Iowa was unbelievably bad at recovering fumbles last year. Kirk Ferentz talked about ball security in his presser yesterday and it’s no wonder why. Iowa had 21 fumbles last year and lost 15 of them. That is terrible. To make matters worse, they only recovered 3 of their opponents' 11 fumbles. That comes in at just 9 recoveries out of the 32 times the ball was put on the turf.
Statisticians have exhaustively analyzed the stats on fumble recoveries and it turns out that both teams have a roughly 50/50 shot at coming up with the ball in most fumble situations. So recovering just 9 out of 32, when the expected value would be around 16 is... pretty bad. And pretty freaking unlucky.
How unlucky is it? I’m glad you asked. So without getting too deep into the math, you can throw this data into a binomial distribution and calculate the odds of recovering 9 or fewer fumbles out of 32 opportunities. The end result? 1.003%. ONE POINT ZERO ZERO THREE PERCENT. That is so incredibly unlikely that it could only happen to Kirk Ferentz and the Iowa Hawkeyes. With three games decided by three points or less last year, just imagine if a couple of those fumbles would have gone to the good guys. (Fumbles recovered by the opponent led to points on the opponent's ensuing drive in both the Wisconsin and Nebraska games.)
There is some good news here, however. Regression toward the mean to the rescue! The basic idea of regression toward the mean is that if you have a sampling that is way off from the average, then the second sampling will tend to be closer to the average. What that means for Iowa, is that this year, it’d be likely that we see an improvement in fumble recoveries.
Now this shouldn’t be mistaken for a correction to make the combined number of fumble recoveries over two years (last season and this season) be close to the overall average. I’m definitely not saying that Iowa will recover 25 out of 32 fumble opportunities this year -- that would be wildly improbable, just like it was wildly improbable that Iowa would only recover 9 out of 32 fumble opportunities a season ago. But it is likely that Iowa should recover more like 50% of their fumble opportunities -- that means that they would recover around 15 fumbles this season if there are 30 fumble opportunities. Which I’ll gladly take at this point -- that would be a notable improvement from what we saw last year. Given the talent that Iowa has in the secondary this year, I think that they'll manage to force some interceptions. And if the reports out of training camp about C.J. Beathard only throwing one interception are close to true, then the combination of all those factors could -- should -- point towards the turnover margin being in Iowa's favor in 2015.
For a team that lives in the margins and relies on the little things going right, that might just be a key factor in Iowa's ability to turn things around this season.