[discostu]

Fell apart right near the beginning: Based on the assumption that fumbles per play follow a normal distribution,

Fumbles per play do NOT follow a normal distribution curve. Some players are more prone to fumbling, some coaches are more prone to allowing sloppy play that makes the team more prone to fumbling, some coaches put more work into having the defense practice forcing fumbles which not only means their opponents fumble more it means their offenses generally fumble less. Fumbles are NOT random events and do not distribute accordingly. Fumble recovery does tend to be random, but that’s not until after the fumble.

[Berlin_Freeper]

You are wrong on the obvious and wrong when you arrive at

Fumble recovery does tend to be random, but that’s not until after the fumble.

There are practice techniques and skills used in fumble recovery like touching a ball in the air so it drops straight down and falling on it, pulling it in to the gut while being balled up so it can't be pulled away etc... So your points can be made for both events.

But back to your original error... as noted it is obvious that fumbles and fumbles lost are not a normal distribution curve and there are skills and practices involved. Otherwise we would be waiting at play count 55 and say OK here comes the fumble.

So to make that observation is really quite pointless to the data being presented. Data, which by the way has skill, practice, etc, inherent in the charts being presented in the whole and is looked at from several ways. You can't separate skill, practice, etc from the data because they helped create the data and in the case of the Pats this includes deflated balls (least we forget). Presenting it in a proper way as a chart is how one can analyze such numbers.

Those things you are transfixed on are actually noted in the second to last paragraph, that judging from the knee-jerk rejection in your post ("Fell apart right near the beginning") could not be bothered reading before wasting my time.

I'd seriously appreciate it if you read the whole thing next time, it's Saturday and I have a family tugging at my elbow... instead I am correcting your errors which you couldn't be bothered with yourself. Thanks.