luckadjusted
CFB Week 4 · 2026

Week 3 2026 · September 19, 2026

Purdue at UCLA, Week 3 2026: who deserved to win?

UCLA beat Purdue 52–38, and the replays lean its way: with the same day's form, it wins 8 of 10.

Before kickoff UCLA was an 81% favorite.

Final

Purdue Boilermakers 38
UCLA Bruins 52

September 19, 2026 · Rose Bowl · Closing line: UCLA −13.5

Luck index +21 UCLA about even · −21 Purdue

Each team’s own plays from this game, resampled into new drives 4,000 times: with the way both teams actually played, how often does this performance win? How the replay works

Ten replays. Purdue wins 2, UCLA wins 8.

  1. 1 PUR 52 UCLA 35
  2. 2 PUR 38 UCLA 35
  3. 3 PUR 20 UCLA 23
  4. 4 PUR 42 UCLA 49
  5. 5 PUR 49 UCLA 59
  6. 6 PUR 38 UCLA 51
  7. 7 PUR 35 UCLA 52
  8. 8 PUR 35 UCLA 56
  9. 9 PUR 24 UCLA 52
  10. 10 PUR 28 UCLA 66

Three other ways to ask

Each answers a different question, so none of them replaces the replay above.

The closing betting line, turned into a probability. What everyone knew before the game.

The line plus what this game showed, weighted by how much one game can tell. It stays close to the line because one game says little.

Every kick, fumble bounce, conversion and judgment call redrawn; everything else fixed. The narrowest view of luck.

How 4,000 replays ended

Purdue wins 21% of them, UCLA 79%, counting the 2% still level after regulation as half each. The line is what actually happened: UCLA by 14.

Show as a table
Final margin in the replaysShare of replays
Purdue by 17+3.9%
Purdue by 9–166.3%
Purdue by 1–89.8%
Level after regulation2.2%
UCLA by 1–820.6%
UCLA by 9–1617.5%
UCLA by 17+39.8%

Where the luck was

What the luck was worth

Every lucky and unlucky play, weighted by the moment it happened: its swing in win probability, added up by kind. A missed kick in a decided game counts for nothing; the same kick at the end of a tie game can be most of a win.

Swings add up across a game, so in one that went back and forth they can pass 100%. On the scoreboard the same luck was worth +4.7 points to UCLA: the final margin was UCLA by 14, without the luck UCLA by 9.3. Luck shares are the fitted ones (how much of each kind is luck).

The kicks

How likely each kicker was to make that kick (his own record, the distance), and what happened.

  • M. Bhaghani, 43 yds missed · 70% expected
  • M. Bhaghani, 34 yds made · 84% expected
  • J. Echeverria Lozano, 27 yds made · 89% expected
Bar: expected chance of a make. Filled dot: made. Open dot: missed or blocked.

Penalties: Purdue 5 for 41 yards, UCLA 12 for 103 yards

Every accepted flag, in order, and what it did to the helped team’s chance of winning at that moment, including any play it wiped out (points on the scoreboard in grey). A flag in a decided game is worth nothing; one at the turn of a close game can be worth most of a win. Judgment calls count toward luck, pre-snap fouls mostly don’t. How calls are valued

WhenFlagWin probability for
Q1 14:49 Face Mask on Purdue, 15 yards +2% UCLA +1.6 pts
Q1 11:45 Holding on UCLA, 10 yards 0% UCLA +0.2 pts
Q1 6:29 Illegal Motion on UCLA, 5 yards, wiping out the play · pre-snap +1% PUR +1.3 pts
Q1 5:39 Roughing the Passer on Purdue, 6 yards +1% UCLA +1.1 pts
Q2 12:32 Pass Interference on Purdue, 15 yards +1% UCLA +0.9 pts
Q2 6:06 Pass Interference on UCLA, 15 yards +1% PUR +0.5 pts
Q2 3:42 Intentional Grounding on Purdue, 0 yards 0% UCLA 0.0 pts
Q2 2:05 Holding on UCLA, 8 yards +1% PUR +0.5 pts
Q3 10:35 False Start on UCLA, 5 yards · pre-snap +1% PUR +0.6 pts
Q3 8:16 Roughing the Passer on UCLA, 15 yards +3% PUR +2.6 pts
Q3 8:12 Offside on UCLA, 5 yards · pre-snap +1% PUR +0.7 pts
Q3 4:43 Illegal Formation on Purdue, 5 yards, wiping out the play · pre-snap +7% UCLA +3.5 pts
Q4 14:56 Offside on UCLA, 5 yards · pre-snap 0% PUR +0.3 pts
Q4 13:06 Roughing the Passer on UCLA, 10 yards +1% PUR +0.8 pts
Q4 10:20 Delay of Game on UCLA, 5 yards · pre-snap +1% PUR +0.4 pts
Q4 10:06 Illegal Formation on UCLA, 5 yards · pre-snap +1% PUR +0.3 pts
Q4 7:13 Face Mask on UCLA, 15 yards 0% PUR +1.6 pts

When it happened

UCLA’s chance of winning

Numbers mark the biggest luck moments, listed below.

Our in-game win probability: the score, the value of the possession and the time left, starting from the betting line. Overtime has no clock and is drawn at a fixed width. Shaded above the middle while UCLA was ahead in it, below while Purdue was.
#WhenWhat happenedWin probability forNext drive
1 Q2 4:34 Purdue converted 3rd & 83rd & 4th downs · 36% expected +3% PUR PUR: Touchdown
2 Q2 2:50 Purdue converted 4th & 23rd & 4th downs · 66% expected +3% PUR PUR: Touchdown
3 Q2 0:33 UCLA converted 3rd & 153rd & 4th downs · 23% expected +5% UCLA UCLA: Field goal
4 Q3 13:31 Purdue converted 3rd & 83rd & 4th downs · 36% expected +6% PUR PUR: Touchdown
5 Q3 10:16 Purdue fumbled and UCLA recoveredFumble bounces · loose balls are close to 50/50 +4% UCLA PUR: Field goal
6 Q3 8:16 Roughing the Passer on UCLA, 15 yardsJudgment calls +3% PUR PUR: Field goal
7 Q3 5:50 Purdue converted 4th & 83rd & 4th downs · 35% expected +5% PUR PUR: Field goal
8 Q3 4:43 Illegal Formation on Purdue, 5 yards, wiping out the playPre-snap flags +7% UCLA UCLA: Touchdown

Adjust the assumptions

Disagree with how much of a fumble bounce or a pass interference call is luck? Change it and the coin-flip replay reruns in your browser.

Only coin-flips re-rolled

Purdue 10% · UCLA 90%

Final margin
UCLA by 14
Luck in points
4.7 for UCLA
Luck-adjusted margin
UCLA by 9.3
Replay with same day's form
Purdue 21% · UCLA 79% unchanged: why
How much of each is luck?

0% treats it as skill, 100% as pure luck. The tick is our fit.

What counts

None by default: our tests measured no carryover after a lucky break. Slide it up to count a break for more than its own swing. Why

Flip a play

Force the other outcome, in this game and in every replay: a made kick becomes a miss, a stop becomes a conversion, a flag goes the other way.

About these numbers

Play-by-play and expected points from CollegeFootballData.com via cfbfastR; win probability from our own model of the score, the possession and the betting line; kicks valued against each kicker’s own record. College play-by-play has no charting of drops or interception-worthy throws and no injury reports, so those kinds of luck aren’t counted here.

Methodology and validation · All Week 3 2026 games · Purdue · UCLA

More from Week 3 2026

Every week 3 game, 2026