Writers of Pro Football Prospectus 2008

02 Jan 2006

2005 Full-Season Quick Reads

This week's Quick Reads looks at the best and worst players for the year, not just for Week 17. That means that all the full-season individual stats pages are now online. Consider this an open thread to talk about not just the Quick Reads column but also all the DVOA and DPAR ratings for individual players in 2005.

Posted by: Aaron Schatz on 02 Jan 2006

43 comments, Last at 05 Jan 2006, 8:00am by big_adventure

Comments

1
by Ferg (not verified) :: Mon, 01/02/2006 - 9:03pm

Hmm, shouldn't Mark Brunell be the most improved QB? Manning the Lesser went from -13.3 DPAR to 36.5 (+49.8), Brunell went from -10.2 to 47.1 (+57.3).

Wacky Mirror Universe Joey Harrington, Snoil

Brilliant!

2
by Shelley (not verified) :: Mon, 01/02/2006 - 9:12pm

Little question concerning Rusher McFumbles. You've got him listed as having an interception, but that was negated by a roughing the passer penalty. So I guess I'm just asking if this is one of those instances when you don't go by the official NFL stats.

3
by David (not verified) :: Mon, 01/02/2006 - 9:18pm

A note on Favre - those "meaningless" game-ending interceptions do have a meaning, or at least a cause. No team was on the losing end of as many one-score games as the Packers this year, which means that, week in and week out, Favre was put in the position of throwing long, desperate lobs from his own territory in the final minute of play with the Packers down by five, or three, or one. So, ironically, his interception total would be lower if his team had only sucked more.

4
by admin :: Mon, 01/02/2006 - 9:18pm

Yeah, I missed Brunell because I was looking at the "minimum 100 passes" table. Cassel's interception came against San Diego in Week 4.

5
by Dean from Oz (not verified) :: Mon, 01/02/2006 - 9:40pm

Its probably been looked at already, but maybe there is a correlation between the highest ranked defensive teams in creating interceptions and the fact that they played against Favre?

6
by jackg (not verified) :: Mon, 01/02/2006 - 10:03pm

(5) my thoughts exactly, especially the teams that played him twice, if you take their games against Fav-re out, and normalize things, do they still look as strong? Not to make too much out of an n of 1, but I don't care how good the Bears D is, that pick-6 he through to them last week was clearly bad QBing not good D. Nice article though, great summary of the year!

7
by Joon (not verified) :: Mon, 01/02/2006 - 10:09pm

i posted this question in some other thread and i don't remember it ever getting answered... do INTs count against the DPAR/DVOA of the intended receiver more than incompletes do? it seems that the QB and the WR share responsibility for an INT just as they do for an incomplete, but i'm not exactly sure how the two should be weighted. probably something like 75% QB's fault, 25% WR's.

8
by Vash (not verified) :: Mon, 01/02/2006 - 11:25pm

Low-DPAR players with high DVOA seems to be a theme of the Pittsburgh passing game.

I guess that's what happens when you only throw the ball 20 times a game.

9
by Travis (not verified) :: Tue, 01/03/2006 - 12:26am

Bottom 5
1. Alex Smith, 49ers: -72.4 DPAR (-66.5 passing, -5.9 rushing)

Worse than his standard statistics made him look: Alex Smith, Buccaneers.

DPAR is clearly biased against players named Alex Smith.

One question: According to NFL.com, Shaun Alexander ran for 1880 yards and Tiki Barber 1860, but the RB page has them with 1892 and 1864. Their carries on FO are the same as on NFL.com. I know that you adjust QBs' stats to remove failed Hail Marys and kneeldowns (e.g. Favre with 26 int rather than 29), but I can't figure out why rushing yardage would be different. Also, Thomas Jones's stats don't seem to include his 12 carry, 62 yard game against Minnesota.

10
by BlueStarDude (not verified) :: Tue, 01/03/2006 - 12:28am

re: "[Wiggins'] numbers were hurt by his schedule, as the Bears and all the NFC East teams were very good against tight ends."

Was this a typo? The Vikings didn't have to play all of the NFC East teams this year.

11
by JonL (not verified) :: Tue, 01/03/2006 - 12:50am

Perhaps I'm confusing things, but couldn't an argument be made that the reason the Bengals led the league in interceptions is because they faced quarterbacks like Favre? I know there was somde discussion at one point as to whether luck has a hand in defensive INT totals.

12
by Catfish (not verified) :: Tue, 01/03/2006 - 12:54am

Re: 9

I don't know for sure, but I think that sometimes the official totals and the totals from the play-by-play don't add up.

13
by paytonrules (not verified) :: Tue, 01/03/2006 - 1:32am

I'm jumping ahead - but how did the Bears defense get better after week 17? Playing all the backups - and getting pasted - made them offically according the the stupid NFL stats the second best defense. As it is they went from (going from memory) 28% to 31% while everybody else didn't so much fall as plummet. Weren't the Giants #2 last week?

14
by Glenn (not verified) :: Tue, 01/03/2006 - 1:41am

Shelley-
Rusher McFumbles did some mopup work against the Chargers and got picked once in that game.

(Guess we should mention that McFumbles is the Pats' Matt Cassell)

15
by admin :: Tue, 01/03/2006 - 2:45am

With the season over, side boxes, with the exception of "WEIGHTED DVOA," have now switched to the total season ratings, not the weighted ratings. That's why Chicago "got better."

Wiggins was just a brain freeze. It happens. I sent a fix.

As to yardage differences, Elias has a tendency to change official numbers weeks later, or I could have a play wrong because of a play-shortening penalty. I often spend the spring cleaning up all these little things. I'll have to check into the Thomas Jones thing.

16
by putnamp (not verified) :: Tue, 01/03/2006 - 4:47am

Is there a way to weigh *games* situationally, not just plays? For example, is there any reason that Indianapolis', or Peyton Manning, should suffer from a lower pro-rated DPAR/DVOA (obviously you can't just make up a performance for him) based off of a game in which they clearly didn't even try?

Much like you would rate a 3rd-and-15 play based on its performance relative to other 3rd-and-15 plays, what about a Week 16 game in which nothing matters for the team in question?

17
by Israel (not verified) :: Tue, 01/03/2006 - 5:18am

Very impressive to see Willie Parker and Jerome Bettis as #15 and #17 in DPAR, but much moreso to see Jerome as #7 in success rate and Parker as #11.

There was a question in the game thread about Jerome's success rate in week 17. Is there an answer on that?

18
by paytonrules (not verified) :: Tue, 01/03/2006 - 10:54am

Thanks Aaron.

Love the drive stats page BTW. It reaffirms something I've known all year. The Bears offense is so bad (how bad is it?) that it's like every time they play they are playing against the best defense in the league, that is a mythical one even better than their own.

19
by pawnking (not verified) :: Tue, 01/03/2006 - 11:11am

The Bears bring up an interesting point. Would it be better to have a totally awesome and dominant defense with a replacement level offense, or vice versa? Any historical evidence we can use here?

Speaking for myself, I can think of any number of great defenses with mediocre offenses which won the superbowl, but offhand I can't think of any great offenses which won without at least an above average defense. Am I mis-remembering?

20
by senser81 (not verified) :: Tue, 01/03/2006 - 11:44am

I liked the Favre comment in the statistic-saturated article: "Better than his standard statistics made him look: Brett Favre, Packers. Yes, he threw 32 interceptions..."

Favre is so much better than his standard statistics, that when he throws 32 interceptions, he really is only throwing 29 interceptions!

21
by admin :: Tue, 01/03/2006 - 12:08pm

There's more to Favre than just the Hail Mary plays, but I wasn't writing a dissertation here.

The Thomas Jones thing is solved. The data parser seemed to have just completely missed the last Chicago-Minnesota game. I have no idea why. I'm going to refigure all the numbers now.

22
by Bill Moore (not verified) :: Tue, 01/03/2006 - 12:31pm

When you look at both JJ Arrington and Marcel Shipp making the worst RBs list, there must be something with the Line and the System. However, I will note this. When charting a couple of the earlier ARI games, I closely watched JJ and Shipp. The most compelling thing I noticed about Shipp is the guy doesn't follow his lead blocker. They run lots of FB sets for him, and the guy continously chooses to find his own holes. By looking at the stats, its easy to see he doesn't find many.

23
by Oswlek (not verified) :: Tue, 01/03/2006 - 1:16pm

I am a Pats fan, but I have a question.

How did Miami's terrible opponent's pass defense hurt Frerotte, yet Deion Branch "played against a number of tough defenses?"

Didn't they face the exact same defenses, excluding themselves? Was Miami's pass D so much better than the Pats' that the overall DVOA of those D's were significantly different?

24
by Led (not verified) :: Tue, 01/03/2006 - 1:17pm

I've heard Madden and other announcers talk about how good Brady is at running sneaks, but has anyone ever tested that? Sure he gets a lot of first downs that way, but that's because they run it a lot. Aaron, do you have numbers on other QBs running sneaks? Personally, I can't remember the last time a QB sneak (any QB sneak) did not pickup the necessary yardage. Seems like, as with many other things, Belichick and company recognized an effective strategy and used it. Other teams should as well. It drives me crazy when the Jets give the defense a chance by handing it off when they only need a yard.

25
by Daniel Warehall (not verified) :: Tue, 01/03/2006 - 1:31pm

Oswlek,

Also, NE faced Pittsburgh and Indy, while Miami faced Cleveland and Tennessee.

26
by big_adventure (not verified) :: Tue, 01/03/2006 - 1:31pm

19 -

The problem is the difference between "mediocre" and "replacement". Mediocre would tend to mean average. There have, perhaps, been some champs that only had an average offense to go with an all-time defense. But I can't think of any team that hoisted the big trophy without at least being "average" on both sides of the ball. Even Baltimore had that going for them. TB was better than average on offense. The other "all-time" defenses that won super bowls, off the top of my head, include the 72-73 Dolphins (best running team in the league, HOF QB, RB, WR and O-line), the mid-70's Steelers (HOF QB, running back, receivers), the mid 80's Bears (HOF RB, killer line) and Giants (very good QB, running back, great TE, awesome line), and those afformentioned Bucs and Ravens (both at least average, and with top kickers).

Having a truly replacement-level offense (like the 05 Bears with Orton starting) will get you run out of town on rails by a good team. Who cares if you hold them to 10 points, when you can't score 7?

-Sean

27
by senser81 (not verified) :: Tue, 01/03/2006 - 1:31pm

Has anyone ever done an analysis on RB receiving stats? I thought it odd that Rudi Johnson was the "worst receiving RB" in the NFL, and teammate Chris Perry (the supposed 'receiving back') was the "2nd worst receiving RB" in the NFL. Both players had rather high receiving percentages (77% & 82%). Could it be that the Bengals throw downfield more to their WRs and less to their RBs? I'm not trying to say that Rudi Johnson is a good receiver; I'm trying to see how much people penalize Johnson for his lack of receiving stats.

28
by pawnking (not verified) :: Tue, 01/03/2006 - 3:13pm

Sean, I meant mediocre as being below league average, not replacement level, which would be worst in the league.

Anyway, looking at the past few years of SB winners through DVOA:
04 - NE had a #4 offense and a #6 defense
03 - NE had a #3 def and was below average (-.8) on offense
02 - TB had the #1 Def and an above average offense (#15)
01 - NE had a slightly better than average defense (-3.4, 19th) and a below average offense (-8.7) What an amazing feat to win the SB that year!
00 - Baltimore had the #1 def and a below average offense (-6.5)
99 - St. Louis had the #1 offense and #4 defense
98 - Denver had the #1 offense and a below average defense (4.1%)

What can I draw from this? Well, taking the 01 NE team out of the picture, every SB winning team had either a top defense of, in the case of Denver, a top offense. A few had both. It seems that if you had to choose, you'd rather have a great defense than a great offense to win the SB.

In the future, I plan to look more at this, seeing if these indications in the past might not give some idea of future games' results.

Has there ever been an "Every Play Counts" article on the 01 Superbowl? SL had the #1 offense and the #3 defense, and lost to what was, by most measures, a mediocre team. How the heck did that happen?

29
by dbt (not verified) :: Tue, 01/03/2006 - 4:09pm

Muhsin Muhammad actually declined even more, but Horn was playing with the same team and the same quarterback, while Muhammad had to endure a year of the Kyle Orton Experience.

Bears vs. Packers, week 16, 4th Quarter:

4-1-GB28 (3:51) R.Grossman pass incomplete to M.Muhammad (N.Collins).

That was a sure touchdown that hit him in the hands, delivered perfectly into coverage.

Muhsin dropped tons of passes that should have been caught, and I'm pretty sure Chicago is getting ready to turn on him in a big way if he keeps it up in the playoffs.

30
by Kibbles (not verified) :: Tue, 01/03/2006 - 4:30pm

Re #24: Personally, I can’t remember the last time a QB sneak (any QB sneak) did not pickup the necessary yardage.
I always thought the same thing, until I saw Josh McCown try the QB sneak against Indy this week and come up short (and fumble, to boot).

Aaron, is it true that you take out INTs that came in hail mary situations? If so, why do you leave INTs that come on 4th down? Jake Plummer threw an INT against Miami on 4th down that was actually a rather decent play. If they had just batted it down, they would have gotten the ball in great field position, but the INT was so far down the field that it was effectively a long punt.

I think that 4th-and-long when trailing big and late should be discounted, too, if hail marys are discounted.

Don't mind me, though. I'm just trying to get Jake Plummer on that short list of QBs who cut their INT totals by over 2/3s that you mentioned earlier this season. So it's either get him credited with one fewer INT this season, or somehow invent another INT last season. ;)

31
by pawnking (not verified) :: Tue, 01/03/2006 - 5:03pm

And RE: INTs at the end of the half/game: I understand that as a counting stat, INTs are flawed, but 32 INTs are 32 INTs no matter what the situation. It is difficult for me to believe that we can effectivly discount Favre's INTs because they came against good defenses/end of the game, whatever.

Lookit, the article quotes this: "In total, Favre had nine games against teams ranked in the top 11 in interceptions, and he threw 20 of his interceptions in those nine games." Soooo, what we're saying is that if you look at 7 games against poor defenses, he still threw 12 INTs, which translates into 27 INTs over a 16 game season. That's still a whoooooole lot of interceptions. No one else had more than 17. In the past 4 seasons, no one had more than 23.

And then let's examine this gem: "By the way, three of Favre's interceptions were meaningless end-of-game plays. Drew Bledsoe had two of these and nobody else had more than one." What if he had not thrown those INTs? What if, instead, they were game-ending TDs? Would they still be "Meaningless?" And doesn't Favre have some responsibility for training in close games which lead to those "meaningless" INTs?

I recall a conservative commentator once said "We damage our credibility if we defend Nixon too much." I think many fine analysis damage theirs be defending Favre too much. Maybe he's not as bad as his stats seem to indicate, but those stats were pretty darn bad, so that's not really going out on much of a limb.

32
by JMM (not verified) :: Tue, 01/03/2006 - 5:24pm

"Muhsin Muhammad actually declined even more, but Horn was playing with the same team and the same quarterback, while Muhammad had to endure a year of the Kyle Orton Experience. "

So "the Kyle Orton Experience" is worth more in "excuse me points" than "the Katrina Experience?"
Hmmm....

33
by putnamp (not verified) :: Tue, 01/03/2006 - 9:07pm

Kyle Orton once ravaged a small rural village in southern Illinois, and there are countless other accounts of similar occurrences that were never properly attributed to a specific cause; however it's likely Kyle Orton was involved there as well. Think of him as the anti-Chuck Norris, if you will.

34
by putnamp (not verified) :: Tue, 01/03/2006 - 9:12pm

#28,

Perhaps it's harder to couple a top offense with an above-average defense than vice versa. Top-tier offensive stars make a lot more, I would venture, and so going for a top-tier offense would leave you with fewer resources on defense.

35
by NF (not verified) :: Tue, 01/03/2006 - 10:17pm

Joe Horn was ineffective at the start of the season, got injured a few weeks in, and did not return to form before the end of the season. Joe Horn's decline was far more significant than Mushin's.

That Muhsin drops a lot of passes is something that could be analyzed in terms of its effect on the DVOA of Orton and Grossman if passes that were obviously incomplete because of poor handling by the receiver were recorded in NFL play-by-play. Also, some passes are not just not caught or dropped, but are actually accidentally tossed by receiver before they start taking off. I'm thinking in particular of Randy Moss tossing a football out of his arms on what would have been a long completion in the season opener against NE. That is as obviously the receivers fault as fumbles are.

If there was only some data on passes that were not complete because of poor ball-handling by the receivers, Aaron could see if dropped passes are statistically significant on the DVOA of QBs.

This discussion of dropped passes has reminded me of the STL@SEA wild-card game last year, which Seattle lost on the last play of the game, in an impressive display of bad karma, when one ogf the Seattle WRs dropped a perfectly thrown pass in the endzone in the last play of the game.

Something I've been meaning to recommend is that if end-of-half or end-of-game interceptions are not counted in DVOA statistics, just as kneel-downs are not counted, every end-of-game play should not be counted. They are usually a high-risk high-reward gamble that is done only a handful of times in a season by any given team, and its success or failure is heavily affected by random chance, so it is essentially the same as a defensive fumble recovery in analogy of predictive value.

36
by Chris (not verified) :: Tue, 01/03/2006 - 10:43pm

In terms of top defenses vs top offenses in the Super Bowl

One of the things charted by Aaron that he talked about is that from year to year, defenses are much less consistant than offenses are.

I take that to mean that a Defense can turn around much quicker (and consequently, fall apart much quicker) than an offense can. We've seen this, when defensive coordinators from good defensive teams and go to other teams and normally within the year can dramatically turn it around. This seems to indicate that personnel don't have as large effect on defense as coordinators, scheme, and playcalling do.

There is no indication that defensive units suffer signifigantly more turnover in their players than offensive units do.

What it could mean is that cap-wise, its more fiscally possible to field a good defense and an average offense than the other way around, as well as depth wise.

37
by putnamp (not verified) :: Wed, 01/04/2006 - 12:23am

#36,

There is no indication that defensive units suffer signifigantly more turnover in their players than offensive units do.

What it could mean is that cap-wise, its more fiscally possible to field a good defense and an average offense than the other way around, as well as depth wise.

Not sure if that was in response or simply on the same train of thought as what I was saying, but .. that's more or less what I was saying :)

38
by LnGrrrR (not verified) :: Wed, 01/04/2006 - 11:10am

Hey, just a question...why is Cassel known as Rusher McFumbles? :)

39
by Andrew (not verified) :: Wed, 01/04/2006 - 12:28pm

Why ISN'T Mike McMahon known as Rusher McFumbles?

40
by Starshatterer (not verified) :: Wed, 01/04/2006 - 12:35pm

LnGrrrR (#38 )--

The preseason.

Andrew (#39 )--

Mike McMahon is more of a Scrambler McPicks.

41
by Andrew (not verified) :: Wed, 01/04/2006 - 12:51pm

senser81 #27:

A really good running back can both run the ball, block, and make receptions out of the backfield for something other than 6 yards on 3rd and 8, or 12 yards on 3rd and 16.

Think Marshall Faulk, Tiki Barber, Brian Westbrook, Edgerrin James, Larry Johnson, Priest Holmes, Charlie Garner.

This allows said back to play any and all downs, and for the playbook to be mixed up with runs or called passes to the running back on almost any down, and discounts the telegraphing of play intention by reading the number of the man in at tailback.

42
by NFC Central Freak (not verified) :: Wed, 01/04/2006 - 2:22pm

pawnking:

As a devoted follower of the NFC North nee Central I tape and review all the games involving any team in the division.

Also, as you may know and if not should be aware I am a "fan" of Favre. I think he has done more with less than any QB this side of John Elway.

That being said Favre's season was defensible based on opponent, circumstances, teammate failures, etc. UNTIL the second time they faced the Vikings. From that game until the LAST game of the season Brett Favre was either inhabited by the spirit of Randy Wright or decided he just plain didn't give a ***k.

There were throws that were not simply bad. They were inexplicable. I have followed his entire career, and I cannot fathom what he was thinking on multiple terrible passes.

Passes that were not simply underthrown or overthrown. Passes that were simply tossed around with no regard whatsoever of the consequences.

Many here have used the term "Favred it" to describe a force into coverage resulting in an interception.

Well, for a decade Favre WON those battles far more than he lost. His ability to wedge a ball into the narrowest of space was amazing. And he had to do because his receivers from 1999-2003 rarely managed sufficient separation to give Favre much leeway. ANY mistake resulted in an interception that the uninformed fan would blame on the QB.

But that was THEN.

Folks talk about what Favre was missing but he DID have a solid receiver in Donald Driver who worked his *SS off to get open. He did have two Pro Bowl quality tackles who were consistently outstanding. Clifton and Tauscher are excellent players. He had a running game materialize out of nowhere for multiple games.

I admire Favre. I think he has done a fantastic job of leading his team over the years.

But I have absolutely no idea what he was doing at the end of November and all of December.

If he was trying to get Sherman fired (which contradicts his public statement of support for the man) he couldn't have been more effective.

GB COULD have won several games in that stretch if the QB had kept from being a blithering idiot with the football. That obviously didn't happen.

For about half the season Brett Favre was having a heroic finish to a wonderful career.

Then he tanked. It wasn't physical. I am telling all of you he can still move and throw.

He simply decided at times during the game to be a moron.

He's either bipolar or unprofessional.

Either explanation makes me sad........

43
by big_adventure (not verified) :: Thu, 01/05/2006 - 8:00am

28 -

Another for Pawnking - :)

Well, by those numbers, we have one team driven by their great offense and mediocre (average) defense, three driven by their defenses and average-ish offenses, and two teams great at both and one team that was one of the all-time fluke winners.

However, and DVOA does not exist to back me up for some of this, I would say that going back thorugh the 90's, we would see a slightly different tale. The 49'rs in the 90's with a not-over-concussed Young were brilliant on offense and just OK on D. The Cowboys were good at both facets. The Redskins were a good offensive team, and I don't remember that D being outstanding. The Packers were close to the best offense in the league when they won. Denver was ALSO an offensive team in 97. Only the Giants, in th other 8 199* superbowls, were a defense-first team.

Maybe it IS something endemic to a d-first team in the modern, capped NFL. The problem is that offense is more consistent than defense, year-to-year, and thus, I would rather have a tip-top O and an average D than the other way around. More chance of a dynasty and multiple rings, I would think.

One aside, about Baltimore, I wonder how their 2000 offensive DVOA from just the last 5-6 games and the playoffs that year would look? Remember, they suffered probably the worst offensive drought I've ever seen in the middle of their SB season. Their kicker was the only guy to score for something like 4-5 weeks. They then destroyed the Titans (DVOA's darlings) twice in short order and annihilated the over-matched Giants. I'm just wondering. I don't really like the Ravens, I hate their "offensive mastermind" of a coach and am sorry to have to sort of say something nice about them.

-Sean