Pot Odds Street by Street
The formula never changes, but what a call buys you does: two cards on the flop, one card on the turn, and pure showdown value on the river. Follow a single hand through all three prices and watch the same percentage mean three different things.
Assumptions: Examples use a 100bb-deep 6-max online cash game with no rake unless a different setup is stated.
The pot-odds machine you've built — count the final pot, divide your call into it — runs identically on the flop, the turn, and the river. What changes street to street is not the math but the merchandise. A flop call purchases two more community cards and all the betting that surrounds them. A turn call purchases exactly one card. A river call purchases nothing but the showdown itself. Three required-equity numbers that look almost identical can therefore describe three completely different decisions, and players who treat "I need 28%" the same way on every street leak in both directions. This lesson follows one hand through all three streets to make the differences concrete.
The hand: you're on the button with 8♥7♥ at $0.50/$1. The CO opens to $3.75, you call, both blinds fold. Pot: $9.00 even.
The flop: a price on two cards — with an asterisk
CO continuation-bets $6 into $9. Final pot if you call: 9 + 6 + 6 = $21. Required equity: 6/21 = 28.6% (tool: 2.5-to-1). So far this is last lesson's arithmetic.
Now the equity side, because on the flop "how often do I win" means "how often do I get there by the river." Your 8♥7♥ on T♠9♦2♣ is an open-ended straight draw: four jacks and four sixes complete it, eight outs. The exact two-card probability, computed by enumeration: 1 − (39/47)(38/46) = 31.45% — about 31%, with the rule-of-4 shortcut estimating 32%. Against a 28.6% bar, 31% clears.
But notice what that comparison quietly assumed: that calling the flop bet buys you both cards. It does so only if no more betting happens — say, the flop bet put one of you all-in. In a normal deep-stacked hand, the turn usually brings another bet, which means your flop call strictly buys one card plus a position in a future decision. Whether that future is good news (you hit and win extra) or bad news (you miss and face another price) depends on material this module doesn't cover — the Implied Odds module takes exactly this question as its starting point. For now, hold the honest version: flop pot odds priced against two-card equity are exact only when the flop call closes the betting; otherwise they're an approximation that future streets will revise. It's a serviceable approximation here — 31% versus 29% — and you call.
It's worth seeing both edges of that approximation, because they bracket the truth. The generous reading uses two-card equity: 31.45%, comfortably over the bar, exact if the $6 had been an all-in. The stingy reading prices the flop call as buying only the turn card: the draw arrives on the turn just 8/47 = 17.02% of the time (tool-computed), nowhere near 28.6% — exact if you knew you'd face an unaffordable turn bet every time you missed. Reality lives between the brackets: sometimes the turn checks through and your one call really did buy two cards; sometimes you hit and win more than the current pot; sometimes you miss and pay again or surrender. Where between 17% and 31% your true prospects sit depends on stack depth, position, and your opponent's barreling tendencies — which is why deep-stacked flop play can't be solved by pot odds alone, and why the next module exists. What pot odds do fix precisely is the boundary case: the shorter the remaining stacks, the closer the flop call sits to the all-in case and the more the clean 31.45%-versus-28.6% comparison governs. Watch for that convergence in tournament contexts especially, where shallow stacks make flop prices nearly exact.
The turn: a price on exactly one card
The turn is the 2♦, pairing the bottom card and changing nothing for your straight draw. CO bets $15 into $21. Final pot: 21 + 15 + 15 = $51. Required equity: 15/51 = 29.4% (tool: 2.4-to-1).
Look how little the price moved — 28.6% on the flop, 29.4% now. A bettor using similar fractions of the pot quotes you similar percentages on every street; that's the sizing-relative principle from the first lesson doing its job. But the merchandise collapsed. A turn call buys one card, full stop, and one card hits an 8-out draw exactly 8/46 = 17.4% of the time (tool-computed; the rule-of-2 estimate says 16%). There is no asterisk this time, no "and then the river might…" — after the river card, the hand's cards are fixed. One card, one exact number.
Seventeen against a bar of twenty-nine isn't close. By direct pot odds, this is a fold by a mile, and you should feel the asymmetry between the streets: the same draw, facing nearly the same percentage, went from a call to a clear mathematical fold purely because half its merchandise expired. Players who learned "open-enders are calls" as a flop rule and apply it on turns are paying 29% prices on 17% chances over and over.
Is folding the final answer? In a vacuum, yes. The full answer in real games is "it depends on the money behind": when you spike a six on the river, CO may pay off a big bet, and those unbet chips can subsidize today's bad price. Pricing future money is the entire subject of the Implied Odds module — flagged here, taught there. What pot odds alone can tell you is the part you just computed: the chips currently in the middle do not pay for this call.
Pause on the bettor's side of this street for a moment, because the symmetry is instructive. CO bet $6, then $15 — roughly the same two-thirds-to-three-quarters fraction of a pot that more than doubled between streets. Consistent fractions produce escalating amounts: the $9 pot became $21 became $51, and a bettor who keeps quoting you 28-29% is nonetheless tripling the chips at stake with each street. This is the squeeze the turn applies to draws. Your price barely moved, your one-card chance got cut nearly in half relative to the flop's two-card figure, and the absolute cost of continuing jumped from $6 to $15. Multi-street betting is designed — sometimes deliberately, by players planning their sizings across streets — to keep each individual price looking reasonable while the cumulative cost of chasing grows geometrically. The defense isn't outrage; it's exactly what you did here: reprice every street from scratch and let the merchandise, not the momentum, decide. Suppose, though, for the sake of the next street, that you talk yourself into calling anyway.
The river: a price on the truth
The river is the K♣. Your draw busted: the board reads T♠9♦2♣2♦K♣, and your best five cards are the board's own K, T, 9, 2, 2 — eight-seven plays nothing. CO bets $30 into $51. Final pot: 51 + 30 + 30 = $111. Required equity: 30/111 = 27.0%, about 2.7-to-1 (tool-confirmed).
On the river, "equity" sheds its probabilistic costume. No cards are coming; nothing can improve or decline. Required equity here means exactly this: out of every hundred times CO bets this river like this, do at least twenty-seven of his hands lose to mine at showdown? That's a question about his betting range — how many busted draws and bluffs it contains versus value — and answering it well is the art of bluff-catching, taught in the hand-reading modules.
But your hand makes the question vanish. Eight-high beats no value bet, obviously — and here is the detail beginners miss — it doesn't beat the bluffs either. CO's busted draws (Q-J that missed, ace-high floats) all outkick your 8♥7♥ at showdown or chop the board's pair-plus-kickers with you at best. A bluff-catcher needs to actually beat bluffs. When your hand wins zero percent of showdowns, every price from 1% to 49% is equally irrelevant. Fold, and notice that the river is the one street where the pot-odds number, by itself, can be completely overridden by your cards — there's no draw to redeem a hand that can't win a showdown it's being offered.
To feel what a real river decision at this price looks like, swap your hand: imagine you'd taken this line with J♠T♠ instead, and the same K♣ river completed the board. Now you hold a pair of tens with a jack kicker — a hand that loses to every value bet CO makes but beats his missed draws and random barrels. The 27% bar suddenly has teeth: your decision becomes a census of CO's $30-bet range. If roughly three bets in ten are bluffs your tens beat, calling exactly breaks even; more bluffs than that and the call prints, fewer and it burns. Notice the inputs: the price came from arithmetic you can do perfectly today, while the bluff count comes from hand-reading you'll build across later modules. Every river call you ever make will be this same two-part sentence — "I need X%, and I believe his range gives me Y%" — with pot odds supplying the X every single time. That's the division of labor this module establishes: the X is now permanently yours.
Three streets, one formula, three meanings
Line up the hand's three prices and what each one bought:
| Street | Bet into pot | Required equity | What the call buys | Your relevant chance |
|---|---|---|---|---|
| Flop | $6 into $9 | 28.6% | Two cards* + future betting | 31.45% (two-card, 8 outs) |
| Turn | $15 into $21 | 29.4% | Exactly one card | 17.4% (one-card, 8 outs) |
| River | $30 into $51 | 27.0% | Showdown only | ~0% (plays the board) |
*exact only if no further betting; otherwise revised by future streets.
Notice also what the table implies about where mistakes live. Flop pricing errors are the cheapest, both because the bets are smallest and because two cards of possibility forgive marginal misjudgments. Turn errors are the expensive habit: the bets have grown, the merchandise has halved, and the flop's permissive math tempts players into carrying calls forward on momentum. River errors are the most violent per instance — the bets are biggest — but also the most preventable, because nothing is hidden: no odds to weigh, no cards to come, just your hand against a range. Audit your own play in that order. Most losing players' pot-odds leak is not the flop; it's paying turn prices with one-card chances.
The street lens works just as hard when you're the one betting. Your sizing charges every draw in your opponent's range a price, and the value of what you're selling changes by street exactly as above: a flop bet sells two cards of possibility, so even a stiff price leaves draws with defensible calls; a turn bet sells one card, so the same fraction of pot that was a fair flop price becomes a toll most draws can't pay correctly. This is why disciplined players often save their biggest sizings for turns — not superstition, but the street where the merchandise-to-price gap punishes chasers hardest. When you barrel a turn and a draw-heavy opponent calls anyway, one of two things is true: they have more than a bare draw, or they just paid 29% for a 17% chance and you've been handed their win rate. Either way, you want to know which streets your opponents misprice — and you find out by pricing every street yourself, from both sides of the bet.
The percentages sit within three points of each other; the decisions were a call, a clear fold, and a trivial fold. That spread is the lesson. Pot odds are a price tag, and a price tag means nothing until you know what's in the box: on the flop the box holds possibility (and an asterisk about money to come), on the turn it holds one card's worth of exact probability, and on the river it holds only the truth about who already won. Run the count-and-divide on every street — it never stops being step one — and then ask the street's own question: both cards or one? cards at all, or just a showdown? The next lesson makes step one fast enough to do mid-hand; the Implied Odds module picks up the asterisk.