Eastern Michigan’s third-down conversion rate with Nick Devereaux and without Nick Devereaux
, ,
Against what the season’s ratings expected, Eastern Michigan was 1.2 percentage points under with Nick Devereaux and 13.1 percentage points over without Nick Devereaux over its last 25 games, 18 games against 4. The plain averages are 38.5% and 51.7%, a difference of 13.2%. All 5 night games were with Nick Devereaux. All 4 games after a bye were with Nick Devereaux.
Nick Devereaux, receiver: 688 receiving yards, 54 receptions and 9 receiving touchdowns in 20 of those 25 games.
third-down conversion rate a game over expected
A −14.3% difference in third-down conversion rate over expected with Nick Devereaux vs without Nick Devereaux is a bit unusual: shuffle which games were with Nick Devereaux and it comes up 13 times in 100.
A −13.2% difference in third-down conversion rate with Nick Devereaux vs without Nick Devereaux is unusual: shuffle which games were with Nick Devereaux and it comes up 2 times in 100.
The games
| Side | Date | Game | Result | 3rd downs | Expected |
|---|---|---|---|---|---|
| Nov 2, 2024 | vs Toledo | L | 45.0% | 39.3% | |
| Nov 13, 2024 | @ Ohio | L | 41.2% | 39.2% | |
| Nov 20, 2024 | vs Buffalo | L | 35.7% | 39.2% | |
| Nov 30, 2024 | @ Western Michigan | L | 66.7% | 40.0% | |
| Aug 30, 2025 | @ Texas State | L | 40.0% | 40.5% | |
| Sep 6, 2025 | vsLLong Island University | L | 41.7% | — | |
| Sep 13, 2025 | @ Kentucky | L | 28.6% | 37.7% | |
| Sep 20, 2025 | vs Louisiana | W | 42.9% | 41.5% | |
| Sep 27, 2025 | @ Central Michigan | L | 50.0% | 42.5% | |
| Oct 4, 2025 | @Buffalo | L | 40.0% | 37.1% | |
| Oct 11, 2025 | vs Northern Illinois | W | 37.5% | 46.0% | |
| Oct 18, 2025 | @ Miami (OH) | L | 46.2% | 43.6% | |
| Oct 25, 2025 | vsOhio | L | 41.7% | 36.3% | |
| Nov 8, 2025 | vs Bowling Green | W | 47.1% | 38.0% | |
| Nov 15, 2025 | @ Ball State | W | 50.0% | 40.2% | |
| Nov 25, 2025 | vsWestern Michigan | L | 25.0% | 37.3% | |
| Aug 29, 2026 | vsSSacramento State | W | 33.3% | — | |
| Sep 4, 2026 | vs San Jose State | L | 18.8% | 40.7% | |
| Sep 12, 2026 | @ Michigan State | L | 46.2% | 33.3% | |
| Sep 19, 2026 | @ Wisconsin | L | 25.0% | 39.5% | |
| Sep 26, 2026 | vsLLindenwood | W | 57.1% | — | |
| Oct 3, 2026 | @ UMass | W | 78.6% | 32.6% | |
| Oct 10, 2026 | @ Akron | W | 44.4% | 37.1% | |
| Oct 17, 2026 | vsToledo | L | 33.3% | 39.8% | |
| Oct 24, 2026 | @Ohio | L | 13.3% | 44.2% |
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 0.6% in 2024, 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 50% in 2024, 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.
- Long Island University, Sacramento State, Lindenwood 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 for 21 of these 25 games; in the other 4 a player is counted as in them when the box score has a stat for him.