People like to torture stats, don't they? While the author admits that Welker and LaFell fumbled MORE for the Patriots than their other teams, he uses the wrong stat to measure this. He is looking at fumbles recovered. What he should use is just plain fumbles.
Plus, he has such a small sample to do this with. For instance, he says Blount fumbled more with the Steelers.
Uh, Blount carried the ball 65 times with the Steelers. 1 fumble.
For the Patriots, he carried the ball 60 times. 0 fumbles.
Does this seem like a significant stat for you?
Not to mention the fact that officially, Blount ran the ball 150 times last year, and fumbled the ball 3 times. Which is a very very high rate, much higher rate than in any of his years at Tampa Bay.
So what's the difference? The writer is measuring fumbles recovered, and not fumbles. If he takes recoveries out of the equation, then it will show that Blount fumbles MORE with the Patriots.