NBA Odds Explained — Fractional, Decimal, American & Implied Probability

Every email I get from a new NBA punter eventually contains the same sentence: “I don’t understand why the price keeps shifting.” The honest answer is that the price isn’t really shifting — it’s being translated, compared and shaved by margin, and once you can see those three layers underneath the number, NBA odds explained as a topic stops feeling like a foreign language. I’ve been on this beat for nine years, and the punters who never quite click into profitability all share one habit: they read a price without ever asking what it actually says.

The UK online betting market is big enough now that this matters more than it used to. Online real-event betting alone produced £596 million in gross gambling yield in the January-to-March quarter of 2025, up about five percent on the year before, and a meaningful slice of that growth is in mobile basketball wagering where the price flickers in real time. If you can’t decode the price on the screen quickly, you’re playing a game whose scoreboard you can’t read.

This guide does three things and only three. It puts fractional, decimal and American odds side by side and shows you which format actually helps you make a decision. It converts those formats into the only number that matters — implied probability — and then strips out the bookmaker’s margin so you can see the fair price hiding underneath. And it shows where the differences between operators come from, why the same NBA line can sit at three different prices across three apps, and how to use that gap intelligently rather than randomly. The maths is light. The discipline is not.

One small note on style. Through this piece I use decimal as the default because the arithmetic is easier on the fly, and I switch into fractional or American where the format makes the point clearer. Pick the one your brain prefers — there’s no right answer, just a familiar one.

Fractional, decimal and American: three dialects of the same price

The fastest way to understand fractional decimal american odds is to look at one bet expressed three ways and notice that the maths underneath is identical. Take a moneyline of 4/5 on a Boston home game. Fractional says you win four pounds for every five risked. Decimal calls that price 1.80 — multiply your stake by 1.80 to get your total return, including stake. American odds, which UK books increasingly show on NBA markets because the sport is built around the format, translates to -125: you’d risk £125 to win £100. Same bet. Same fair value. Three numbers that look completely different.

Fractional is the format I grew up with on a Saturday football coupon, and it’s still how UK books default-render their main page. It tells you the ratio of profit to stake — 5/2 means a fiver back for every two staked, plus your stake — and the format is fine when you’re choosing a bet, slow when you’re calculating returns and useless when you’re trying to combine multiple selections in your head. Try multiplying 11/4 by 9/5 at speed and you’ll see what I mean.

Decimal is the format I use for every actual calculation. It’s a multiplier — a £20 stake at 2.45 returns £49 in total. To work out profit, subtract one and multiply: 2.45 - 1 = 1.45, times £20 is £29 profit on a £20 stake. Implied probability is a simple division: 1 / 2.45 = 0.408, or 40.8 percent. Combine four selections in an accumulator? Multiply the decimals together. Nothing else gets close for usability, which is why every odds-comparison tool and every modern app shows you decimal somewhere on the slip.

American odds are the format you’ll see on every prop menu and most live trackers, because the NBA’s home market runs on this dialect. Positive numbers tell you the profit on a £100 stake — +155 means £155 back for £100 risked. Negative numbers tell you what you’d need to risk to win £100 — -180 means £180 staked to get £100 in return. Positive is underdog, negative is favourite, and the closer to zero on either side, the closer to a coin flip. The format is awkward at first, fluent after a fortnight, and indispensable on a US-style prop menu where everything else has been translated except this.

The practical recommendation: set your sportsbook account to display decimal as the primary format and learn American as the secondary dialect for props. Leave fractional for the slip preview, where the format genuinely helps a UK eye sanity-check whether the operator has settled at the right price.

Converting between formats without a calculator

I’ll spare you the textbook tables. Here’s the only handful of conversions I actually do in my head, and the rest I let a calculator handle without shame.

Decimal to fractional is the cleanest. Subtract one, then express the remainder as a fraction. 2.50 minus one is 1.50, which is 3/2. 1.80 minus one is 0.80, which is 4/5. 3.40 minus one is 2.40, which is 12/5. The reverse — fractional to decimal — divides the top by the bottom and adds one. 5/2 is 2.5 + 1 = 3.50. 9/4 is 2.25 + 1 = 3.25. Two operations, no notebook.

Decimal to American splits into two cases. If decimal is 2.00 or above, the underdog rule applies: multiply the profit portion by 100. 2.50 becomes +150; 3.40 becomes +240. If decimal is below 2.00, the favourite rule applies: divide 100 by the profit portion and add a minus sign. 1.80 means profit per unit is 0.80, so 100 / 0.80 = 125, giving you -125. 1.50 means profit is 0.50, so 100 / 0.50 = 200, giving you -200. The minus on a favourite always tells you the stake required to win £100.

American back to decimal mirrors that logic. A positive American number divided by 100 plus one gives decimal: +150 is 1.50 + 1 = 2.50. A negative American number is converted by dividing 100 by the absolute value and adding one: -200 is 0.50 + 1 = 1.50. The two rules feel awkward for a fortnight and then they don’t.

The conversion that catches more punters than any other is -110. This is the default juice on a standard NBA point spread, and it converts to decimal 1.91 or fractional 10/11. Memorise that one. You’ll see it on roughly half the spreads you ever look at, and being able to read -110 as “I have to win 52.4 percent of these to break even” lives at the centre of every break-even calculation in basketball betting.

One last conversion habit I recommend: when you’re considering a parlay or builder, convert every leg to decimal first, multiply them together, then convert the result back to whichever format your eyes find easiest. The compounded price in fractional looks ugly — 21.85/1 is not a number anyone wants to read — but in decimal it’s just 22.85, and you can see immediately whether the combined price feels right relative to your sense of the legs.

From odds to NBA implied probability in one operation

The single most important number in any betting decision isn’t on the screen. It’s hiding behind the price, and you have to extract it before you can decide whether the bet is good. That number is the implied probability — the bookmaker’s assessment of how often this outcome occurs, expressed as a percentage between zero and a hundred.

From decimal odds, implied probability is one operation: divide one by the decimal. A price of 2.00 implies a 50 percent probability. 1.50 implies 66.7 percent. 3.00 implies 33.3 percent. 1.91 — your default NBA spread price — implies 52.4 percent. That last figure is the break-even point for any bet at standard juice. If you can’t realistically beat 52.4 percent over a meaningful sample, you can’t beat the standard NBA spread market in the long run. The maths is not optional.

From fractional odds, divide the denominator by the sum of numerator and denominator. 4/5 becomes 5 / (4+5) = 0.556, or 55.6 percent implied. 11/4 becomes 4 / (11+4) = 0.267, or 26.7 percent. From American, the positive-number rule is 100 / (American + 100): +150 is 100 / 250 = 0.40, or 40 percent. The negative-number rule is -American / (-American + 100): -180 is 180 / 280 = 0.643, or 64.3 percent.

The reason I drill these is that you can’t possibly evaluate a bet without translating it into a probability your own brain can grade. When a friend offers me a tip on a Lakers -7.5 at -110, my next sentence is always “what’s your probability they cover by eight or more?”. If they can’t say “55 percent” or “59 percent” or “I genuinely don’t know”, they don’t have a bet — they have a vibe. Implied probability turns every betting argument into a comparison of two estimates: yours and the market’s. If yours is higher, the bet is at least defensible. If it isn’t, walk away.

One subtlety worth knowing. The implied probability on the screen is not the bookmaker’s true estimate — it includes the margin. The “true” or “fair” probability is always lower than the displayed implied probability for both sides of a two-way market, because the gap between the two figures is where the operator’s hold lives. Stripping that gap out gets you closer to fair value, which is the next step and the one almost no casual punter bothers with.

A small drill that pays for itself. Pick any NBA spread tonight, in any format, and convert it to implied probability before you place a bet on it. If you wouldn’t take that probability against a friend with no money on the line, don’t take it from a bookmaker who has margin to spare.

Overround, vig and where the bookmaker’s margin actually lives

The first time a trader explained overround to me, he drew two columns on the back of a napkin: implied probabilities for each side of a two-way market, summed at the bottom. The total was 104. I asked why it wasn’t 100. He laughed and said “that four is my mortgage”.

Overround — known interchangeably as the vig, vigorish, juice, hold percentage or simply margin — is the amount by which the implied probabilities of all outcomes in a market sum above 100 percent. On a perfectly fair NBA moneyline with no margin, the two implied probabilities would add to exactly 100. In reality, they add to 104, or 105, or sometimes 108 on softer markets. That extra slice is the bookmaker’s structural advantage. You can win bets against the price; you can’t win bets against the price plus the slice — at least not over a long enough sample.

On a standard -110 / -110 NBA point spread, both sides imply 52.4 percent, summing to 104.8 percent. The overround is 4.8 percent. On a -115 / -105 spread, the sides imply 53.5 percent and 51.2 percent, summing to 104.7 percent — almost the same overround, distributed slightly differently. On a moneyline with a heavy favourite, say -300 / +240, the sides imply 75 percent and 29.4 percent, summing to 104.4 percent. The overround is structurally similar across markets; what varies is how the book chooses to distribute it.

The trajectory is worth tracking. Average hold percentage at US sportsbooks rose from 6.7 percent in 2018 to over 9 percent across the 2024-2025 period, and that drift is partly product-mix — more props, more builders, fatter margins on each — and partly operators realising the recreational customer doesn’t shop prices. UK books haven’t lifted hold quite that aggressively because the comparison-shopping culture among UK punters is sharper, but the direction of travel is the same. Hold is a quiet problem because it doesn’t show up on any single bet; it shows up after a thousand bets, when the punter realises they had a 51 percent win rate on their spreads and somehow still lost money.

The practical implication is brutal. If the spread market carries a five percent overround, you don’t need to win 50 percent of your bets to break even — you need to win roughly 52.4 percent. The market doesn’t open the gates for you at coin-flip rates; it demands a real edge. And that edge has to overcome compounding hold every time you place a bet, which is why bet selection matters more than bet size and why most casual punters lose slowly even when their conviction is roughly right.

The skill, then, isn’t beating the line. It’s beating the line by enough to clear the slice the book has already taken before you sat down.

Removing the vig to see the fair line underneath

Strip the margin out of any market and what’s left is the bookmaker’s actual probability assessment. This is the no-vig price, and it’s the number I anchor every bet decision on. Without it you’re comparing your view to a price; with it you’re comparing your view to a probability, which is the only honest comparison.

The simplest method is the proportional approach. Take a moneyline of -180 / +160. The implied probabilities are 64.3 percent and 38.5 percent, which sum to 102.8 percent overround. To strip the vig, divide each implied probability by the sum: 64.3 / 102.8 = 62.5 percent and 38.5 / 102.8 = 37.5 percent. Those two figures sum to a clean 100. The favourite’s fair implied probability is 62.5 percent, not the 64.3 percent the price implied — meaning the fair decimal odds would be 1 / 0.625 = 1.60, not 1.56.

That gap between 1.56 and 1.60 is exactly what you’re paying for the privilege of betting. If you genuinely think the favourite wins 65 percent of the time, the displayed 1.56 is a marginal value bet — you implicitly disagree with the book by 0.7 percent. If you think the favourite wins 64 percent of the time, the bet has negative expected value once you account for the margin. The decision pivots on tenths of a percent, which is why doing the calculation matters and why feel doesn’t.

For NBA spreads at -110 / -110, the no-vig fair price for both sides is 50 percent / 50 percent, with fair decimal odds of 2.00. Anything you pay above 2.00 on either side of a balanced spread carries the operator’s full margin against you. This is why pricing differences between bookmakers matter so much on spreads — a single half-point of line difference or a price of -105 instead of -110 can flip a bet from negative expected value to neutral or marginally positive.

For three-way markets — race-to-twenty, first basket scorer, certain prop tier markets — the proportional method works the same way. Add up the implied probabilities of all three outcomes, divide each by the sum, and you have the no-vig estimate for each. For markets with twenty or more selections, like an outright on the championship, the proportional method introduces some distortion because it assumes the bookmaker spreads margin evenly across all selections, which they don’t — favourites typically carry more proportional juice than long-shots. The advanced fix is called Shin’s method, and it adjusts for the fact that bookmakers shade long-shots away from fair odds. For a quick desktop sanity check, proportional is good enough.

The habit to build is simple. Before you click “place bet”, convert both sides of the market to implied probability, sum them, calculate the no-vig fair price for your side, and compare that to your own estimate. If your estimate is meaningfully above the no-vig figure, you have a value bet. If it’s at or below, you have a bet that requires luck rather than skill to win.

Why the same NBA line costs different amounts at different books

I keep five UK sportsbook apps on the same iPhone home screen, and on any given NBA evening I’ll see the same primary spread at five subtly different prices. That’s not malfunction — that’s competition, model variance and margin policy stacking on top of each other.

The first reason for the gap is independent pricing models. Each operator runs its own algorithm to estimate fair odds, weighted with adjustments for sharp money, public exposure and recent line movement across the market. Two models built on different feature sets will reach slightly different conclusions about the same NBA game, and those differences live in the half-points and the juice on either side. A Lakers -7.5 (-110) at one book might be -7.5 (-115) at another and -8 (-105) at a third. All three describe the same view of the game with different margin distributions.

The second reason is exposure management. Operators take very different volumes on different teams — a London-based punter base will skew towards certain franchises, which means the book has more liability on those names and will shade the price slightly to discourage further action. This is rare on small-market games and obvious on prime-time fixtures involving the Lakers, Celtics, Warriors or whichever team the casual UK audience has gravitated to that month.

The third reason is the broader online market continues to grow. The whole UK remote sector posted £1.42 billion in online gross gambling yield in Q2 2025, an eight percent year-on-year increase, which means competition for the recreational punter is intense and books are willing to lead with sharper prices on highlighted markets to win the click. The corollary is that sharpness varies — a sharper price on the main NBA market often comes with worse juice on the props menu, because the book has to make the margin somewhere.

Practically, this means line shopping is the highest-value habit a UK NBA punter can build. Even a single half-point on a spread, taken systematically across a season of forty or fifty bets, swings the long-run result meaningfully. There’s a structured walkthrough of how to do this efficiently in the step-by-step line shopping for NBA odds across UK bookmakers guide — the workflow is dull but it’s the closest thing to free money the betting market offers.

The one warning is don’t shop yourself into nonsense. If you find one book consistently priced ten percent off the rest, the question to ask is not “which book is right” but “what does my book see that I don’t”. Outlier prices on NBA markets are almost always reactive, not predictive — they reflect liability or a missing piece of news rather than a smarter model.

Mobile and desktop pricing: subtle but real differences

Until two years ago I’d have told you it didn’t matter which device you bet from. Now I’d hedge — and the hedge is worth knowing about.

The first practical difference is speed of price refresh. Mobile apps tend to update live prices on a slightly different cadence than the web product, and on fast-moving in-play markets a half-second of latency can move a bet from +115 to +110 or worse. For pre-match bets this is irrelevant — the price you click on is the price you’ll get — but for in-play wagers on NBA quarters and totals, mobile latency is the most common reason a bet “the price has changed” prompt appears between tap and confirm.

The second is the user-experience nudges that mobile apps lean into more aggressively. Bet builders, same-game parlays, request-a-bet — all of these are designed for thumbs, and the suggestion engine is sharper on the mobile product than on the desktop equivalent. Roughly 78 percent of online sports bets globally were placed via mobile devices in 2024, so the operators have correctly identified where to focus their friction reduction. The downside is that the same friction reduction makes it easier to size a bet wrong, leg up a builder without thinking, or take a price that’s worse than you’d accept on a laptop where you can see the full slip at once.

The third difference, and the one most punters miss, is that the boosted-price markets and the request-a-bet markets are more aggressive on mobile. The headline odds boost on a Lakers same-game builder will appear on the app’s home screen, while the equivalent desktop view buries it three menus deep. Books are not being charitable with these boosts — they’re advertising margin in a different shape — but it does mean genuine value is sometimes accessible on the app that you’d miss entirely on the web.

The recommendation I give friends moving from football to NBA is simple. Build your slips on desktop for any bet you’ve thought about for more than two minutes, and use the app for opportunistic in-play and for the times you’ve already done the analysis and just need to pull the trigger. Mobile-only NBA betting is fine; mobile-first NBA betting is where people end up surprised by how their balance evaporated over a season.

Common odds questions UK punters keep asking

Below are the questions I get most often after walking through this material the first time. The thread running through them is the same: the format isn’t the story, the probability is. Grainne Hurst, who runs the Betting and Gaming Council, framed the broader industry version of this responsibility well when she said the latest figures “underline the industry’s ongoing commitment to raising standards and ensuring the millions of people who enjoy a regular flutter do so in a safe and responsible environment”. That sentence is corporate by design, but the underlying point is right — the punter has to be able to read the price they’re being offered, and a regulated market makes that legible in ways an unregulated one doesn’t.