the way outs are. PotCheck's semi-bluff drill states the fold% for you so the EV math stays gradable; treat it as practicing the formula, not the read.
Why it matters: this is why aggression with a live draw usually beats passively calling or checking — you're not just hoping to hit, you're also getting paid immediately a real chunk of the time. It's also why the same bet is a mistake with a weak draw and no fold equity: check the numbers, don't just bet every draw on autopilot.
Working backwards
Every number so far answers "is this good enough?" Flip each one around and you get the break-even point itself — the price at which the answer changes. These are the numbers experienced players actually carry around, because they turn a calculation into a threshold you can recognise instantly.
6. The three break-even points
Biggest bet you can call. Pot odds told you the equity a given bet demands. Solve it the other way and you get the largest bet your draw can profitably call:
biggest callable bet = pot × equity ÷ (1 − 2 × equity)
A nine-out flush draw is 19.1% to hit on the next card, so into a 10bb pot it can call up to 10 × 0.191 ÷ 0.618 ≈ 3.1bb on the immediate price alone — less than a third of the pot, which is the same fact from section 3 seen from the other side. Note the denominator: as equity approaches 50% the ceiling runs off to infinity, because a coin flip can call anything.
Put 35% into that formula instead and you get 11.7bb — a draw that can call an overbet. It is a seductive number and it is wrong for any spot where money is still behind, because the 11.7bb is being paid for one card while the 35% counts two.
Fold equity a bluff needs. When betting loses money against a caller, fold equity has to make up the difference:
fold% needed = loss if called ÷ (pot + loss if called)
Bet 8bb into 10bb with a gutshot that loses 4bb when called, and you need 4 ÷ 14 ≈ 29% folds to break even. Below that the bet is burning chips no matter how much you like your draw.
Minimum defence frequency. The mirror image, from the defender's chair:
minimum defence = pot ÷ (pot + bet)
Against a half-pot bet you must continue with 10 ÷ 15 ≈ 67% of your range; against a pot-sized bet, 50%; against a double-pot overbet, only 33%. Defend less than that and your opponent can bet any two cards at a profit.
Use it as a sanity check, not a rule. MDF assumes your opponent is bluffing at a game-theory-correct rate. Most opponents are not — against someone who never bluffs you should fold far more than MDF says, and against a maniac, far less. It tells you what a bet is threatening, not what you must do.
Why it matters: thresholds beat calculations at the table. "This bet is over what my draw can call" is a decision you make in a second; recomputing pot odds every hand is not.
The assumption underneath
Everything so far has taken your outs at face value. That is the one assumption in this whole guide that regularly costs people money, so it is worth spending a section on the cases where it breaks.
YOUR HAND
K♥T♥
BOARD (TURN)
Q♣J♥2♣5♦
You hold K♥T♥ on Q♣J♥2♣5♦. One card to come. Pot is 20bb and villain makes a small stab of 4bb.
7. Dirty outs
A dirty out completes your hand and still loses. You have K-Q-J-T, so any ace or any nine makes a straight — eight outs, and the rule of 2 says 16%.
Except the board has two clubs. If the A♣ or the 9♣ arrives, that is a third club, and anyone holding two clubs has a flush that beats your straight. Those two cards make your hand and lose the pot. Eight raw outs, six clean — 12% by the rule of 2, not 16%.
Now the price. Villain's 4bb into 20bb means calling 4 to win 28, so you need 4 ÷ 28 ≈ 14.3%. Sixteen percent calls. Twelve percent folds. The dirty outs are the entire decision.
This is why small bets matter more than big ones here. Against a large bet you need so much equity that two outs either way rarely changes the answer. It is the cheap-looking bets, where you are close to the line, that punish a lazy out count.
Three ways an out goes dirty, all of them readable straight off the board:
it pairs the board → a full house is out there it brings a third card of one suit → a flush beats you it makes only the low end → a higher straight is available
That last one catches people. Holding 5-4 on 8-7-2-K, a six gives you 8-7-6-5-4 — and gives anyone with T-9 the higher T-9-8-7-6, and anyone with 9-5 a better one too. You hit your gutshot and you are still behind.
Where we draw the line. PotCheck's drill counts an out dirty whenever the board makes a better hand available — not weighted by how often villain actually holds it. That is deliberately strict: the A♣ is only truly dead when they have two clubs. We do it this way because "is a flush possible" is a fact about the board, while "do they have it" is a read, and this site does not pretend to read ranges. In a real hand, discount by what they can credibly have — which will land you somewhere between the raw count and ours.
Why it matters: every number in this guide is built on the out count, so this is the one input where being sloppy corrupts everything downstream. Counting eight when you have six does not make you 2 outs wrong — it makes your equity, your price, and your decision all wrong at once.
The full sequence
Count your outs → throw out the ones that don't actually win → convert to equity with the rule of 4 and 2 → compare to the pot odds you're being offered → if equity falls short, check whether realistic implied odds cover the gap. That's a complete call-or-fold decision, and it's what PotCheck's drills ask after every hand, with worked math so you can see precisely where your read went wrong. This is still a simplified model: it doesn't weigh opponent range or reverse implied odds the way a full equity calculator would — treat it as the mechanical foundation, not the whole picture.
Pot