There was a time when I could find the season ending release NE sent out to all the media with NE's own coach's stats, those are the ones I'd like to see Patriots.com publish instead of linking to those pitiful in-game stats compiled by NFL.com. I recall an article this winter from one of our local reporters which mentioned Guyton's Tackles credited him with almost a game's worth more than NFL.com's final numbers, so I know the season ending release is still going out to the media. It's frustrating from a fan's perspective to not have the most accurate data if we're going to dissect our team.
Absolutely agree. However, for me, the big thing is consistency - comparing apples to apples.
So, PFF (whose guys base their numbers on their own study of TV tape) has Chung with 13 tackles in 194 snaps (I misquoted 125 in my previous post). That extrapolates to about 40 tackles in 600 snaps which is right in line with what PFF credits to several other safeties who did have right around 600 snaps --- and to 50-something tackles in about 775 snaps (also right in line) - and to about 65 tackles in 970 snaps, just a bit less than the guys who did actually get that many snaps (mostly top starting veterans). So, PFF's numbers appear to be internally consistent, at least.
The numbers on NFL.com are internally consistent as well and they match the numbers from PFF very closely,
IF you subtract ST tackles (from PFF) and NFL.com assists (NFL.com appears to be much more liberal in crediting assists) from the NFL.com "Combined" number for a given player. For Chung, this yields 15 tackles (PFF has him at 13).
Kyle Arrington is credited with 18 TT by NFL.com (14 + 4 assists) even though he had NO tackles from scrimmage AFAIK. PFF has Arrington with 13 ST tackles and 0 tackles from scrimmage. So, 14/13 is pretty close an makes it clear that NFL.com
does include ST tackles in their total.
McGowan: NFL.com = 79 TT (59 + 20 assists). Subtract PFFs 7 ST tackles from 59 = 52 and PFF credits McGowan with 54. Again, 52/54 is pretty close.
The numbers worked out within 1-3 tackles either way for each of several other players I chose at random.
The point is, of course, to use one set of numbers exclusively for the sake of consistency - apples to apples. I prefer PFF only because they break things down into more detail (including snaps played at what position, number of times thrown at with completion percentage, etc.). Perhaps the coaches numbers would be more accurate, but perhaps more accurate ONLY for Pats players and they might not be quite as trustworthy as those from neutral observers for comparing Pats players to the rest of the league.