Odds vs Probability: The Hidden Sports Betting Distinction

Every sportsbook number is a price. Understanding the difference between odds and probability is the key that unlocks everything else.

GhostLeg, 2026-06-28, 9 min

The gap between odds vs probability is the most misunderstood concept in sports betting. Most explainers teach you how to read odds but skip the more important question: what odds actually are.

Every major explainer (from NYT to Investopedia) walks you through American odds notation. Negative number means favorite, positive number means underdog. -110 means you risk $110 to win $100. Got it. Fine. But that explanation answers "how do I decode the number" while completely skipping: what is the number telling you about the real world?

The answer is that betting odds are not predictions. They are prices. And the translation layer between a price and a prediction is implied probability. Once you understand that translation, every other concept in sharp betting: devigging, expected value, closing line value. Each one falls into place. Without it, you're just reading numbers on a screen.

Odds vs Probability: The Core Distinction

Think of any outcome that hasn't happened yet. There's some real-world probability that it will, call it P. A sportsbook takes that P, converts it to American odds, and then bakes in a margin for themselves before posting the number you see.

The odds are not P. They are a price derived from P, plus the house's cut.

When a book posts the New York Knicks at -136 on the moneyline, they're not saying "the Knicks have a 57.63% chance to win." They're saying "we need you to risk $136 to win $100, because our pricing (which includes our margin) implies approximately that probability." The true probability they're estimating could be 55%. Could be 57%. The posted line hides it.

This is the distinction that matters: implied probability (what the odds say) versus true probability (what the market actually believes). The vig, the overround, the juice: all of those words refer to the same thing, the gap between the two.

How to Convert Any Moneyline to Implied Probability

The math is two formulas, one for favorites and one for underdogs. Commit these to memory.

Favorite (negative American odds):

Implied probability = |odds| / (|odds| + 100)

Underdog (positive American odds):

Implied probability = 100 / (odds + 100)

No exotic math. Let's run it on a real game.

A Live NBA Playoff Example: Knicks vs Cavaliers

GhostLeg tracks moneylines across seven sportsbooks in real time. Here's what the Cavaliers vs Knicks playoff game (May 26, 2026) looked like in our odds table as of the morning of the game, with Pinnacle as the reference:

Start with Pinnacle, because Pinnacle is the tightest market. They post the sharpest lines in the industry and consistently have the lowest vig. That's why you always devig against Pinnacle.

Step 1: Calculate raw implied probabilities from Pinnacle:

Knicks at -136: 136 / (136 + 100) = 136 / 236 = 57.63%

Cavaliers at +120: 100 / (120 + 100) = 100 / 220 = 45.45%

Add them up: 57.63% + 45.45% = 103.08%

There's your problem. Probabilities for an event with two outcomes have to sum to 100%. The extra 3.08% is Pinnacle's margin, the vig. Even the sharpest book in the market doesn't offer true-probability odds.

Step 2: Devig to get the true probability:

To strip the vig out, divide each side's implied probability by the total:

Knicks true probability: 57.63% / 103.08% = 55.91% Cavaliers true probability: 45.45% / 103.08% = 44.09% Check: 55.91% + 44.09% = 100.00% ✓

That's what Pinnacle actually thinks: the Knicks are roughly 56% to win this game. Not 57.63%, which is what the raw line implied. The difference is vig.

What Do Odds of -200 Mean? (And Why the Standard Answer Misses the Point)

The standard answer: -200 means you risk $200 to win $100.

The complete answer: -200 implies a 66.7% win probability (200/300), which becomes roughly 64-65% true probability after you strip the vig. The real question is whether the book's 65% probability estimate matches reality.

That's where betting decisions actually live. If you think the true probability is 70% and the market is pricing it at 65%, you have a potential edge. If you think it's 62%, you should pass. The number on the board is a claim about probability. Your job is to assess whether that claim is accurate.

This reframe is exactly what our intel dashboard surfaces for every game: not just the odds, but whether those odds look efficient or mispriced relative to our model's probability estimates. The data underlying those comparisons is also available programmatically through GhostLeg's Data API for anyone building their own models.

Why Retail Books Are Worse Than Pinnacle (The Vig Math)

Look at the Knicks-Cavaliers numbers again. Compare DraftKings to Pinnacle:

DraftKings: Knicks -130, Cavaliers +110

• Knicks implied: 130 / 230 = 56.52%
• Cavaliers implied: 100 / 210 = 47.62%
• Sum: 104.14%, a 4.14% overround

Pinnacle: Knicks -136, Cavaliers +120

• Sum: 103.08%, a 3.08% overround

That 1.06% difference might sound trivial. Over 500 bets, it compounds into a meaningful drag. Pinnacle's lower vig is why sharp bettors use it as a reference price and why devigging against Pinnacle gives you the most accurate picture of where the market actually thinks probability sits. The sibling post How Do Sportsbooks Actually Set Their Odds? explains in more detail why Pinnacle posts first and why retail books shadow their prices.

How Betting Odds Work: The Market Feedback Loop

Here's where it gets interesting. The odds at tip-off are almost never the same as the opening odds from 48 hours ago. Markets move because sharp money (professional bettors and syndicates) bets the sides they believe are mispriced, pushing the line toward true probability.

This is the feedback loop:

1. A book opens a line based on their probability estimate
2. Bettors who disagree bet the other side
3. The book adjusts the line to balance their exposure and incorporate the new information
4. By tip-off, the line reflects the consensus of thousands of bettors, including the sharpest ones

The closing line, that end result, is the most accurate implied probability the market produces. It's why closing line value is the gold standard for measuring whether you're actually getting good prices: if you consistently beat the closing line, you're consistently finding mispriced probability before the market corrects it.

GhostLeg's intel dashboard tracks line movement in real time so you can see when the market is moving sharply on a game, a signal that someone with information is correcting the implied probability. That same line-velocity and movement data is available via the Data API for teams building automated pricing models.

Betting Odds Explained: The Three Market Structures

Not every betting line works the same way, and the vig profile differs by market type.

Moneylines are the cleanest. Two outcomes, two prices, one devig calculation. Straight application of the formulas above.

Point spreads are priced close to -110/-110 by design. The spread exists to manufacture a 50/50 proposition, pricing each side at roughly 50% implied probability with vig layered on top. When a spread moves from -110 to -115, the book is signaling that the market is imbalanced toward one side, not that the true probability has shifted dramatically.

Totals (over/under) work identically to spreads, priced around -110/-110 to split money evenly, with the number adjusted when money comes in heavy on one side.

Each of these is a probability market. Each carries vig. The underlying math is always the same formula.

The Bridge to Every Other Sharp Concept

Understanding odds as probability prices unlocks a chain of ideas:

Expected Value (EV): If you think the Knicks' true probability is 60% but the market prices it at 55.91%, your expected value is positive. You've found a mispriced probability. The expected value betting formula walks through the full EV calculation, including how GhostLeg applies it across 5,000+ graded legs.

Devigging: You just did it above. It's the mechanic that converts a posted line into a fair-market probability estimate.

Closing Line Value: Comparing the probability you bought at your bet price vs. the probability implied at closing time. If you bet Knicks at -134 (FanDuel) and the line closes -145, you paid 56.3% implied and the market settled at 59.2%. You beat the closing probability. That's positive CLV.

Vig comparison: Understanding that different books run different overrounds lets you shop lines intelligently. The book with the lowest vig on a given market is giving you the closest price to true probability. We cover the vig math in sports betting in full, including real 30-day hold data across six books, in the companion post.

Bankroll sizing: Once you know your true edge (the probability gap after devigging), the bankroll management framework tells you exactly how much to stake. The Kelly criterion uses the same devigged probability estimate you just computed; so this formula is literally the first step of proper bet sizing.

All of this is visible, updated, and tracked in the GhostLeg intel dashboard. The raw odds data feeding those dashboards: Pinnacle's lines, our devigged probability estimates, and line movement history, is licensed through the Data API for builders who need it in their own systems.

One More Thing About Reading Odds

The biggest mistake new bettors make is letting the formatting distract them.

-136 looks huge. +120 looks like a gift. Neither feeling is useful. What matters is the probability gap between what you think and what the market thinks, and whether the vig on the line still leaves you with positive expected value after the house takes its cut.

The number on the board is a claim. Every time you place a bet, you're claiming you know something the market doesn't. That's why understanding how odds translate to probability isn't a trivia exercise: it's how you know whether your claim has any basis.

If you want to see this in practice rather than theory, the intel dashboard shows current moneylines, our model's implied probabilities, and the devigged market consensus for every game we cover. For more background on why sportsbooks price the way they do, read How Do Sportsbooks Actually Set Their Odds?; it covers the mechanics behind why Pinnacle posts first and why retail books follow.

For entertainment purposes only. Past performance does not indicate future results.