RSS Feed
21 Nov 2008
Worst Division Ever?
Jim Baker wonders whether any of this year's divisions are the worst in NFL history.
Most Recent FO Features
» Audibles at the Line: AFC Wild Card Games
In the AFC Wild Card edition of Audibles: The Chargers eke out an overtime squeaker thanks to a pinball and a punter, while the Ravens assert their superiority and expose the holes on Miami's roster.
Re: Worst Division Ever?
Okay--when you have a "worst ever" on the other end you frequently have a "best ever." Or should I be asking this question to the pro-football-reference guys?
I'm not for sure, but I'd bet that the AFCE, NFCE, and NFCS all have a chance to produce 4 teams with at least .500 records. (Didn't the NFCE do it last year too?) The Eagles tie even gives the NFCE have all teams OVER .500--a remarkable achievement.
Re: Worst Division Ever?
Last year both the AFC South and NFC East had all four teams at .500 or better. Each sent three teams to the playoffs, with Houston and Philly missing the cut. Cumulatively, the AFC South as a division finished two games up on the NFC East.
Re: Worst Division Ever?
I'm always fascinated by how you define "best" or "worst" division. Assume you have some way of objectively saying exactly how good each team is--be it W/L record, DVOA, etc. What would make a division be the "best"? Highest sum of all involved teams (or, equivalently, the mean "goodness" of the division)? Best median team? Best worst team? Consider four divisions: one with three above average teams and one world beater, one with two really good teams and two average teams, and one with four well-above-average-but-a-little-below-really-good teams, and one with three really good teams and the Oakland Raiders. Which of these divisions is the "best" division?
And it gets harder because teams in a division play each other twice...so if you're using W/L or something that doesn't correct for opponent adjustments, you are overlooking that every measure of goodness for teams in a good division ought to count for more. For example, the Jets, Bills, and Dolphins last year each got two losses almost automatically just by being in the same division as the Patriots.
The engineer in me argues for a mean (equivalent to a sum) based measure, or a median measure. The football fan in me argues for a "best worst team" measure. The statistician in me wants to combine all the measures and define the best division as the division that produces the highest probability that a randomly selected team from that division will beat an "average team" for that year (or perhaps a randomly selected out-of-division opponent). In other words, having at least one really awesome team helps, but having top-to-bottom quality also helps.
I just did a quick calculation based on what Maximum Likliehood Estimation thinks of the teams this season. By my (admittedly rough) rapid calculations I get that the odds of a randomly selected team from each division beating an "average" team is:
AFCE: 54%
AFCN: 52%
AFCS: 58%
AFCW: 35%
NFCE: 65%
NFCN: 42%
NFCS: 61%
NFCW: 34%
So MLE thinks the best division this year is the NFCE, and the worst is the NFCW. I don't know where these rank historically, but it seems pretty impressive that if you pick a random NFCE team and make them play an "average" team, they'll win nearly 2/3 of the time.
It would be interesting to do this kind of calculation using DVOA instead of MLE, since DVOA is probably more accurate at predicting win probabilities...
Re: Worst Division Ever?
Back when Doug did his divisional DVOA article last summer, I did a data dump of the divisions since 1970 sorted by PythWins. Last year's AFC South would check in at 3rd on the list, .01 behind the 04 AFC East, while the NFC West and AFC West fit in on the bottom 10, and the NFC South not far ahead. Overall, though, there doesn't seem to be a hugely strong connection between strong and weak divisions.
Re: Worst Division Ever?
Can we get a DVOA article to look into this? I doubt any of these divisions, except maybe the AFC West, are as bad as some of the NFC Wests of the past that were "dominated" by mediocre Seattle teams, but it looks like there's a solid chance one of the three will set the record for fewest wins by a division this year.
Re: Worst Division Ever?
The first full season I followed football seriously was in 1978, which as a Vikings fan came on the tail end of their run of 10 division titles in 11 years. I recalled it as being one of the weakest divisions I could recall... lemme see if that holds up.
Their standings:
Minnesota: 8-7-1 3-5 out of division
Green Bay: 8-7-1 3-5 out of division
Detroit: 7-9-0 3-5 out of division
Chicago: 7-9-0 4-4 out of division
Tampa Bay: 5-11-0 3-5 out of division
That makes for an out of division record of 16-24. 40%. Huh, remembered them as worse...
Re: Worst Division Ever?
Any discussion of "best" division has to include 1975 AFC Central.
...............Overall......in division
Steelers.......12-2............6-0
Bengals........11-3............3-3
Oilers.........10-4............2-4
Browns.........3-11............1-5
20 losses total (14 game season). 12 within the division, 8 outside. A 75% winning percentage outside the division.
The Bengals lost 2 to the Steelers and one to the Browns. The Oilers lot 2 to the Steelers and 2 to the Bengals. The Steelers lost to the Bills (227 yards for OJ) and to the Jawarski (AKA Jaws) led Rams in the season's final game. Its hard for a division to do better than that.
Re: Worst Division Ever?
Here is the link to Doug's article about the worst divisions of the DVOA era. It confirmed my belief that the 2004 NFC West was by far the worst division I have seen. And in case anyone is wondering if the season ended today the NFC West would have a combined DVOA of -57% making it the 5th worst ever and giving the division the dubious honor of having 4 of the 5 worst seasons ever.
Re: Worst Division Ever?
While I don't disagree that the 2004 NFCW was fraught with suck and may well have been the worst division ever, I do disagree with trying to quantify that with cumulative DVOA--i.e. that having the worst combined DVOA makes a division the "worst". In fact, Doug admits in the article to which you link that this method is oversimplistic, for exactly this reason: You could have a situation where you have three above average teams (say +5% DVOA) sharing a division with an abysmally horrible team (say, -60% DVOA) (tot DVOA: -45%), and I would argue that that division is much better than one where all four teams are somewhat bad, having say -10% DVOA (tot DVOA: -40%), even though cumulative DVOA says otherwise.
I would stick by the metric I suggested before--qualify a division's DVOA by the probability that a randomly selected team from that division would beat an average (say 0 DVOA) team. I know there are some rough probabilistic heuristics about the odds of team A beating team B given their respective DVOA's (it's what FO uses to calculate the playoff odds report, if I recall correctly). Anyone recall what they are?
This would also factor in "diminishing returns". A ROBO-team with a DVOA of, say, +200% would only have a slightly greater chance of beating an average team than a really, really good team with a DVOA of +100%--in both cases, the good team will almost certainly win. But if the ROBO-team is in a division with three -10% DVOA teams and the good team is in a division with three +20% DVOA teams, then the ROBO-division will end up looking stronger, even though I'd far rather play my out-of-conference games against that division than against the one with four above-average teams.
Re: Worst Division Ever?
I would stick by the metric I suggested before--qualify a division's DVOA by the probability that a randomly selected team from that division would beat an average (say 0 DVOA) team. I know there are some rough probabilistic heuristics about the odds of team A beating team B given their respective DVOA's (it's what FO uses to calculate the playoff odds report, if I recall correctly). Anyone recall what they are?
The linked article suggested estimated wins might be a better measurement, and it sounds exactly like what you're talking about. Their standard definition of estimated wins: uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
16 games against a league average schedule should be roughly equal to 16 games against a league average team. So the best way might just be to average the sum of all the teams' estimated wins. (If you're only comparing the 32 team era, you don't even have to average, since each division has exactly 4 teams.)
Re: Worst Division Ever?
1984 AFc West great deivision. Maybe best ever,
Raiders 11-5
Seahawks 12-4
Broncos 13-3
Chiefs 8-8
Charers 7-9
All wild card teams came from AFC west. So 3 out of 5 playoff teams in Afc were from West division
Re: Worst Division Ever?
I love that you listed the Raiders on top of the division as with the third best record.
That is a pretty good division record.
Post new comment