The bet, in one paragraph
A spread is the bookmaker's way of turning an unbalanced game into a coin flip. By assigning a handicap — usually three to fourteen points in football, two to twelve in basketball — the book makes the favorite earn its price by winning by more than the number, and rewards the underdog for keeping the game close. Both sides typically price at -110, which means a $110 risk to win $100 either way. The mechanic was invented by Charles K. McNeil in Chicago in the 1940s and rewrote American sports betting. Before McNeil, books offered odds; after McNeil, books offered margins. The spread became the central organising idea of the industry.
How a spread is set
A spread is the output of three inputs: a model, a market, and a margin. The model is the book's internal probability estimate of the final score margin. The market is the expected behaviour of the public — books deliberately shade lines toward the side they expect casual money to back, so that when sharp money arrives on the other side the book ends up with balanced action. The margin is the vig, almost always around 4-5% on a -110/-110 spread.
The opening number — what bettors call the 'overnight' or 'sharp open' — is rarely the model's pure output. It is the model adjusted for the book's risk tolerance, then rounded to a half-point or whole number, then shaded for known public bias. Sharps test the opening line within minutes of release. If the sharps push the spread half a point, that move becomes the market's first piece of public information.
Reading the price next to the number
The spread number and the price next to it carry separate information. A line listed as Chiefs -7 (-110) / Bills +7 (-110) says the book believes a 7-point Chiefs win is the median outcome. A line listed as Chiefs -7 (-120) / Bills +7 (+100) says the book has had enough money on the underdog and is now charging an extra 10 cents to bet the favorite, while paying full value on the dog. The number tells you the median. The price tells you the imbalance.
When you compare two books on the same game, the spread move from -7 to -7.5 is roughly equivalent to a price move from -110 to -125. Either change costs the bettor about 3-4% of win probability. Sharps use this equivalence to choose between books — a -7 at -120 may be a worse trade than -7.5 at -110, even though the headline number looks better.
Key numbers in NFL spreads
NFL margins of victory cluster around the values the scoring system produces most efficiently. A field goal is three points. A touchdown plus extra point is seven. A field goal plus a touchdown is ten. Two unanswered touchdowns is fourteen. These numbers — and a few others (4, 6, 11) — appear far more often as final margins than the numbers in between.
| Margin | NFL frequency | Spread sensitivity | Half-point value |
|---|---|---|---|
| 3 | ~14.7% | Highest in football | 3-4% win probability |
| 7 | ~9.1% | Very high | 2-3% win probability |
| 10 | ~5.9% | High | 1.5-2% win probability |
| 14 | ~4.6% | Moderate | 1-1.5% win probability |
| 4 | ~3.7% | Low | ~0.5% win probability |
| 6 | ~6.0% | Moderate (PAT misses) | 1-1.5% win probability |
| 1, 2, 5, 8, 9, 11+ | Each <3% | Low | <0.5% win probability |
The practical implication is that a spread move from -2.5 to -3 (crossing the 3-point key number) is almost twice as expensive in true EV as a move from -1.5 to -2 (not crossing a key). Books price this asymmetry into their buy/sell point menus, but rarely fully — a careful bettor can occasionally find half-points priced as if all numbers were equal when they are not.
NBA spreads versus NFL spreads
NBA spreads behave differently because basketball scoring is dense and continuous. There is no key number in basketball with anything close to the influence of 3 and 7 in football. Half-points still matter — moving from -6 to -6.5 is meaningful — but no single number dominates. NBA spreads also move more during the week because injury news arrives later (load management decisions are often game-day) and because the smaller per-possession value of basketball points makes the market more responsive to small information.
Reverse line movement and the sharp signal
The classic sharp indicator on a spread is reverse line movement: the spread moves opposite to the public ticket count. If 75% of NFL bets are on Team A but the spread moves toward Team B, the book is willingly accepting a lopsided liability because the sharp money on Team B is more informed than the casual money on Team A. Books that publish ticket counts (DraftKings, FanDuel, BetMGM) are a useful sanity check; books that do not (Pinnacle, Circa) tell the same story through their line itself.
Not every reverse line move is sharp action. Sometimes it is simply asymmetric stake size — one $50,000 bet outweighs fifty $200 bets. The signal-to-noise ratio is highest in the 12-48 hours before kickoff, when sharp money has done its work and casual money has accumulated.
Buying and selling points
Books offer point-buying menus where the bettor moves the spread in their favor for a worse price, or sells the spread against themselves for a better price. Buying a half-point off the 3 in the NFL (moving -3 to -2.5) typically costs 25 cents — the price goes from -110 to -135. The implied win-probability gain of crossing the 3 is roughly 3-4%, and 25 cents of price costs roughly 6-7% of EV. The bettor loses on the trade. The same half-point on a non-key number (moving -5 to -4.5) gains roughly 0.5% in win probability but still costs 5-10 cents — also a losing trade.
The math is consistent: buying points is a losing trade in expectation. The exception is when the bettor sees a key-number half-point priced at a non-key cost (10 cents instead of 25), which occasionally happens at smaller books late in the week. Selling points is the same math in reverse — accept a worse spread for a better price — and is occasionally useful for hedging existing positions, but rarely a positive-EV play on its own.
The regulator angle
In the United States, the spread is the most heavily regulated betting product because it dominates handle. State gaming commissions monitor spread integrity through unusual line-movement tracking, and books are required to report suspicious-betting alerts to state regulators and to leagues. The Tim Donaghy case (NBA referee 2007), the Calvin Ridley suspension (2022), and the Jontay Porter ban (2024) all originated from spread-market or prop-market integrity alerts. The mechanism is the same: a single bettor or syndicate moves a line in a direction that the public action does not justify, the book flags the trade, and the integrity team investigates.
The recreational bettor's biggest spread mistakes
Three behaviours separate casual spread bettors from break-even-or-better bettors. The first is line shopping. A bettor who takes the best available spread across five books gains roughly 1.5-2% in EV on every bet, larger than most modeling edges. The second is respecting key numbers — never blindly betting -3 when -2.5 is available at the same price, and never paying 25 cents to move a non-key number. The third is parlay discipline — spread parlays compound vig multiplicatively, and a four-leg spread parlay carries an effective margin of 17-22%, more than triple a single-game spread.
The professional checklist
- Compare the spread to the no-vig consensus across at least three books. A spread two cents off Pinnacle's number is a market signal.
- Identify the key numbers for the sport you are betting. NFL: 3, 7, 10, 14. NBA: no dominant key. CFB: similar to NFL, weaker.
- Track reverse line movement against public ticket counts. Sharp money is the dog when the spread moves against the public count.
- Avoid buying points except across a key number at a non-key price.
- Never confuse the headline number with the price. -7 (-120) and -7.5 (-105) carry similar EV; the headline can mislead.
- Document CLV against the closing spread. Beating the close by 0.5 points or 5 cents is the only durable evidence you are betting better than the market.
Sources & further reading
- Levitt, Steven D. "Why are gambling markets organised so differently from financial markets?" Economic Journal, 2004 — bookmaker price-setting behaviour vs. exchange markets.
- Paul, Rodney J. & Weinbach, Andrew P. "Investigating the inefficiency of the NFL betting market." Journal of Economics & Finance, 2008 — key-number analysis and recreational bettor bias.
- Pro-Football-Reference — NFL historical margin-of-victory distributions, 1980-2024.
- Nevada Gaming Control Board — Sports Pool Monthly Reports, 2010-2025 (handle and hold by sport).
- Roxborough, Michael "Roxy" — interviews and published material on the history of spread-setting in Las Vegas, 1990-2010.