BJGE fumbled with the Patriots during a preseason game.
https://www.youtube.com/watch?v=JueRWjZ-Hms
Why isn't run blocking a factor? Football Outsiders ranked the Patriots #1 and #2 in 2010 and 2011, respectively, and BJGE went to a team with a markedly worse run-blocking offensive line (ranked #11 in both years in CIN). I can't see all of his fumbles online, but the first one I see on Youtube is him
getting blown up behind the line of scrimmage. Anecdotal, yes, but I'm not sure how an under-inflated ball would help in that situation.
I apologize if it's listed somewhere, but if they are comparing players with and without the Patriots, why isn't it compared to how other teams' players perform when leaving their teams as well? Are the Patriots clearly outliers there?
Though why look at only 2010-2014? Why wouldn't you go
back to 2006, when the rules change allowed teams to supply their own balls? Well, simply because it would make the stats look worse for the WSJ I would guess. Adding one more year to the sample size adds Ben Watson to the cohort (who you did not leave out, and not coincidentally your totals were far from extreme as WSJ), which would hurt the "analysis" quite a bit. He fumbled 6 times against 167 receptions for the Patriots, and just 1 time against 193 receptions for other teams. That brings the WSJ number of "w/Pats" down from 185.3 to 117.8, and raises the "other teams" up from 77.3 to 82.3. In that scenario, one extra fumble
every 271 touches would result in an identical rate to that group's with the Patriots. That's a bit over one a year, which could easily be attributed to an organizational emphasis on ball protection.
Going back to 2008 allows you to add David Thomas to the cohort. This brings the w/Pats number from 117.8 to 119.3, and "other team" total from 82.3 to 85.9. Hell, simply adding Julian Edelman to the group will bring the w/Pats touches total to 1936, quite close to the "other team" total of 1975 (yes, I'm aware he's only been on the Patriots), and suddenly the "with Patriots" cohort fumbles once every 62.5 touches, while the "other team" cohort once every 85.9 touches. Yes, I'm moving the goalposts here, but that's the point: choosing 2010 as a starting date seems arbitrary, gives a smaller sample, and appears to have been done so to screen out Ben Watson, who makes the shocking "185.3 touches" figure look a far more reasonable 117.8.