I’m a huge nerd when it comes to statistical research and advanced metrics in sports, but so far my geekiness has primarily stayed in the world of baseball along with some dabbling in college basketball.
The idea is to find out how many attempts (or whatever denominator you are using) it takes before the observed data is half real (skill or true variance) and half noise (luck or random variance).
The most interesting result of the article was that it takes about 5,000 passing attempts (or about 10 season’s worth) before an individual quarterback’s interception rate carries more weight (in a predictive sense) than the league average.
For example, if a quarterback averaged only a 1% interception rate for 10 seasons, and the league average rate was 3%, we would predict that his future interception rate would be 2% (half-way between his personal average and the league).
To me, this was an incredible discovery. It means that NFL quarterbacks actually have very little control over their interception rate. (This has been found to be the case by others as well.)
But what does this mean in a more practical sense for us as Detroit Lions fans?
It means that we should pay very little attention to Matthew Stafford’s 5.4% interception rate as a rookie or his 1.0% interception rate in his (limited) second year. Even his 4.4% career rate carries very little meaning at this point.
It also means that you probably shouldn’t shy away from quarterbacks that threw lots of INT’s last season for your fantasy football team this year. If you find that a guy is getting dinged in the rankigs due to an unusually large number of interceptions, he may be a good sleeper pick because of the expectation that his interception rate will, in all likelyhood, regress toward the league average rate.