A Closer Look At Variance

Yesterday we took a look at our astonishingly low ranking for Adrian Peterson, and we found that not penalizing for variance moved Peterson from 4th in our running back ratings to second.

So what's the deal?

Most running backs have a variance in the 10-13% range, while Peterson's is a sky-high 21%. Is this deserved? Peterson has been injured every year since his freshman year in college. He has an upright running style that is conducive to injuries. Sorry, but he will get injured this year. Drafting him is almost guaranteeing that you will have at least one game with 2 carries for 8 yards before Adrian goes down with another injury. Its gonna happen, trust me. So penalizing his increased injury risk with a higher variance makes absolute sense.

But is the base rushing number that we gave Peterson a good number? Last season, Adrian rushed for 5.6 yards per carry. Our original projection had him at 4.6. Now, 4.6 yards per carry is very, very, good, but is it fair to penalize Adrian a full yard per carry just because his average last season was off the charts. Let's look at some other great rookie rushers, and see what they did in their second season.

Eric Dickerson averaged 4.6 yards per carry as a rookie, then improved to 5.6 yards per carry in his record-setting sophomore season. George Rogers averaged 4.4 yards per carry as a rookie and in his second season. Ottis Anderson went from 4.8 yards per carry to 4.5 yards per carry. Looking at a recent example of someone with similar rookie yards per rush, Maurice Jones Drew had 5.7 yards per rush in his rookie season, then fell to 4.6 in his next season. So even though MJD fell by about a yard per carry, the other guys did not, and Dickerson actually improved by a complete yard per carry.

So we will re-calculate Peterson's numbers with 5.1 yards per carry as the base, which brings him back up to the second running back, even with the 21% yardage variance penalty.