Week 4 2025 · September 20, 2025
Ball State at UConn, Week 4 2025: who deserved to win?
UConn beat Ball State 31–25, but the replays lean toward Ball State: it takes 6 of 10 with the same day's form.
UConn won on an off day: it played far below what the matchup promised. In a fresh game against Ball State it would win 89% of the time; with this day’s form, 37%. The numbers
Before kickoff UConn was a 91% favorite. The biggest swing (Q4, 0:09 left): Ball State’s two-point try was good, worth 17 points of win probability to Ball State.
Final
 Ball State Cardinals | 25 |
|---|---|
 UConn Huskies | 31 |
Replay with same day's form
Luck index +63 UConn very lucky · −63 Ball State
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. Ball State wins 6, UConn wins 4.
- 1 BALL 42 CONN 14
- 2 BALL 31 CONN 13
- 3 BALL 17 CONN 3
- 4 BALL 21 CONN 10
- 5 BALL 24 CONN 17
- 6 BALL 24 CONN 21
- 7 BALL 14 CONN 16
- 8 BALL 12 CONN 17
- 9 BALL 20 CONN 30
- 10 BALL 0 CONN 20
Three other ways to ask
Each answers a different question, so none of them replaces the replay above.
Before kickoff
The closing betting line, turned into a probability. What everyone knew before the game.
Same teams, fresh 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.
Only coin-flips re-rolled
Every kick, fumble bounce, conversion and judgment call redrawn; everything else fixed. The narrowest view of luck.
UConn won on an off day
It played far below what the matchup promised. In a fresh game against Ball State it would win 89% of the time; with this day’s form, 37%.
Same teams, fresh game
Replay with same day's form
52 points apart for UConn: the gap is how differently this game went from what the two teams usually are.
| Unit | This game | Usual level | Better or worse than usual |
|---|---|---|---|
| Ball State offenseEPA per play gained | −0.06 | +0.13 | worse by 0.19 |
| Ball State defenseEPA per play allowed | +0.27 | +0.12 | worse by 0.16 |
| UConn offenseEPA per play gained | +0.27 | +0.09 | better by 0.18 |
| UConn defenseEPA per play allowed | −0.06 | +0.02 | better by 0.08 |
EPA (expected points added) per play on competitive snaps, turnovers left out because they are valued as luck. Usual level: the team’s earlier games this season plus half of last season, pulled toward the league average. One team’s defense and the other team’s offense are the same snaps, so one game can’t fully say whose day it was; it shows how far this one was from both teams’ norms.
How 4,000 replays ended
Ball State wins 63% of them, UConn 37%, counting the 7% still level after regulation as half each. The line is what actually happened: UConn by 6.
Show as a table
| Final margin in the replays | Share of replays |
|---|---|
| Ball State by 17+ | 22.5% |
| Ball State by 9–16 | 18.2% |
| Ball State by 1–8 | 20.6% |
| Level after regulation | 6.9% |
| UConn by 1–8 | 16.9% |
| UConn by 9–16 | 8.8% |
| UConn by 17+ | 6.1% |
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.
The kicks
How likely each kicker was to make that kick (his own record, the distance), and what happened.
- C. Freeman, 30 yds made · 87% expected
- C. Holmer, 23 yds made · 91% expected
Penalties: Ball State 4 for 32 yards, UConn 1 for 10 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
| When | Flag | Win probability for |
|---|---|---|
| Q1 9:48 | False Start on Ball State, 5 yards · pre-snap | 0% BALL +1.3 pts |
| Q3 11:49 | Roughing the Passer on Ball State, 2 yards | +1% CONN +1.8 pts |
| Q3 3:14 | Penalty on Ball State, 10 yards | +1% CONN +0.6 pts |
| Q3 0:33 | Holding on UConn, 10 yards | +1% BALL +1.6 pts |
| Q4 0:05 | Penalty on Ball State, 15 yards | +2% BALL +0.9 pts |
When it happened
UConn’s chance of winning
Numbers mark the biggest luck moments, listed below.
| # | When | What happened | Win probability for | Next drive |
|---|---|---|---|---|
| 1 | Q2 6:46 | UConn converted 3rd & 23rd & 4th downs · 63% expectedUConn was stopped on 3rd & 33rd & 4th downs · 54% expected | +2% BALL | CONN: Punt |
| 2 | Q2 2:19 | Ball State converted 3rd & 43rd & 4th downs · 47% expectedBall State converted 3rd & 43rd & 4th downs · 47% expected | +3% BALL | BALL: End of half |
| 3 | Q3 11:49 | UConn converted 3rd & 23rd & 4th downs · 63% expectedUConn was stopped on 3rd & 83rd & 4th downs · 31% expectedUConn converted 4th & 43rd & 4th downs · 46% expectedUConn was stopped on 3rd & 63rd & 4th downs · 38% expectedUConn converted 4th & 63rd & 4th downs · 37% expectedRoughing the Passer on Ball State, 2 yardsJudgment calls | +7% CONN | CONN: Touchdown |
| 4 | Q3 3:14 | Ball State was stopped on 3rd & 13rd & 4th downs · 70% expectedPenalty on Ball State, 10 yardsJudgment calls | +3% CONN | BALL: Field goal |
| 5 | Q3 0:37 | C. Holmer made a 23-yard field goalKicks · 91% expected for him | +2% CONN | BALL: Turnover |
| 6 | Q4 15:00 | UConn converted 3rd & 143rd & 4th downs · 22% expected | +3% CONN | CONN: Touchdown |
| 7 | Q4 0:09 | Ball State’s two-point try was good3rd & 4th downs · 46% expected | +17% BALL | – |
| 8 | Q4 0:05 | Penalty on Ball State, 15 yardsJudgment calls | +2% BALL | CONN: End of game |
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
Ball State 59% · UConn 41%
- Final margin
- UConn by 6
- Luck in points
- 7.4 for UConn
- Luck-adjusted margin
- Ball State by 1.4
- Replay with same day's form
- Ball State 63% · UConn 37% unchanged: why
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 4 2025 games · 
Ball State · 
UConn
More from Week 4 2025
- Troy at Buffalo TROY 21–17 · replay 30% for the winner
- Louisiana at Eastern Michigan EMU 34–31 · replay 42% for the winner
- UNLV at Miami (OH) UNLV 41–38 · replay 50% for the winner
- Toledo at Western Michigan WMU 14–13 · replay 71% for the winner
- Northern Illinois at Mississippi State MSST 38–10 · replay 88% for the winner
- Gardner-Webb at Ohio OHIO 52–35 · replay 95% for the winner
- Bowling Green at Louisville LOU 40–17 · replay 99% for the winner
- Duquesne at Akron AKR 51–7 · replay 100% for the winner
- Wagner at Central Michigan CMU 49–10 · replay 100% for the winner
- Purdue at Notre Dame ND 56–30 · replay 100% for the winner
- Kent State at Florida State FSU 66–10 · replay 100% for the winner
- Fresno State at Hawai'i FRES 23–21 · replay 24% for the winner