Why 3 and 7 Decide Every NFL Spread

Of 6,967 completed NFL regular-season games in NFL game results from 1999 through 2025, 1,046 ended with a three-point margin and 634 ended with a seven-point margin. Together, those outcomes account for 1,680 games, or 24.1 percent of the window.

That is the spine of the key-number argument. The market does not treat every half point alike because games do not finish on every margin equally. A move across three or seven touches a larger chunk of the historical result set than a move across a quiet number. The hook is not decoration. It changes the set of final scores that cash, lose, or push a ticket.

Three is the value of a field goal. Seven is the common value of a touchdown followed by the extra point. When a close game is separated by one ordinary scoring event, those are natural landing spots. The historical rows show the result: three is the most frequent margin in the window, and seven is second.

This does not make either number a side by itself. A game can land on three after a chaotic sequence that no pregame model anticipated. The point is narrower: when two otherwise identical prices sit on opposite sides of a key number, they do not offer identical protection. The bettor who takes the worse half point is giving away outcomes that appear often in the completed schedule.

The three-point margin appears in 15.0 percent of the 6,967-game window. The seven-point margin appears in 9.1 percent. Those rates describe final margins, not a strategy record, and they do not say that buying every hook is profitable.

In the captured August 30 Week 1 schedule, twelve of sixteen openers sit within a half point of three or seven. Four matchups are listed exactly on one of those numbers: Atlanta at Pittsburgh, the Jets at Tennessee, and Denver at Kansas City at three; New Orleans at Detroit at seven. Arizona at the Chargers is the only opener more than three points from both key numbers.

The current board does not prove why each bookmaker chose each number. It does show how tightly the market is arranged around the historical landing spots. That is useful before any handicap enters the conversation. It tells you where line shopping can change the ticket's outcome set and where a move is mostly cosmetic.

A spread of -2.5 and a spread of -3 may look adjacent. They are separated by every game that lands on exactly three. The same applies to +3 and +3.5, or the comparable pairs around seven. A bettor should record the actual number available at the actual book, not summarize all of them as “around a field goal.”

New Orleans at Detroit is captured at Detroit -7 at both Bovada and Pinnacle. The books hold the whole number in the same snapshot. Denver at Kansas City and the Jets at Tennessee show the other behavior: the schedule feed lists a clean three while Bovada's live spread is -2.5.

Those rows are evidence of price differences, not evidence of a bookmaker's private motive. A book may be managing liability, following a market maker, reacting to information, or simply sitting at a different point in the update cycle. The data here cannot choose among those stories. It can tell you that the half point existed and that the ticket terms were different.

The sharp question is not “which number sounds right?” It is “which outcomes does this price add or remove?”

Cleveland at Jacksonville appears at Jacksonville -7.5 in the August 22 Bovada rows and at -8.5 by August 25. The feed then holds -8.5 through August 30. Atlanta at Pittsburgh remains Pittsburgh -3 in the available Bovada window from August 22 through August 30.

One market moved away from the seven neighborhood. The other stayed pinned to three. That contrast is the useful observation. The snapshots do not reveal the order flow, injury information, or internal decision that produced either path, so the article does not supply one.

Line history is strongest when it is treated like a ledger. Save the book, market, number, and capture time. Then compare actual rows. A narrative built after kickoff can always make the move sound obvious. A timestamped series can show whether the move was there before the result.

Start with the handicap, then price it. If your case for a side survives only at +3.5 and the board is +3, the missing hook is the decision. If your case survives at either number, the hook is still worth shopping but it does not have to dominate the thesis. The key number tells you where the ticket changes; it does not tell you which team is better.

Do not backfill a record from cherry-picked examples. A valid performance claim would need the side, closing spread, grading rule, full window, and sample. This article makes no such claim. It reports historical final margins and current pregame prices, each with its source class named.

For the market mechanics, read How NFL Spreads Work. For judging the price you actually took, use Closing Line Value Explained. For the cost embedded in two-sided prices, continue with the vig and hold guide.

Watch the games sitting one half point from three or seven and record which book moves first. The next useful fact is not a prediction about why the market will move. It is the next timestamped row. Around the NFL's most common margins, that one row can change the ticket more than the size of the move suggests.

NFL ATS cover-margin distribution

Bars count completed NFL schedule rows by closing-spread cover margin using the repository canonical home-margin grading convention.