About 15% of NFL games end with a margin of exactly 3 points. Another ~9% land on 7. Another ~5% on 10. These aren't random — they're the structural artifacts of how points score in football (field goals worth 3, touchdowns worth 7), and they create the most-studied edge in football betting: key numbers. This guide covers which numbers matter, the historical push frequencies on each, the math of buying or selling a half-point through them, and the framework sharp bettors use to decide when the move is worth the price.
The "key number" concept in NFL betting is one of the cleanest, most replicable edges in sports betting math. Football's scoring structure produces a distinctive distribution of game margins: scores cluster at 3-point intervals (field goals) and 7-point intervals (touchdowns + extra point), with massive density at exactly 3 and 7. That density doesn't appear in basketball, hockey, baseball, or soccer the same way — it's specifically a football phenomenon, and it's the basis for half-point shopping at retail sportsbooks.
Across 2000–2024 NFL regular-season and playoff games (roughly 6,400 games), the share of games ending on each margin breaks down approximately as follows:
| Margin | Share of games | Why |
|---|---|---|
| 3 | 14.8% | One field goal — the most common winning margin |
| 7 | 9.2% | One touchdown + PAT |
| 10 | 5.1% | One TD + one FG |
| 6 | 4.4% | Two FGs or TD with missed PAT/no two-point try |
| 4 | 4.3% | Two FGs minus one (one TD with missed PAT possible) |
| 14 | 2.6% | Two TDs + two PATs |
| 1 | 2.5% | Safety or extra-point miss in tight games |
| 2 | 3.7% | Common with two-point tries / safeties |
| 5 | 2.1% | Mid-margin, less common |
| 8 | 2.2% | TD + successful two-pointer |
| 11 | 1.4% | Mid-margin |
| 13, 17, 20, 21 | 1−3% each | Other TD-derived margins |
The hierarchy is striking. 3 appears on roughly 15% of all games. 7 appears on 9%. Together, those two margins account for nearly a quarter of all NFL games — vastly more than any non-key-number margin. 10 and 14 follow as secondary key numbers; 4 and 6 are tertiary. All other margins land in the 1–3% range.
When you place a spread bet, your bet wins or loses depending on the final margin. If the spread is −3 and the favorite wins by exactly 3, the bet pushes (refund). If the favorite wins by 4+, the bet wins. If the favorite wins by 2 or less, the bet loses.
The push-rate table tells you how often the bet you're considering will land exactly on the spread. A spread of −3 has about a 15% chance of pushing. A spread of −7 has about a 9% chance of pushing. A spread of −4.5 (a half-point off the 4 key number) has about a 4% chance of not pushing in the bettor's favor relative to a spread of −4.
This is what makes half-point shopping interesting. Moving the spread by half a point doesn't just shift the breakeven margin by a small amount — when the move crosses a key number, it shifts the bettor's cover probability by the entire share of games that land on that number. Going from −3 to −2.5 picks up the ~15% of games ending in a 3-point favorite win (converting them from push to bettor win). Going from −5 to −4.5 picks up only the ~2% of games ending in a 5-point margin.
Books offer "alternate spreads" with their own pricing. Typical structure:
Current line: −3 at −110
Alternate (buy): −2.5 at typically −130 to −145
What you gain: Cover-rate boost equal to the push frequency on 3 (~15%) since 3-point favorite wins now count as a win instead of a push.
What you pay: The implied-probability cost of moving from −110 to −130 is roughly 4.9 percentage points (52.4% → 56.5%). Or for −145, it's ~7.5 points.
Is it worth it? If the push-rate gain exceeds the implied-probability cost, yes. 15% gain vs 4.9% cost = +10 percentage points of cover probability — one of the clearest +EV plays in NFL betting.
The decision rule for any half-point buy: cover-rate gain > implied-probability cost. When you cross a key number, the cover-rate gain is large (5–15%); when you don't, it's small (1–3%). Books price half-point buys based on which side of the key-number landscape the move is on, but they don't always price perfectly — especially on derivative or less-popular markets.
NFL key-number decisions reduce to seven common scenarios:
A teaser combines multiple half-point or 6-point shifts into a single bet at a fixed price. The most famous +EV teaser format is the "Wong teaser" — a 2-leg 6-point NFL teaser where both legs cross both 3 and 7 in the bettor's favor.
Setup: take a favorite of −7.5 to −8.5 and tease it down by 6 points (becomes −1.5 to −2.5). Take an underdog of +1.5 to +2.5 and tease it up by 6 points (becomes +7.5 to +8.5). Both legs cross 3 and 7 in the bettor's favor. Combined, the cover-rate gain from those four key-number crossings is large enough to overcome the typical 6-point teaser price (−110 to −120 for a 2-team teaser).
Historically, Wong teasers have been profitable enough that sharp bettors used them as a foundational NFL strategy. Books have responded over the years by raising teaser pricing and limiting eligible games. Wong teasers still work but with thinner margins than 10 years ago — the edge persists but the size is smaller.
Three rules sharps follow:
NFL key numbers are point-margin values where games end disproportionately often, creating push and near-push situations on spread bets. The biggest are 3 (about 15% of games), 7 (~9%), 10 (~5%), and 14 (~2.5%). Lesser key numbers include 4 (~4%) and 6 (~4%). These frequencies are driven by football's scoring structure (touchdowns of 7, field goals of 3) and have remained remarkably stable across decades of NFL data.
Because field goals are worth 3 points and most close NFL games are decided by one or two field goals. About 15% of all NFL games end with a margin of exactly 3 points — by far the highest single-number frequency. A spread sitting on 3 will push or near-push at meaningfully higher rates than any other line, which makes buying or selling off 3 the most studied half-point decision in football betting.
Sometimes. When the half-point move crosses a key number (3, 7, 10, 14), the cover-rate gain is real and often justifies the price the book charges. When the half-point move doesn't cross a key number (going from -2.5 to -2, or -5 to -4.5), the cover-rate gain is small and the book's price usually overcharges. Sharp bettors only buy half points when crossing a key number, and only when the offered price is fair relative to the historical push rate.
Standard pricing on a half-point buy off 3 typically moves the line from -110 to around -125 or -130 (a 15-20 cent premium). Buying off 7 typically costs 10-15 cents. Buying off non-key numbers usually costs 10 cents — and is usually -EV because the cover-rate gain doesn't justify the price. The Half Point Calculator computes the implied probability cost and flags whether the move is +EV or -EV.
Not automatically — it depends on the offered price. Buying -3 down to -2.5 captures roughly 9-10 percentage points of additional cover probability (the share of games ending on exactly 3 that flip from push to win). If the book charges more than that in implied probability, the buy is -EV. Sharp bettors check the math each time rather than assuming key-number buys are always good.
Buying a half point is changing one bet's number for a worse price (-110 to -125). Teasing combines multiple bets, each with a 6 or 6.5 point shift in your favor, into a parlay at a single set price (typically -110 or -120). Teasers are essentially round-robin half-point buys at scale. The math is similar but compounds across legs; 'Wong teasers' (NFL 6-point teasers built specifically around crossing 3 and 7) are the most-studied +EV teaser format.