This is intended as a companion to
the recent Post article on team improvements and declines at the quarterback position. I looked at the year-to-year change in total EPA at the QB position for each team since digital records began in 2000. The table below lists the change in total EPA from the previous season. For example, Arizona improved by 47 EPA in 2001 compared to the 2000 season.All numbers are from the regular season only. The table is color coded so that the biggest improvements are in green and the biggest declines are in red. The 2012 numbers are adjusted for a 16 game season.
Look at TB's last 6 years... Freeman's rollercoaster !!
You'll notice "the regression to the mean". Almost every QB over 100 fell back the next year by a large amount.
Even the very best QB's regress, Rodgers 143 last year and his -136 this year.
Mannings big year in 2004, a 95, not quite a 100, but he fell off big time to -45 the next season.
Bree's had a 109 in 2011 and a -57 this year.
Rivers had a monsterous 172 in 2004 then fell way off to -93.
Farve goes from 152 to -212.
Seems to indicate, Manning, Luck, Freeman and RGIII will fall-off big time next season.
I think there is some lookup error in the calculation. Indy should have over 200 points of EPA in 2004, right?
Its change in EPA from the previous season...not total EPA
I did a really crude regression analysis on these values and came up with roughly...
8 - 40% [last years change]
So yeah - definitely regression to the mean. But then again, it wasn't a very accurate model - these numbers are all over the place.
Joe-That't pretty interesting. But...did you type in all those numbers on your own? It's a picture file, not a table. Impressive dedication!
I'm surprised DET is positive over last year.
SD 3years of regression