DVOA Analysis
Football Outsiders' revolutionary metrics that break down every single play of the NFL season

DVOA Z-Scores

Baltimore Ravens QB Lamar Jackson
Photo: USA Today Sports Images

Guest column by Matt Bittle

Who had a better regular season: the 1988 Bengals, who posted a DVOA of 27.5%, or the 2004 Patriots, who had a DVOA of 34.2%? Obviously, the 2004 Patriots, right? Well, not necessarily. If we're asking which team stood out more in that season, the answer is the 1988 Bengals, because they had a superior z-score.

For those unfamiliar, z-scores, also known as standard scores, measure how many standard deviations an item is from the mean. To put it another way, they indicate how widely dispersed a set of values (in this case, team DVOAs in a specific season) are. Applied to football, it essentially reflects the fact different seasons are just that -- different, not identical. Quality varies from season to season; just compare the extreme individual team records of 1976 to those of 1979. Put simply, it's easier to excel when there are more bad teams.

DVOA may be normalized, but a DVOA of 20.0% in one year is not necessarily equal to a DVOA of 20.0% in another year. Take the 2016 Patriots, who led the league with a DVOA of 24.9%. They would have ranked sixth in 2009 with that DVOA. Their z-score, however, based on the standard deviation in 2016, comes in at 1.71. In 2009, that same DVOA would translate to a z-score of 1.13.

Continuing with this example, the 2016 Patriots are tied for 96th since 1985 (the first year that DVOA is available) in DVOA, but 35th by z-score. The hypothetical 2016 Patriots and their 24.9% DVOA transplanted to 2009 would come in at 160th in a ranking of z-scores.

The 1988 Bengals, meanwhile, are 27th all-time in z-score but tied for 65th in DVOA. Their z-score in 2016, by the way, would be 1.88, which would be 18th among all teams! The 1988 Bengals are not remembered as a great team, but that may be mostly because they didn't win the Super Bowl.

So, back to our initial example. The 1988 Bengals may have had a lower DVOA than the 2004 Patriots, but their z-score is higher, 1.78 compared to 1.69. That's not a huge difference, of course, and it certainly doesn't mean Cincinnati was definitely superior; it's just a new way of looking at things.


Best/Worst Teams by DVOA Z-Scores

Here's a list of the top 25 teams by DVOA z-score with rank, as well as actual DVOA and rank. We've also averaged the ranks in both categories for each team.

Best DVOA Z-Scores, 1985-2019

Year Team Record Z-score Rank DVOA Rank Avg. Rank
1991 WAS 14-2 2.75 1 56.9% 1 1.0
1987 SF 10-2 2.64 2 47.2% 4 3.0
2007 NE 16-0 2.60 3 52.9% 2 2.5
1985 CHI 15-1 2.56 4 52.5% 3 3.5
1995 SF 11-5 2.39 5 40.0% 8 6.5
2010 NE 14-2 2.38 6 44.6% 5 5.5
1996 GB 13-3 2.25 7 42.0% 6 6.5
2019 BAL 14-2 2.24 8 41.5% 7 7.5
2013 SEA 13-3 2.21 9 40.0% 8 8.5
2015 SEA 10-6 2.14 10 38.1% 11 10.5
2018 KC 12-4 2.05 11 32.9% 21 16.0
1994 DAL 12-4 2.04 12 32.9% 21 16.5
1989 SF 14-2 2.02 13 36.0% 14 13.5
1995 DAL 12-4 1.95 14 32.7% 23 18.5
2012 SEA 11-5 1.93 15 38.7% 10 12.5
2014 SEA 12-4 1.90 16 31.9% 27 21.5
2010 PIT 12-4 1.89 17 35.4% 15 16.0
2004 PIT 15-1 1.86 18 37.6% 12 15.0
1998 DEN 14-2 1.86 19 32.6% 25 22.0
1994 PIT 12-4 1.85 20 29.8% 40 30.0
2003 KC 13-3 1.84 21 30.3% 36 28.5
1992 DAL 13-3 1.84 22 35.1% 16 19.0
2012 DEN 13-3 1.82 23 36.5% 13 18.0
1999 STL 13-3 1.82 24 34.0% 19 21.5
2013 DEN 13-3 1.81 25 32.7% 23 24.0

The lists are a little different, but there really is not too much movement here. The 1991 Redskins are first in both categories despite the fact that 1991 had the third-highest standard deviation for any season (more on that later). The 1987 49ers (strikebreaker games not included) and 2007 Patriots each claim second in one of the two categories. The 1995 Cowboys and 2014 Seahawks look a bit better by z-score, as do the 1998 Broncos. Notably, the 1994 Steelers shoot up 20 spots, while the 2003 Chiefs jump 15. This ranking also makes it even wilder that the 1995 49ers won only 11 games.

The 2018 Chiefs and 2000 Titans (who rank 20th in DVOA at 33.3%) offer another great illustration of why z-scores are useful: the teams have very similar DVOAs, but the NFL was more balanced, in a sense, in 2018. 2000 had its share of awful teams, and so the Titans, while slightly higher in DVOA than the Chiefs, are 27 spots lower by z-score.

Going farther down the list a bit, we find some real outliers. To name a few: The 2001 Rams are 87th in DVOA but 44th in z-score, and the 1988 Vikings are 96th and 49th, respectively. The 2005 Colts are 26th in DVOA but 67th in z-score, while the 2005 Broncos come in tied at 34th in DVOA but rank 80th in z-score.

Notably, 2005 saw the 49ers post a DVOA of -55.5%, the worst ever. Well, I'm pleased to report they are also worst by z-score. Consider them the inverse of the 1991 Washington team. Here are the worst DVOA z-scores:

Worst DVOA Z-Scores, 1985-2019

Year Team Record Z-score Rank DVOA Rank Average Rank
2005 SF 4-12 -2.62 1 -55.5% 1 1.0
2003 ARI 4-12 -2.56 2 -42.0% 9 5.5
1987 ATL 2-10 -2.55 3 -45.7% 6 4.5
2018 ARI 3-13 -2.53 4 -40.7% 12 8.0
1986 TB 2-14 -2.52 5 -43.8% 8 6.5
2008 DET 0-16 -2.42 6 -48.4% 3 4.5
2008 STL 2-14 -2.35 7 -47.1% 5 6.0
2009 DET 2-14 -2.34 8 -51.6% 2 5.0
1991 IND 1-15 -2.30 9 -47.7% 4 6.5
2002 HOU 4-12 -2.22 10 -41.9% 10 10.0
2016 NYJ 5-11 -2.22 11 -32.4% 44 27.5
1999 NO 3-13 -2.15 12 -40.3% 13 12.5
1993 IND 4-12 -2.15 13 -37.2% 24 18.5
1999 CLE 2-14 -2.12 14 -39.7% 18 16.0
2011 STL 2-14 -2.11 16 -35.4% 32 24.0
2013 JAX 4-12 -2.11 15 -38.2% 21 18.0
1992 NE 2-14 -2.10 17 -40.1% 16 16.5
2016 CLE 1-15 -2.08 18 -30.4% 51 34.5
2004 SF 2-14 -2.07 19 -41.8% 11 15.0
2009 STL 1-15 -2.05 20 -45.1% 7 13.5
1989 DAL 1-15 -2.03 21 -36.2% 28 24.5
2000 CLE 3-13 -2.03 22 -40.2% 14 18.0
1994 HOIL 2-14 -2.00 23 -32.3% 45 34.0
2012 KC 2-14 -2.00 24 -40.1% 16 20.0
2010 ARI 5-11 -1.98 25 -37.1% 25 25.0

By z-scores, the 2009 Lions fall (jump?) from second to eighth. The 2003 Cardinals and 1987 Falcons come in at second and third in z-score (the 2003 Cardinals are slightly worse if you go out one more decimal place). Those two teams are ninth and sixth, respectively, in a ranking of the worst teams ever by straight DVOA.

The 2009 Rams are 20th in z-score but seventh in terms of actual DVOA. The inclusion of the 2016 Jets also jumps out, as they're just the 22nd-worst team ever by DVOA but come in at 11th by z-score.


Year-by-Year Standard Deviations

I have mentioned 2016 and 2009 several times. Well, there's a good reason I picked those two seasons: no two seasons since 1985 were more extreme than 2009 and 2016. Teams were more "spread out" in 2009 than in any other year. That year, 14 teams posted DVOAs of 10.0% or greater, with six exceeding 20.0%. In 2016, there were just seven teams with a DVOA in the double digits, including two of 20.0% or more. On the flip side, 2009 saw nine teams with DVOAs of -10.0% or worse, including the Rams at -45.1% and the Lions at -51.6% (there really was a tremendous race to the bottom between St. Louis and Detroit in 2008 and 2009). In 2016, although seven teams had a DVOA of -10.0% or worse, the bottom-feeders -- those Jets at -32.4% -- would have ranked just fourth worst in 2009.

2009 and 2016 are both pretty clearly the most extreme years in our sample. The standard deviation for 2016 was 14.6%, the lowest on record; 1988 is in second place at 15.4%. The standard deviation for 2009 was 22.0%, the highest on record; 2005 is in second place at 21.2%.

NFL Seasons by Standard
Deviation of DVOA, 1985-2019

Year Standard
Deviation
2009 22.013
2005 21.219
1991 20.725
1985 20.501
2007 20.381
2004 20.189
2008 20.015
2012 20.005
2000 19.813
1990 19.567
1992 19.053
2002 18.838
2010 18.728
1999 18.705
1996 18.662
2019 18.534
2013 18.101
1987 17.887
1989 17.790
2015 17.788
1997 17.734
1998 17.566
1986 17.397
1993 17.266
2006 17.072
2017 17.035
2011 16.783
2014 16.775
1995 16.771
2003 16.433
1994 16.110
2018 16.076
2001 15.781
1988 15.442
2016 14.593

There's no obvious factor causing 2009 and 2016 to be so extreme; it's just seasonal variance -- again, compare 1976 to 1979. And speaking of those two "pre-DVOA" seasons…


Best/Worst Teams by Estimated DVOA Z-Scores

I've also calculated the standard deviation for all teams post-merger using estimated DVOA for the 15 seasons where DVOA is not available. Using z-scores for these years produces some interesting results. I'm not sure how much of that is due to these DVOA ratings being estimated and not based on actual play-by-play and how much stems from other factors, like how top-heavy the league was for much of the 1970s.

There's a big divergence there; the 1979 Steelers have the best z-score for estimated DVOA (by far, at 2.49) but rank ninth in DVOA. The 1978 Cowboys are second by z-score but 17th by DVOA. The 1976 Steelers, who have the highest DVOA of all teams in this timespan at 42.7%, are only ninth in z-score. (1976 was a weird year -- probably the most imbalanced of all post-merger seasons.)

Here are the top 25 teams from 1970 to 1984 by z-score.

Best Estimated DVOA Z-Scores, 1970-1984

Year Team Record Z-score Rank DVOA Rank Average Rank
1979 PIT 12-4 2.49 1 33.4% 9 5.0
1978 DAL 12-4 2.21 2 29.7% 17 9.5
1983 WAS 14-2 2.21 3 29.4% 18 10.5
1971 DAL 11-3 2.18 4 36.3% 4 4.0
1983 SF 10-6 2.07 5 27.6% 20 12.5
1977 DAL 12-2 1.98 6 35.1% 6 6.0
1974 WAS 10-4 1.96 7 36.0% 5 6.0
1980 PHI 12-4 1.91 8 25.8% 25 16.5
1976 PIT 10-4 1.84 9 42.7% 1 5.0
1984 SF 15-1 1.81 10 32.5% 10 10.0
1973 MIA 12-2 1.79 11 38.4% 2 6.5
1982 NYJ 6-3 1.77 12 25.2% 27 19.5
1972 MIA 14-0 1.76 13 34.5% 8 10.5
1973 LAR 12-2 1.71 14 36.7% 3 8.5
1975 PIT 12-2 1.70 15 35.0% 7 11.0
1979 SD 12-4 1.67 16 22.4% 32 24.0
1970 DET 10-4 1.64 17 30.2% 15 16.0
1974 PIT 10-3-1 1.64 18 30.1% 16 17.0
1981 PHI 10-6 1.57 19 22.0% 37 28.0
1974 OAK 12-2 1.51 20 27.8% 19 19.5
1975 OAK 11-3 1.48 21 30.5% 13 17.0
1973 DAL 10-4 1.46 23 31.3% 11 17.0
1977 DEN 12-2 1.46 22 26.0% 24 23.0
1973 OAK 9-4-1 1.46 23 31.3% 11 17.0
1977 PIT 9-5 1.45 25 25.8% 25 25.0

The 1979 Steelers are not generally considered the best of those great Steel Curtain teams -- I wager most fans would pick Pittsburgh's 1975 or 1978 squads -- but by this metric, they're not just the best team from that dynasty, they're one of the best ever. Look how big the gap is between them and the second-ranked 1978 Cowboys!

There are some truly great teams on the list above that may not be considered such. The 1978 Cowboys went 12-4 and lost to the 14-2 Steelers in their Super Bowl, while the 1983 49ers were much better than their 10-6 record would suggest and nearly beat a great Washington team in the conference title game. The 1980 Eagles came up short in the Super Bowl, while the 1979 Chargers and 1973 Rams both had outstanding regular seasons and then lost their first playoff game. (Also, seeing the 1983 Redskins ranked third makes me wonder: how many coaches had two different teams as good as Joe Gibbs' best Washington squads?)

The 1976 Steelers are a fun bunch: they went 10-4 despite outscoring their opponents by 14.6 points per game, 2.4 more than anyone else. Pittsburgh started 1-4, with their losses by a combined 19 points. Then, over the ensuing nine games, the Steelers allowed 28 points, total. Pittsburgh did lose in the AFC Championship Game to what I consider a vastly overrated 13-1 Raiders team (6-0 in one-score games). Thus died the dreams of a threepeat, the closest any team has come to winning a third consecutive Super Bowl.


Best/Worst Offenses/Defenses by Z-Score

Anyway, how about the top units? Which offenses and defenses had the top z-scores? The best offenses by z-score remain the 2007 and 2010 Patriots, in that order, but the third-ranked team is a surprise. Take a guess. Seriously, go ahead.

Ready?

It's the 1988 Bengals. An offensive DVOA of 30.7% is very, very good, but it's only 16th all time. 1988, however, happened to be one of those years when the league's teams were grouped more closely together, and so the Bengals jump up the list when considering standard deviation.

Best Offensive DVOA Z-Scores, 1985-2019

Year Team DVOA Z-score
2007 NE 43.5% 2.79
2010 NE 42.2% 2.69
1988 CIN 30.7% 2.62
2003 KC 33.4% 2.56
1993 SF 31.4% 2.51
1990 SF 26.2% 2.36
2002 KC 35.4% 2.32
2013 DEN 33.5% 2.32
1998 DEN 34.5% 2.29
2017 NE 27.3% 2.25

Those are the top 10 offenses by DVOA z-score. Nine of these teams either made it to the conference title game or won at least 13 games. And then there's the 2002 Chiefs, who led the league in scoring and went 8-8.

If you include estimated DVOA, the best offense in the 50 post-merger seasons is the 1982 Chargers. Although they would rank below both of the aforementioned Patriots teams in DVOA, San Diego tops New England by z-score. (Granted, the Chargers only played nine regular-season games.)

The 2002 Texans remain the worst offense in z-score as they are in DVOA.

The title of best defense in the DVOA era by this method stays with the 1991 Eagles. That team isn't thought of as one of the best ever in most circles, but by DVOA, those Eagles were legendary.

The 2002 Buccaneers are second by z-score, which is not surprising, but the third-ranked team once again will probably make you raise your eyebrows. It's the Bears.

No, not that Bears team. Not that one either.

The 2018 Bears. Khalil Mack and company benefit here big-time, shooting up from the 10th spot. Those Bears may not be thought of as an all-time great unit, but by DVOA, the second-ranked defense in 2018 (Buffalo) was closer to the No. 10 Colts than they were to Chicago.

Best Defensive DVOA Z-Scores, 1985-2019

Year Team DVOA Z-score
1991 PHI -42.4% -3.41
2002 TB -31.8% -3.12
2018 CHI -26.0% -2.96
1986 CHI -33.6% -2.92
1988 MIN -26.5% -2.85
1985 CHI -33.6% -2.70
2012 CHI -26.7% -2.68
1997 SF -21.2% -2.52
2013 SEA -25.9% -2.51
1995 SF -23.7% -2.46

Again, note the repeat offenders. For what it's worth (not much), one of those top 10 offenses and three of those top 10 defenses won the Super Bowl.

The 2015 Saints and 1986 Buccaneers are the two worst teams by defensive DVOA and also rank as one and two in z-score. The 1997 Bengals jump from 28th to fourth worst going from straight DVOA to z-scores.

With estimated DVOA added, the 2015 Saints aren't the worst defense in the post-merger era, however. Both the 1976 Seahawks and 1981 Colts come in below the Saints, with the Seahawks having a worse (estimated) DVOA and the Colts having a worse z-score.

While Baltimore is a close second in DVOA, no one else comes anywhere near the Colts in z-score. (Fun fact about the 1981 season: Both the Patriots and Colts went 2-14 despite the fact Baltimore was outscored by 14 MORE POINTS PER GAME than New England. Both the Colts' victories came against the Pats, bookending the season.)

If you have any questions or whatnot, please share. I'm happy to give the z-score for any team in this sample upon request.


Matt Bittle is a writer in Delaware. Prior to COVID-19, he was really hoping the 2020 Ravens would follow in the footsteps of the 1988 49ers.

Comments

22 comments, Last at 10 Aug 2020, 2:39pm

1 Some of this is because DVOA…

Some of this is because DVOA gets rejiggered on a yearly basis, right?

What does the annual DVOA mean look like if you normalized on a decadal basis, instead of annual?

2 Well, At Least Five

“how many coaches had two different teams as good as Joe Gibbs' best Washington squads?)”

Lombardi, Noll, Landry, Brown, and Belichek.
Carroll would seem to deserve a nod for four teams in the top 16 of the DVOA era.

and I am surprised Walsh did not have a second team.

And, of course, John Madden.

13 Another 49ers coach

They may not be quite on that level, but George Seifert had the 1989, 1992, and 1994 49ers (to say nothing of the 14-2 1990 team and the amazing 11-5 1995 team).

Walsh's two best were probably 1984 and 1983/88 depending on how you weight postseason. Don Shula probably qualifies for the 1972 and 1973 Dolphins.

Guys like Tony Dungy and Andy Reid have had a bunch of really good teams (even a few that flirted with greatness) but none on the very highest level.

 

15 Reid's 2019 team did pretty…

Reid's 2019 team did pretty well considering they never got healthy until the playoffs. They spent most of the year down 2 LBs, two linemen, and the baby goat gimping around.

Then it turned out their death star actually was fully functional.

16 None of those teams were as…

None of those teams were as statistically dominant as Gibbs' teams, therefore the answer is none of them from the context we are using here.   We've well documented the 1991 Gibbs team as the best statistically of the DVOA era, and now here is an argument that the 1983 team was the best of the pre-dvoa era (even though they fell apart in the superbowl, which despite its lobsided margin, I still think could have gone the other way if it had started differently.   Apart from that game, the lost two games on MNF by one point each and that's it.

 

 

3 "..the 1983 49ers were much…

"..the 1983 49ers were much better than their 10-6 record would suggest and nearly beat a great Washington team in the conference title game."

They were also a gakked Eddie Murray field goal away from being upset at home in the divisional round by a thoroughly mediocre 9-7 Detroit team.

4 Ahh Detroit, where every…

Ahh Detroit, where every trip to the playoffs involves facing one of the 15 best teams in the last 50 years.

\Had Detroit gotten past #15 SF, they would have had #9 DC
\\In 1991, they were considered fortunate to not get SF and got DC.
\\\Really curious what that Lions' team's DVOA would have been if Washington were not on their schedule

5 If that infamous pass…

If that infamous pass interference flag doesn't get picked up against Dallas in the 2014 WC game, and the Lions held on to their lead, their reward would have been to face.the (#16 by Z-score) the 2014 Seahawks.

6 If there was a title for…

If there was a title for most cursed franchise; would the Lions be the prohibitive favorite? They are certainly among the inner circle of contenders. Throw in the Chargers, Jets, and Bills I guess? Vikings seem to be usually too professional to put on this list. 

Its a fun exercise. Any team can be thoroughly horrible and mismanaged for a period of time, but it takes special skills to be a good team and perpetually flop due to comedies of unplanned errors. 

7 It depends on the curse. The…

It depends on the curse.

The Vikings, and to a lesser extent the Bills, are cursed like Tantalus -- success is always right there, it just always eludes their grasp.

The Chargers are the masters of snatching defeat from the jaws of victory, losing in increasingly unlikely ways despite superior play and talent.

The Lions, and really the Cardinals, are like Satan -- cursed to be ignored by God's favor. Really, I think the Lions made a Faustian bargain, and after the demon Bobby Layne's terms expired, their penalty was being owned by the degenerate wing of the Ford clan, and the whole thing rolled over like a late-90s Explorer. 

8 Ahh, late-90s Fords. My Dad…

Ahh, late-90s Fords. My Dad's Expedition had a part of the suspension (sway bar?) fall off twice while backing out of the driveway. Good thing that didn't happen in traffic - that was the last Ford we ever had!

One of the best testaments to the Lions luck in the modern era is what happened after they drafted 3 star players at #1 & #2 overall in 3/4 seasons from '07-'10 (and they didn't have a high pick in the "off" year of '08). Their reward for a nearly-unprecedented run of success picking at the top of the draft? They got capped out and were unable to keep the core together once they finally started winning, and by that point the NFL had changed the rules so that would never happen to any other team again. The hilarious poison pill they negotiated into Calvin Johnson's rookie contract was probably all on them, though.

9 My first car was a 96 Ford…

My first car was a 96 Ford Sable. Granted, I didn't take the greatest care of it(I got it when it had 70k miles on it). When I was in high school, the glove box locking mechanism broke and I had to use a stick to prop it shut, forcing whoever sat in the passenger seat to place a leg on either side.

The rest of the interior, which wasn't in great shape to begin with, continued its rapid decline as the car got older. It was like watching the eagles injury report last year. 

The coup de grace was the transmission, which blew out at 92k miles while I was driving down the santa Cruz Mountains. Needless to say, I have not bought a Ford since. 

10 "Put simply, it's easier to…

"Put simply, it's easier to excel when there are more bad teams."

I'm having trouble seeing how this stat tells you more about a teams quality than DVOA already does. I don't think it's pointing out when a high DVOA is just due to "more bad teams"; Because having more bad teams in the league doesn't necessarily translate into a lower mean or more ease of being x standard deviations above that mean, right?

You could have a lot of bad teams, but a bunch of good ones too... have a fairly high mean, and have a smaller z-score for the top teams, even though there are a bunch of bad teams. Maybe this would lead to depressed DVOA scores because of the opponent-adjustment already built in? 

You could also have very few bad teams but a big grouping of mediocre teams.... have a fairly high mean, and have a smaller z-score for the top teams, in a different way than having a lot of bad teams. Again, would this show up already in their DVOA as higher DVOA because of the opponent-adjustment?

What does the mean DVOA and standard deviation look like season-over-season? (A graph would be awesome.) And how does that correlate with z-scores? What actually makes it harder to get a bigger z-score? 

11 I agree with dbostedo

DVOA supposedly measures the amount by which a team exceeds the average team in a season, so the 2004 Patriots were more impressive than the Bengals of, well, any year. If the z-score analysis is adding value then... Football Outsiders has some explaining to do. :-)

 

What the z-scores really tell us: the seasons with the widest spread had the most impressive champions. The point is that playoffs keep only the good teams, so the 2004 Patriots beat a bunch of very good teams to win the title. Those Bengals? Well, they didn't have to beat anyone particularly strong while coming up short. (But season-ending DVOA should account for playoffs, too, so we should not need z-scores to tell us this, either.)

Anyway... my impression is the opposite of what the author says.

14 Author here

Maybe I'm misunderstanding you, but for me part of it comes back to how much things can vary year to year. Were five teams in 2009 really better than the 2016 Patriots in the regular season? I mean, I guess rank tells you that somewhat, but it doesn't indicate the separation.

I am curious now how the results change if I group multiple years together.

18 1976 Steelers and season stuff

As I'm sure most know, that Steelers squad (not just 28 pts in 9 games, but 5 shutouts, and over half those points came in one game) were wiithout Franco Harris in that AFC championship game vs the Raiders.   He had rushed for 132  yards vs the Colts but was injured, and without him they had to rely on Reggie Harrison and John Fuqua (Rocky Bleier was also hurt).  They managed 72 yards rushing a week after having 225 total (I'm sure the Raiders run defense was better than the Colts, but still, Harris was a huge part of their offense.    

The Raiders had actually had a chance to end the Steelers' run for a 3 peat earlier that year,  at 11-1 and having clinched their division, they faced a 9-3 Bengals squad that was clinging to the AFC Central lead, having just lost to the hard charging Steelers 7-3, and were in their midst of their epic run looking like world beaters, but still trailing due to that 1-4 start, being 9-4 after that sunday, with this game being on Monday Night Football.    If the Bengals won, they would be 10-3 and could clinch the division over the lowly Jets, who were 3-10 at the time (and who they would indeed go on to annihilate 42-3).   At the time, there was only one wild card team, and the Patriots and Colts were both also 10-3, with games against the 2 win Bills and winless Bucs remaining.   

So... if the Raiders were to simply lose to the Bengals... say rest their starters...  the Bengals would likely finish 11-3 and win the AFC central, and the way the tiebreakers lined up, the Colts would win the AFC East (they had tiebraker vs Patriots, having split season series but with better net pts, plus the Patriots had an additional division loss), and then the Patriots would win any tiebreaker over the Steelers, having beaten them 30-27 during their 1-4 start.  This would leave the Steelers out of the playoffs.

The Raiders, of course did not do this, they played their starters and won 35-20.   The Steelers finished winning out and won a tiebreaker over the Bengals due to having swept them and thus the AFC Central champs, while the Bengals missed the playoffs at 10-4.

In looking back over this, the NFL schedule makers were not what they are today,  that 13-1 Raiders squad opened at home vs the Steelers (an epic come from behind 31-28 win), then played the next five weeks on the road.  Can you imagine the outcry over such a schedule were it to happen today?    They then got to play at home six of the final 8 weeks of the season.    Was there maybe some kind of stadium renovation going on?    Need to talk to quirky research, see if there are any more unblanced home/away splits on schedules....

 

19 The Raiders also opened 1974…

The Raiders also opened 1974 with 6 of their first 8 on the road.  They generally played a lot of road games early in that era because they shared a stadium with a baseball team that everyone expected to play into October, and also the league at the time was more concerned about good weather for late season games, especially on Monday night.

20 Unbalanced Schedules

The Oakland A's made the playoffs 1971-1975, so they were just trying to keep the Coliseum open for baseball.

If you want truly unbalanced schedules, look at any Jets schedule between 1967-1976, when they were not allowed to play any games at Shea Stadium until the Mets' season had ended.  In 1973, the Jets played their first 6 games on the road (the 6th was moved to Pittsburgh after the Mets made the World Series, leaving the Jets with an unbalanced number of home/road games).

22 Was the shorter length of…

Was the shorter length of the 1987 season (12 games vs 16) taken into account? I'm guessing that if we looked at the standings of every season after Week 12 or 13, there would be a lot more teams with |DVOA| > 40%. The presence of two 1987 teams in the top & bottom 3 is perhaps telling.

I can't wait for the 1982 numbers!