Purdue
UCLA
The books 1 account books · tap any price — the ticket builds at that book
| Book | PUR spread | UCLA spread | Over | Under | PUR win | UCLA win | PUR TT over | PUR TT under | UCLA TT over | UCLA TT under |
|---|---|---|---|---|---|---|---|---|---|---|
| FanDuel | — | — | — | — |
The gold price is the best available on that side; a gold ★ marks the side our number implies — a read, not a pick. TT is a team total — that team’s own points, not the game’s; columns are empty where no book hangs one. Send to book opens whichever book you tapped, not the one we would have picked. Prices are the books’ own.
Ours vs the market
Both percentages are UCLA’s chance to win — ours, then the book’s with the vig removed, medianed across the books we track.
What the gap has been worth 10,740 games · 2016–2025
How often the side we favoured covered, by how far our number sat from the market’s. Break-even at −110 is 52.4% — no bucket reaches it, and the rate falls as we disagree more. The further we sit from the market, the worse we have done.
Ranked by team-unit strength (rush & pass offense vs. run & pass defense) across all 265 D1 teams, not by individual players. Identity, tempo and unit strength are measured from the 2025 season.
The evidence 2025 production and power ratings · supporting the number above
Who carried each team in 2025
Top-rated players from last season, ranked within their position across D1. A grade measures what a player produced, and no 2026 game has been played yet — so this is who produced for these programs in 2025, not a claim about who is on the field this year.
Power breakdown
| Team | Power | Off | Def | Rank |
|---|---|---|---|---|
| Purdue Boilermakers |
+9.7 | +7.1 | +2.6 | #87 |
| UCLA Bruins |
+19.9 | +9.4 | +10.5 | #39 |
Def is higher = better. The projection assumes a 2.4-pt home edge (0 at neutral sites).
Where the number comes from
An efficiency adjustment from season-to-date EPA and success rate applies once teams have played enough games; before then it is 0 and the number is the power rating alone.
Situational spots are shown for context, not baked in: travel, altitude and rest were tested against 6,000+ games and the power and efficiency ratings already account for them. Adding them again would double-count.