Writers of Pro Football Prospectus 2008

10 Dec 2014

Scramble for the Ball: Song of the South(s)

by Tom Gower and Mike Kurtz

Tom: Some divisions are better than others. Those that stand out this year are the AFC North, which appears to be quite strong in a manner reminiscent of the 1989 AFC Central, and the NFC South, which will not see a team finish with a winning record.

Mike: It is so bad.

Tom: Before we start this discussion, I am going to run quick numbers.

Mike: OK. Do we want to go off the same numbers or should I start mathing at things?

Tom: I'm doing division records, so if you want separate mathing, math yourself out.

A BRIEF DELAY ENSUES, in which Tom and Mike engage in separate mathing in the privacy of their own homes. Enter TOM, stage left.

Tom: Very interesting.

Mike: I did average DVOA. The AFC South is significantly worse than the NFC South, with an average DVOA of -15.7% vs. -12.7%. The AFC East is slightly better than the AFC North, 7.1% vs. 7.9%:


Division Average DVOA
ACFW 6.7%
AFCN 7.1%
AFCS -15.7%
AFCE 7.9%
NFCW 6.2%
NFCN 3.4%
NFCS -12.7%
NFCE -5.3%

Tom: We talk about divisional strength, but it may actually be kind of hard to compare teams in different divisions. Teams play six of their 16 games against teams in their own division. Unless you're an SEC fan, those games are not evidence of how good your division is relative to other divisions. Teams play four games each against another division in the same conference, and another four games in a division in the other conference. That does give us pretty good data to compare those divisions, but we're left with just two other games per team to directly compare other divisions.

Mike: We have this problem every time we try to compare divisions, however. And we have yet to find a way to control for the separate conference scheduling, at least, one that doesn't take literally minutes to calculate.

Tom: I'm outlining the basic problem for people who haven't thought about this before.

Take the AFC North. Is it actually good, or just better than the NFC South? They're 11-2-1 in interconference play and just 11-9 against the rest of the AFC:


Division
Conference
Record
Interconference
Record
AFCE 11-11 7-7
AFCN 11-9 11-2-1
AFCS 8-13 4-11
AFCW 12-9 7-8
NFCE 7-16 11-4
NFCN 15-7 7-7
NFCS 6-16 2-11-1
NFCW 17-6 8-7

Mike: To be fair, the AFC North has performed horrifically against the NFC South. Granted, the wins are there, but just look at Tampa's games. Yes, Baltimore destroyed them. But the Buccaneers beat the Steelers, outplayed Cleveland in a loss and lost to the Bengals on a last-minute miracle drive.

Tom: That's who Tampa is, though. They're mostly close except when they get blown out.

Mike: That's why I think DVOA is more instructive for these purposes than record.

Tom: Oh, you don't have to tell me DVOA is better than record. If the NFL didn't actually rank teams by record and make important decisions off it, I might never look at records.

Mike: Right, but for specifically this purpose, because it's adjusted league-wide rather than who played whom. Which brings us to the curious case of the AFC South.

That other southern division has gotten most of the media's attention for generalized terribleness. Yet according to our statistics, the AFC South is significantly worse. The Titans have even taken over Tampa's spot at the bottom of the rankings this week.

Tom: The AFC South stands out in our rankings because it has two really bad teams instead of just one. That goes back to something I've struggled with since before I was on staff, how to best aggregate DVOA rankings to judge division strength. Likelihood of a good team winning games against division? Likelihood of an average team winning games? Likelihood of a bad team winning games?

Mike: I'm not sure what you would do other than average that would provide any more accuracy. Having two horrible teams is indeed worse than having one horrible team.

Tom: So because Carolina is better than Tennessee, and the other three teams in each division have rough analogs in the other one, the NFC South is better than the AFC South?

Mike: That sounds perfectly reasonable to me

Tom: I get the logic behind it, but that idea has always bothered me. It sounds like you want to create something akin to the BaseRuns of awfulness.

Tom: It bothers me more in the context of, which is better, a division with two good teams and two bad ones, or one with four average ones? It reminds me of the difference between the strength of schedule calculation in DVOA and that in FEI.

Mike: What, that DVOA's dartboard is fancier?

Tom: In college, we don't care about how good the 64th-best team is, so Brian Fremeau calculates strength of schedule in FEI based on how a top team would do against a particular schedule. I'm not trying to contest the basic point, which is that the AFC South and the NFC South both suck.

Mike: Well, that is true that we don't care how good the 64th-best team is. On the other hand, there is something approaching parity in the NFL, which makes the kind of generic calculation we get from a raw average of divisional DVOAs pretty useful.

Tom: I should probably just get over myself and accept that because the third-worst team is really bad instead of kind of bad, the division as a whole is worse.

Mike: You are the only football fan on earth that needs to be cajoled into admitting that his divisional rivals are bad.

Tom: I know they're bad! I know my team is ridiculously awful and has now reached their rightful 32nd place in DVOA. I'm just trying to think about how bad we think they are.

Mike: Fair enough. What is interesting is that the AFC East is now actually, legitimately, good. Everyone has fallen back into the comfortable assumption that it is again New England and the three stooges, but Buffalo and Miami have some seriously good defenses. Miami even has something resembling an offense!

Tom: By record, the AFC East is only .500 against the rest of the AFC and also .500 against the NFC (North). The Patriots, two legitimate teams, and a bottom-feeder.

Mike: In any case, it's interesting to see how DVOA averages back up most of my perceptions of the league, but throws a few curveballs. I don't think any of these divisions will be near historic highs or lows, but who knows? The South has three more weeks to prove me wrong!

Tom: To me, the most interesting record information is the NFC East, which is 11-4 against the AFC South and 7-16 against the NFC North 3-12 against the NFC West. Given the DVOA rankings of each division, though, that's pretty much what we should expect.

I can't let the NFC South discussion go, though, without bringing up everybody else's favorite topic: hosting a playoff game. Is it somehow wrong, or should it not happen that the NFC South winner will not have a winning record and will host on wild card weekend?

Mike: I was bitterly disappointed that Tampa Bay was mathematically eliminated from the playoffs this past week.

Tom: On TB, you and me both. You and me both. I was rooting for the 6-10 Falcons, at 6-0 inside the division and 0-10 outside it, to win. Sadly, they had to go and beat Arizona.

Mike: Well, it is wrong in a kind of cosmic morality sense. It is an unavoidable reality of what is by far the best playoff system for the NFL. Unfortunately. Also, let's not stop with "not having a winning record." I want a 7-9 playoff home team!

Lessons Learned

Mike: I hate fantasy football. I won't elaborate, because nobody cares about your fantasy team, but seriously. Fantasy football. The worst.

Tom: Our playoffs start this week. I may be saying that next week.

Hmm. I could come up with a meditation about how surprisingly fun it is to watch your football team when you've fully acknowledged they're terrible and don't even have any hope for them. I kind of ran out of anger and indignation in the Eagles game, which has made watching them lose badly to the Texans and Giants not too unpleasant.

The two most interesting results of the weekend came from the two most... competitive divisions. Division co-leader New Orleans got rocked at home by Tampa Bay Carolina. Division leader Cincinnati lost by three scores at home to Pittsburgh. Apparently, parity is the name of the game, I guess.

By DVOA Cincinnati had the edge through three quarters, then fell apart in the fourth. The Bengals had a 34.5% offensive DVOA through three quarters. It was -72.7% in the fourth. Their defensive DVOA in the first three quarters was -0.2%. It was 97.5% in the fourth. That's one very bad quarter, and with it Cincinnati went from a potential near-stranglehold, up a game-and-a-half with three games to play in the season, to just a half-game lead and a very tenuous grasp of the playoffs. NFL games are 60 minutes, and teams are behooved to play well for all 60 of them, not just 45.

Loser League Update

Full results for the current week and the season to date are available on the Loser League results page. Each week, Scramble for the ball highlights the top performers at each position.

Quarterback: Late-round rookie quarterback, banged-up offensive line, bad banged-up receiving corps, bad running backs. Zach Mettenberger doing anything positive would have been a surprise. Getting 2 Loser League points and finishing last in DYAR was not a surprise.

Running backs: Where have you gone, RG3, Alfred Morris turns his lonely eyes to you. (Woo woo woo.) What's that you say, Mr. Gruden, the passing game has left and gone away. (Hey hey hey.) Just 1 point for Alf (not to be confused with Gordon Shumway). Runners-up with 2 were Mettenberger's teammate Bishop Sankey (who has fewer runs of 15-plus yards than Trent Richardson) and Reggie Bush, since Loser League is non-PPR.

Wide receivers: Keenan Allen got to hang out on Revis Island on Sunday, and your Scramble writer actually forgot he did catch two passes to avoid the penalty. Matching his 0 were Danny Amendola and Cecil Shorts.

Kicker: Greg Zuerlein has never missed an extra point before Sunday. Nor had he ever missed from inside 30 yards. He did both against St. Louis Washington, and missed a 38-yarder after missing the 28-yarder to finish at -5 despite making a field goal and an extra point.

Awards!

Keep Chopping Wood: The Arizona Cardinals were always likely to give Kansas City's offense, particularly its passing offense, fits on Sunday. Alex Smith, playing because he can manage the game and avoid making big mistakes, did neither of those at crucial times. On third-and-20 in field-goal range with his team up five points, he threw a bad interception right to the Cardinals' Alex Okafor. Okafor's return gave Arizona excellent field position and led to the game-winning score. Given a chance to redeem himself at the end of the game, needing to drive into field-goal range from his own 24 with 69 seconds to play, Smith delivered a shambles of a two-minute performance, featuring time-consuming short pass after time-consuming short pass and burning most of that clock just to get to his own 36.

Mike Martz Award: When you are down two scores against one of the AFC's top teams and facing a fourth-and-reasonable in the middle of the fourth quarter, do not punt the ball. Doug Marrone did so on fourth-and-2 with 5:33 to play down 14 against the Broncos from his own 16. The Bills got the ball back, converted a fourth-and-16 from even worse field position, and narrowed the gap to one score before running out of time. Then Mike McCoy topped him, punting on fourth-and-4 from midfield with 6:28 to play down 9 against the Patriots. Like Buffalo, the Chargers ended up facing another fourth down, in even worse field position (fourth-and-5, their own 25), with less time left, and needing to convert there to extend the game. Just go for it in the first place.

Lock of the Week

Tom: We are both now 4-8 on the year. I put my trust in Kansas City. They played it close in Arizona but had a touchdown negated by offensive pass interference and turned it over twice in scoring territory. Arizona came away with the covering win. For the sake of 49ers fans, the less said about Jim Harbaugh's team's performance against his possible future employer, the better.

As always, odds are courtesy of Pinnacle Sports and were accurate as of timing of writing. All picks are made without reference to the FO Premium picks.

Mike: I'm going to say right now that I'm not actually trying to jinx anything. Nor am I projecting my horrible bitterness over Atlanta's role in my embarrassing defenestration in my league's quarterfinals. But Pittsburgh has basically no pass defense. It's almost comical at this point.

Atlanta's isn't much better, but Matt Ryan is able to spread it around a bit better than Ben Roethlisberger has in recent weeks, and I think that's a big advantage in a game of poor safety play. I don't see Pittsburgh's inconsistent offense getting it done on Atlanta's turf. Atlanta Falcons +3 vs. Pittsburgh Steelers.

Tom: Darn, I have to take a different game. That line really stood out, as Weighted DVOA suggests a very even contest and perhaps even Atlanta as a slight favorite. Instead, I'm going to go local and examine a contest between a pair of teams coming off losses. The Bears were down early, but made a game of it against Dallas without Brandon Marshall before falling short. They now get to face the second-worst defense in the NFC South, and in the NFL.

Mike: You are literally the only person I've talked to, media or otherwise, that thought the Bears made a game of that contest.

Tom: It was fake close more than actually close. The Bears showed continuing signs of life after going down 35-7 in the third quarter. That was a lot more than the Saints did against Carolina. I've seen games I thought were less competitive described as "close" and "games Team X could have won."

Perhaps we've run into the Bizarro Saints, losers of four in a row at home and winners of two in a row on the road. But I'm trusting in a long-term view of who the Saints are, a struggling defense that will give up big plays and an offense that isn't as crisp or as efficient away from the friendly confines of home. The long-term forecast for Monday night suggests temperatures in the 30s and a good chance of rain. I know, it takes an awful lot to trust the Bears defense (or their offense, at this point), but I'm betting against the Saints more than favoring Chicago. Chicago Bears +3 vs. New Orleans Saints.

Posted by: Mike Kurtz and Tom Gower on 10 Dec 2014

13 comments, Last at 12 Dec 2014, 2:43pm by Tom Gower

Comments

1
by steveNC :: Wed, 12/10/2014 - 10:19pm

Interesting stuff. Not sure the NFC East can be 7-16 against the NFC North, though.

2
by techvet :: Thu, 12/11/2014 - 12:30am

Get the copy editor - stat!

#1. New Orleans was rocked by Carolina this past Sunday, not Tampa Bay.

#2. Was Greg the Leg actually making field goals against his own team? Was he traded at halftime to the Redskins?

3
by Vincent Verhei :: Thu, 12/11/2014 - 4:08am

All errors fixed.

4
by Revenge :: Thu, 12/11/2014 - 9:27am

The NFC S gets more negative attention because there's no "good" team in it. Fans (and media) give the AFC S more slack because they know that at least it's playoff representative is an actual "playoff caliber" team. The difference between the Panthers and Jaguars is "bad" vs "very bad". The difference between the Colts and the Falcons is the difference between "belongs in the playoffs" and "it's a sham that they're in the playoffs". The latter is a much more compelling case to me.

5
by nuk :: Thu, 12/11/2014 - 2:58pm

"The AFC South is significantly worse than the NFC South"?
What do you think "significantly" means? To me, it means it's worse in a meaningful way, or passes some statistical test. I don't think it qualifies in either of those ways. I respect DVOA, but it's not *that* good that it can distinguish between -13% and -16% with confidence.

If you look at PFR's SRS, the average of the AFCS is a little better than the NFCS, but both systems are really saying the same thing - it's too close to call.

6
by RickD :: Thu, 12/11/2014 - 3:29pm

"Unless you're an SEC fan, those games are not evidence of how good your division is relative to other divisions. "

Sweet.

7
by RickD :: Thu, 12/11/2014 - 3:34pm

"So because Carolina is better than Tennessee, and the other three teams in each division have rough analogs in the other one, the NFC South is better than the AFC South?"

The NFC South doesn't have any analog to the Colts. Both the Falcons and Saints are far behind the Colts in both DVOA and W-L.

11
by Mr Shush :: Fri, 12/12/2014 - 1:44pm

I know DVOA disagrees, but I'm pretty sure the Texans would and should be favoured on a neutral field against any NFC South team.

The Colts, meanwhile, are just plain actually good.

8
by MJK :: Thu, 12/11/2014 - 8:18pm

" I'm not sure what you would do other than average that would provide any more accuracy."

How about median? It's often a more accurate way than averaging. Of course, it's not perfect either... It will tell you that a division with a good team, two mediocre teams, and a bad team is the same as a division with three mediocre teams and a terrible team...

9
by RickD :: Fri, 12/12/2014 - 11:45am

Using the median of a data set with only four points is really asking for trouble.

10
by Kevin from Philly :: Fri, 12/12/2014 - 1:33pm

The average of four points isn't much better. If you define a better division as one that the winner can be expected to go far in the playoffs, I'd say a division with one really good team and three crap piles would be better than a division with four equally mediocre teams, even if their average ratings are the same.

12
by dbostedo :: Fri, 12/12/2014 - 2:24pm

There's the real problem... how do you define "a better division"? I wouldn't use your definition because it's only reliant on having one great team.

I'd say that a division where all four teams are mediocre-to-good (teams who's "true" value is, say 7-10 wins) is better than one with a great team, two mediocre ones, and an awful team.

13
by Tom Gower :: Fri, 12/12/2014 - 2:43pm

The thing we didn't mention is the NFC South stands out because it was reasonable to have expectations for them. The Panthers were the #2 seed, the Saints seemed like one of the best teams in the NFC, Atlanta was in the NFC championship game in 2012, and Tampa wasn't bad for a 4-12 team and got rid of the toxic Greg Schiano.

The AFC South, on the other hand, is going mostly according to form-Indianapolis the only good team, Houston with a reasonable record but below average, Jacksonville bad. The only "surprise" is Tennessee is really bad instead of mediocre.