Across 1,693 NFL games from 2020 to 2025, the consensus spread changed between our first and last pre-game snapshot in 85.4% of games, moving a full point or more in 63.4%. Our snapshots are daily, so these figures measure day-to-day change and understate total movement.
Current numbers are on the NFL odds board.
The headline
Movement is the norm, not the exception. But the size of the move matters more than the fact of it: 63.4% of games moved a full point or more, and 38.8% moved two or more.
Across three markets
| Market | Games | Snapshots | Span | Moved at all | Moved 1+ | Moved 2+ | Mean move |
|---|---|---|---|---|---|---|---|
| NFL spread | 1,693 | 18 | 11.0 d | 85.4% | 63.4% | 38.8% | 1.96 |
| NFL total | 1,693 | 13 | 10.0 d | 89.3% | 70.8% | 39.3% | 1.69 |
| MLB total | 10,142 | 2 | 0.7 d | 39.9% | 3.1% | 0.03% | 0.22 |
Consensus of at least three sportsbooks at each snapshot, first versus last pre-game observation, 2020–2025. Snapshots and span are medians per game. Games with only one pre-game snapshot are excluded.
Read the span column before comparing the rows. The NFL numbers cover a median of 18 observations across 11 days, because NFL games sit on the board for over a week. The MLB number covers a median of 2 observations across 0.7 days, because a single baseball game is typically priced a day or so out. So the MLB figure is not evidence that baseball totals are stickier — it is a much shorter observation window, and the comparison between leagues is not like-for-like.
12.7% of NFL games saw the consensus spread cross the key number 3 — the market moving from one side of the most common margin to the other. On the key-numbers page you can see why that particular crossing changes the bet more than any other.
What we can measure, and what we cannot
This is the part most published line-movement numbers leave out, so it goes here rather than in a footnote.
The historical archive was built by walking one snapshot per day at a fixed hour. So we can compare the market across days — and that is exactly what the figures above do — but we cannot see intraday movement, we cannot time a move to the hour, and we do not hold a kickoff-minute closing line. Anything requiring one, closing line value included, is not computable from this archive and is not published on this site.
Concretely, for the NFL spread series: the median game has 18 pre-game observations spanning 11 days, and the last one sits a median 3 hours before kickoff. So “85.4% moved” means: between the first day we saw the game priced and the last day before it started, the consensus number was different. Some part of the real movement happens after our final look and is invisible to us — which makes these figures a floor, not a full accounting.
Post-game rows are excluded everywhere, which also removes a known artifact in the archive where an import re-stamped its final slate under later dates.
Real wall-clock captures, going forward
Since July 2026 our live boards also write every refresh to the archive at its true timestamp, several times a day. That series is genuinely intraday — and it is still only days old, far too short to publish a movement statistic from.
| Sport | Rows | Distinct capture times | Window |
|---|---|---|---|
| americanfootball_nfl | 36,132 | 18 | 2026-07-23 to 2026-07-28 |
| baseball_mlb | 14,696 | 21 | 2026-07-23 to 2026-07-28 |
| mma_mixed_martial_arts | 4,906 | 19 | 2026-07-23 to 2026-07-28 |
Real wall-clock captures from the live board refresh. Stated here for transparency about what will become measurable later; no figure on this page uses it.
When that series is long enough to support an intraday measurement, this page will say so and the number will appear with its own window and sample size.
How this was measured
- For each game we take the consensus number at our first and last pre-game snapshot and compare them. Consensus = discrete median across at least three sportsbooks at that exact snapshot instant.
- Spread movement is measured on the home side only, so the sign is stable. Totals are measured on the over.
- Games with fewer than two qualifying pre-game snapshots are excluded: 1,693 of 1,693 NFL games with a spread consensus qualified.
- Historical snapshots sit on a daily grid at a fixed hour. These figures therefore measure day-to-day change and understate total movement; no intraday or closing-line claim is made.
- Produced by
repo_check/seo/f31_lake_stats.pyon a read-only database session. - Odds archive current to 2026-07-28. Figures refresh when the source scripts are re-run.
Next steps
Frequently asked questions
How often do betting lines move?
In our NFL sample the consensus spread changed between the first and last pre-game snapshot in 85.4% of 1,693 games, and moved a full point or more in 63.4%.
How much does an NFL spread typically move?
The mean absolute change was 1.96 points and the median 1.0, measured between our first and last pre-game observation.
Can you measure the closing line?
No. Our historical snapshots sit on a daily grid at a fixed hour, so we hold no kickoff-minute price. Closing line value is not computable from this archive and we do not publish it.
Do totals move as much as spreads?
Less often, in our sample: 89.3% of NFL totals and 39.9% of MLB totals changed, against 85.4% of NFL spreads.
Related: NFL odds board · NFL key numbers · how often the over hits · methodology
Every figure on this page is a historical measurement of what already happened, produced by a committed, re-runnable script over our own odds archive. Past frequencies are not forecasts and are not betting advice: Nebula Insights publishes no picks, no selections and no recommended wagers. 18+ only.