The Screwy Way the NFL Counts Tackles

[TodMaher]

An interesting new post on the Turney/Webster/Willis "Pro Football Journal" about tackles:
http://nflfootballjournal.blogspot.com

Now, I'm not going into either the whole "do tackles stats mean anything" or how accurate are the "official" tackles. But I do have a few quibbles.

One, play-by-plays are NOT now called gamebooks. They are a part of each gamebook.
The other is the the screwy way the NFL counts tackles. What do I mean. Well, the NFL does not simply count tackles and assists - even though that's how the totals are presented for each game and season. They're are actually three types of tackles - to of which are counted under the "Tackle" column. In other words, Tackles does not equal Unassisted Tackles.

The three types of tackles that the NFL counts are:
Solo Tackle
Assisted Tackle
Tackle Assist

It is the "Assisted Tackle" that gets lumped in the with "Solo Tackle."

Assisted Tackle and Tackle Assist occur on the same play. So, when you see two players in a play-by-play listed on the play the first player gets credited with an Assisted Tackle and the second with a Tackle Assist.

How do I know this? Well, you have to dig deeper than the gamebook and look at a file on the NFLGSIS site for each game the NFL calls "Play Statistics Report."

[JohnTurney]

One, play-by-plays are NOT now called gamebooks. They are a part of each gamebook.

The other is the the screwy way the NFL counts tackles. What do I mean. Well, the NFL does not simply count tackles and assists - even though that's how the totals are presented for each game and season. They're are actually three types of tackles - to of which are counted under the "Tackle" column. In other words, Tackles does not equal Unassisted Tackles.

The three types of tackles that the NFL counts are:
Solo Tackle
Assisted Tackle
Tackle Assist

It is the "Assisted Tackle" that gets lumped in the with "Solo Tackle."

Assisted Tackle and Tackle Assist occur on the same play. So, when you see two players in a play-by-play listed on the play the first player gets credited with an Assisted Tackle and the second with a Tackle Assist.

How do I know this? Well, you have to dig deeper than the gamebook and look at a file on the NFLGSIS site for each game the NFL calls "Play Statistics Report."

My understanding is those "PSRs" began in 1998. Though I have spent no time looking at them since there was never a tackle total or forced fumble total, etc on them. But I have always thought it odd how the NFL scored tackles, especially tackles for loss

Nick and I have been planning on writing about that. With the NFLGSIS a tackle for a loss is a solo or assisted tacked (not a tackle assist) that goes for a loss run or pass . . .AS LONG AS IT'S NOT a FORCED FUMBLE. If a guy tackles a running back for a loss and the rb fumbles, it's not a tackle, there can only be one tackle per play, the person downing the fumble recoverer gets credit for that.

But while it makes the computer program work and all the things add up, it takes away from the ever loving point, that a defender made a play in the backfield for a loss and the fact that he forced a fumble is all the better, but the NFL doesn't see it that way.

Stats, LLC, on the other hand, counts tackles the old way, if there are two names in the () then each gets a half sack. Or a half stuff. So, a Geno Atkins could have 7.5 stuffs in a season for example. Nick Webster has expanded on what Stats does by adding in pass stuffs. NFLGSIS counts tackles for loss in both run and pass (as long as it's a lead tackle or solo and it's not a forced fumble) and that includes sacks. So there is double counting.

With NFLGSIS a player could get a solo tackle on a QB on a pass play. He gets 1 tackles 1 tackle for loss and 1 sack.
Stats would have 1 tackle and one sack. For them a stuff is on run plays.

As far as the name for Gamebooks and play-by-play, I am sure you are right and totally take your word for it. It's just that when you go to team sites to get old play-by-plays they are are linked as "Gamebooks" even though they never were called that in the old days.

I never heard of "Gamebook" until about 1996 or 1998, whenever NFL.com started posting them on that site. I'd always called them play by plays but for the purposes of the article I just went with Gamebook. But I suppose I could do a correction...

But I totally appreciate the info on the "assisted tackles" that was a new one for me. I can only guess they are counted as an assist?

NFLGSIS has tackles, assists and then "combined".

I knew about the two names in the () thing was scored different between Stats LLC and NFLGSIS. That is why I think there are different totals for Ray Lewis
PFR =1562 solo-- 493 assists-- 2055 combined
Fox =1567 solo-- 493 assists-- 2060 combined
ESPN =1558 solo-- 492 assists-- 2050 combined
I think the scoring differences are partially that, though I will never spend the time to find out for sure.

Anyway, good post, tons of food for thought

[TodMaher]

"Gamebook" is a term the NFL came up with in the mid-1990s when they went online. Basically, I guess they needed to call final stats package for each game something, so that's what they came up with. Or maybe it was internally called that before then, but who knows. Then guy who now runs the NFLGSIS doesn't answer questions like that anymore (well, at least when I ask).

Before that there was the "gamebook" and the Official Score Sheet. But with digital age the NFL decided to do away with Official Score Sheet make the gamebook serve the same function.

[JohnTurney]

"Gamebook" is a term the NFL came up with in the mid-1990s when they went online. Basically, I guess they needed to call final stats package for each game something, so that's what they came up with. Or maybe it was internally called that before then, but who knows. Then guy who now runs the NFLGSIS doesn't answer questions like that anymore (well, at least when I ask).

Before that there was the "gamebook" and the Official Score Sheet. But with digital age the NFL decided to do away with Official Score Sheet make the gamebook serve the same function.

Thanks, that's kind of my memory of the timeline...Gamebook, which I never heard of showing up. I just adopted in the article because that is the "recentism" needed to convey ideas to younger folks. But I am sure you are right, play by play is still written aat top of those 4-5 pages with the line by line description.

Also gamebook is seems a bit cleaner than play-by-play with the hyphens, etc...

[NWebster]

Yeah, so a sack that's a forced fumble isn't a Tackle for Loss because the QB wasn't taken down with the ball, a Quarterback drops back to pass and is tackled for an 8 yard loss, Sac + TFL, but drops back to pass, is chased and is tackled at the line of scrimmage, Sack (for 0 yards) but no tackle for loss. Inconsistencies abound.

[TodMaher]

Yeah, so a sack that's a forced fumble isn't a Tackle for Loss because the QB wasn't taken down with the ball, a Quarterback drops back to pass and is tackled for an 8 yard loss, Sac + TFL, but drops back to pass, is chased and is tackled at the line of scrimmage, Sack (for 0 yards) but no tackle for loss. Inconsistencies abound.

Well, I have always disliked all those sacks of zero yards. If it did not lose statistically lose yardage then it's not a sack. I don't care if he's one nanometer short of the line of scrimmage - it ain't a sack.

[JohnTurney]

Yeah, so a sack that's a forced fumble isn't a Tackle for Loss because the QB wasn't taken down with the ball, a Quarterback drops back to pass and is tackled for an 8 yard loss, Sac + TFL, but drops back to pass, is chased and is tackled at the line of scrimmage, Sack (for 0 yards) but no tackle for loss. Inconsistencies abound.

Yeah, oddities everywhere. I don't mind a sack for zero, not sure who gets to define "sack" though the term, iMO, was first used in George Allen playbooks but maybe he got it from Clark Shaughnessy, or maybe he didn't. I just wish it was the same as stuffs

A tackle for no gain is as important as a tackle for loss in that no damage was done. If it's 4th and 1 and there is a no gain, then defense wins. So in a sense, that is a run stuff. But Stats, Inc (what it was called at the time) chose tackles for loss and called them stuffs.

For years coaches has listed tackles for loss, so I guess that is where it got its impetus. I guess that's why we all watch the games, because football, unlike baseball has all these quirky things.

That is why I chuck a bit when these guys who do all the QB evaluations come up with metrics slicing passing stats here and there. Ballas are marked at the yard line, there are no half-yard gains or passes...it's a step function and that ends all those things, really, the raw data is too imprecise to make too much of footballish sabermetrics

[Reaser]

That is why I chuck a bit when these guys who do all the QB evaluations come up with metrics slicing passing stats here and there.

I roll my eyes but probably same difference.

Sacks of zero yards sums up capital F Football v. football statistics.

In Football, using John's 4th and 1 example, no gain is a win for the defense. 100%.

Statistically, 0 yards is neutral. Didn't lose yards, didn't gain yards, there's 3 categories there; did you lose yards, did you get no yards or did you gain yards. While in Football you either got that 1 yard on 4th and 1 or you didn't, it's black and white.

[rhickok1109]

This is rather tangential, but there used to be a college football commentator--maybe Bob Griese--who would talk about "gaining negative yards." He would say things like "the one thing you really want to avoid is gaining negative yards."

He never mentioned that losing positive yards is just as bad

[LJP]

I like the three categories for tackles, its just a shame they don't record them that way in the Gamebook. The combined (COMB) TKL (ST+AT) and AST (TA) numbers in the Gamebook are somewhat misleading since the assisted tackle/tackle assist is essentially counted as 2 tackles on the same play. If an assisted sack counts as half to each player, why doesn't an assisted tackle/tackle assist?

There is far too much duplication in the defensive stats. Sacks count as so many different things that the numbers are just reported as is and not necessarilly what actually happened.

Personally, I would separate the different stats so that one play doesn't count in multiple categories, e.g.:-

Sacks (SS/AS/SA) would not count as Tackles, QBH or Stuffs. AS & SA count as half a sack to each player. I have no problem with the 0 yard Sack. If there was an intent to pass and the player is brought down for no gain or a loss, then it's a Sack (but not a Tackle, etc).

Tackles (ST/AT/TA) with AT & TA counting as half a tackle to each player. Would excludes Stuffs.

Stuffs (STF) would be Rushing/Receiving plays that result in 0 yards or less. Half each for two players making a Stuff. No more TFLs.

QB Hit (QBH) would exclude Sacks.

Pass Defensed (PD) would exclude Interceptions.

Forced Fumble (FF) would still count as a Tackle, Sack or Stuff.

Looking at the Bears defensive numbers in the Week 1 Gamebook they were credited with - 46 Tackles (36 TKL + 10 AST), 2 Sacks, 4 TFLs, 5 QBH and 4 PD. Using the above criteria that would equate to 34 Tackles (30 ST + 4 AT + 4 TA), 2 Sacks, 5 Stuffs (2 for a loss), 2 QBH and 4 PD. On an individual basis, Sam Acho was credited with 2 TKL + 1 AST, when actually he had 2.5 Stuffs, with all of his "tackles" resulting in no gain.

The Gamebook credits the Bears D with 59 snaps, although 4 plays were nullified by penalties. Of the remaining 55 plays, 9 resulted in incompletions (4 PD). Of the remaining 46 (23 rushing, 21 receptions + 2 sacks), 5 resulted in no tackles, etc (1 TD run, 1 TD pass, 1 QB Scramble/Slide & 2 QB Kneeldowns) leaving 41 plays. My method has the Bears with 34 Tackles, 5 Stuffs and 2 Sacks (i.e. 41), whereas the Gamebook says they had 46 tackles, 5 TFLs & 2 sacks. Which is a more accurate view of what actually happened?