Take Away His 'X' Best Carries and He's Average

Actually, I learned just today that I don't need to show that, before I can I argue that its likely.

 

Interesting. I learned today that showing it may not be enough

 

So I went ahead and did this, correlating yards per carry with the carry #, over all of 2014's games, and that correlation is -0.30. So there is a relatively meager and inverse correlation between the carry number and the average number of yards gained on it in 2014.

91% of the variance in yards per carry is associated with things other than the carry number. The carry number explains 9% of the variance in yards per carry.

So if one's argument is that Miller "wears down" as he carries the ball more, or that he "gets worse" or is "less effective" as he carries the ball more, be aware that when the carry occurs (either early or later in his tally of carries in a game) is associated with only 9% of the variance in his yards per carry. It's true that the relationship is indeed negative (inverse), though it's very meager.

 

Lamar Miller 1,300 yards and 9 touchdowns next season.

 

Assuming the calculations are correct, good job. Since you seem to have the data, can you also check and see whether the correlation is more negative than -0.30 if you just look at carries 11-20? It would give us a better indication of what to expect if he carried it 20+ times. Oh, and what % of total carries are 1-10 vs. 11-20 (just raw number of carries)?

 

The correlation for carries 11 through 19 is -0.31 (0.24 for carries 1 through 10). Carries 11 through 19 were 28.7% of the total carries; carries 1 through 10 the other 71.3%.