Interesting article posted today by Bill Connelly, ESPN staff writer. The analysis tries to quantify the amount of “luck” that went into turnovers by college defenses. While causing fumbles is a skill, who actually recovers the ball is more luck than skill. He cites the example of USC which had a remarkable amount of “turnover luck”. In the first five games of the season, there were 9 fumbles by opposing offenses. USC recovered all nine. If you stretch things out over time, fumble recoveries should be 50-50. Things regress to the mean. Only 8 defenses recovered less than 35% of fumbles in 2022. UH was one of those with a recovery rate of 34.2% (124th in the nation). UH was -6 on expected recoveries. If a turnover is worth 4 to 5 points per analytics, then UH had bad fumble recovery luck that cost us 24-30 points on the season. A couple of key fumble recoveries could have turned an L into a W in several close losses.
Interceptions take a deeper analysis. This study analyzes interception to breakup rate. Passes “defensed” is defined as interceptions and PBUs. Analytics say that 21% of passes defensed are intercepted. I take that to mean that about 1 out of every 5 PBU turns out to be an interception. Some teams had very good “turnover luck” and intercepted almost 33% of passes defended.
In the overall analysis of turnover luck, the article had UH next to last. Number 130 out of 131. Only Rutgers was below UH in turnover luck. Putting it in terms of points per game, bad turnover luck cost UH 3.7 points per game.
So if last year’s team had just average luck regarding turnovers by the defense, how much would our record have improved? 3.7 points per game would have turned 2 overtime losses into wins.
Interesting article.