Maryland’s third-down conversion rate with DeJuan Williams and without DeJuan Williams
, ,
Against what the season’s ratings expected, Maryland was 2.6 percentage points under with DeJuan Williams and 1.0 percentage points under without DeJuan Williams over its last 25 games, 16 games against 2. The plain averages are 36.3% and 32.5%, a difference of 3.8%. Expected was 4.1 percentage points higher with DeJuan Williams than without DeJuan Williams. All 7 away games were with DeJuan Williams. All 3 ranked opponents were with DeJuan Williams. Both top-10 opponents were with DeJuan Williams.
DeJuan Williams, running back: 839 rushing yards, 185 carries and 55 receptions in 18 of those 25 games.
2 games without DeJuan Williams is still too small a sample for the statistics to say much, but maybe you can find something if you squint.
third-down conversion rate a game over expected
A −1.5% difference in third-down conversion rate over expected with DeJuan Williams vs without DeJuan Williams is about what you’d expect.
A +3.8% difference in third-down conversion rate with DeJuan Williams vs without DeJuan Williams is about what you’d expect.
The games
| Side | Date | Game | Result | 3rd downs | Expected |
|---|---|---|---|---|---|
| Oct 26, 2024 | @ Minnesota | L | 42.9% | 37.2% | |
| Nov 9, 2024 | @ Oregon | L | 31.6% | 34.9% | |
| Nov 16, 2024 | vs Rutgers | L | 47.1% | 41.1% | |
| Nov 23, 2024 | vs Iowa | L | 30.8% | 38.7% | |
| Nov 30, 2024 | @ Penn State | L | 16.7% | 38.5% | |
| Aug 30, 2025 | vs Florida Atlantic | W | 29.4% | 44.5% | |
| Sep 5, 2025 | vs Northern Illinois | W | 33.3% | 47.5% | |
| Sep 13, 2025 | vsTTowson | W | 35.7% | — | |
| Sep 20, 2025 | @ Wisconsin | W | 28.6% | 32.7% | |
| Oct 4, 2025 | vs Washington | L | 52.9% | 34.1% | |
| Oct 11, 2025 | vs Nebraska | L | 46.7% | 36.7% | |
| Oct 18, 2025 | @ UCLA | L | 35.3% | 42.8% | |
| Nov 1, 2025 | vs Indiana | L | 25.0% | 30.9% | |
| Nov 8, 2025 | @Rutgers | L | 33.3% | 36.2% | |
| Nov 15, 2025 | @ Illinois | L | 25.0% | 40.1% | |
| Nov 22, 2025 | vs Michigan | L | 35.7% | 35.3% | |
| Nov 29, 2025 | vs Michigan State (neutral) | L | 56.3% | 33.7% | |
| Sep 5, 2026 | vsHHampton | W | 57.1% | — | |
| Sep 12, 2026 | @ UConn | W | 53.3% | 35.6% | |
| Sep 19, 2026 | vs Virginia Tech | L | 58.3% | 33.3% | |
| Sep 26, 2026 | vsUCLA | L | 6.7% | 33.8% | |
| Oct 3, 2026 | @Nebraska | L | 7.1% | 32.7% | |
| Oct 10, 2026 | @ Ohio State | L | 25.0% | 36.3% | |
| Oct 17, 2026 | vsRutgers | L | 30.8% | 42.4% | |
| Oct 31, 2026 | vsIllinois | L | 42.9% | 39.9% |
Method
- Expected is an offense and a defense a team, fit over every game of its season between two teams in the table with the game being read left out, plus 0.7% in 2025, plus 1.4% in 2026 for the home team (the edge the fit finds once strength is in), nothing at a neutral site.
- Each expected number is then fit to the season’s games: a game bears out about 56% in 2025, 40% in 2026 of the difference the model predicts.
- The shuffle keeps the season’s shape: the games are cut in two where one date leaves the least variance, each game is read against its stretch, and the sides are shuffled only inside each stretch.
- Towson, Hampton have no other game in the table, so those games have no expected number and sit out the second figure.
- Who played is read from the schools’ box scores’ participation lists, which name every player who entered the game.
- 5 games in 2024 are neither with DeJuan Williams nor without DeJuan Williams: DeJuan Williams had no game for the team that season, and sit out.