≡ Menu

Ranking the Almost Dynasties, Part II

Andrew Healy is back with a sequel to his popular post. As always, we thank him for his generous contributions. Andrew Healy is an economics professor at Loyola Marymount University. He is a big fan of the New England Patriots and Joe Benigno.


A couple of weeks ago, I went decade-by-decade since the 1970 AFL-NFL merger to identify the teams that were the best of their eras and the teams that nearly became the teams we remember most instead. In those rankings, I used Pro Football Reference’s Simple Rating System to estimate team strength. Today, I use Football Outsiders’ DVOA ratings and go back an additional twenty years. Using DVOA produces some pretty notable differences that were bigger than I would have guessed.

What are some of those changes?

  • The Steelers have been supplanted as the true team of the ‘70s.
  • The best team to win no titles changes for three of the decades.
  • The ‘70s Vikings get replaced by a more recent what-might-have-been team as the best to win nothing in the Super Bowl era.

Before we get to that, I cover the 1950s and 1960s, identifying the true teams of those decades and the what-might-have-been teams. In a follow-up post, I’ll bring it all together and identify the franchises that have maximized their championship potential the most, and those that have left the most money on the table.

Here are the ground rules for the simulations:

  • 1) The playoffs were run under the rules in place in a given year. All of the results are based on 50,000 simulations performed under those rules. As was the case then, it is substantially easier to win multiple championships before 1967 than it is currently, where teams have to win at least three playoff games.
  • 2) DVOA ratings are used for 1989 and estimated DVOA from 1950-1988. I thank Andreas Shepard for sharing his estimated DVOA ratings from before the DVOA era. [1]I also thank Bryan Frye for sharing SRS ratings going back to 1941. Predictions for the 1950s and 1960s using SRS generally look pretty similar to what you get with DVOA. I’m happy to share those … Continue reading I created a betting line based on the DVOA difference between the teams, adding 2.5 points for home field. [2]The standard deviation of DVOA is about three times that of SRS, so I took the DVOA difference and divided by three to create the point spread. That line was used to simulate wins and losses.
  • 3) I adjusted the DVOA estimates to account for AFL/NFL differences for the first four Super Bowls. As others have discussed, any adjustment will be a little ad hoc, but about a seven point penalty seems right. [3]Chase note: For those interested in reading a more quantitative approach to comparing the two leagues, Jason Lisk’s series of articles comes highly recommended. So instead of being a pick ’em against Green Bay, the simulations make Kansas City a 7 point underdog against the Packers were the teams to meet in Super Bowl I. [4]If we look at 1970, the SRS for the former AFL teams averaged about seven points below the average for the NFL teams that year. We could use the actual point spread, of course, but those lines were … Continue reading
  • 4) Pre-Super Bowl championships count, but only for NFL teams. My condolences to the to the ’60 and ’61 Oilers, Lamar Hunt’s 1962 Dallas Texans, the ’63 Chargers, and the ’64 and ’65 Bills.

Let’s begin with a look at the 1950s. The inclusion of the % symbol would inhibit your ability to sort the table, so as a reminder, allow me to work through an example. In the decade of the ’50s, the Browns had a 0.2% chance of winning zero titles, a 2.3% chance of winning one, a 9.2% chance of winning two, and so on. Cleveland even had a 3.8% chance of winning 7 titles, and would be expected to win 4.12 championships.

Team01234567E(Wins)
CLE0.22.39.220.62823.712.13.84.122
RAM13.637.834.412.61.60001.507
NYG12.439.837100.80001.471
DET23.342.226.87.10.70001.195
BAL26.552.521000000.946
CHI48.648.92.5000000.538
SF78.321.70000000.217
PHI10000000000
PIT10000000000
WAS10000000000
GB10000000000
STL10000000000

The True Team of the Decade: Cleveland Browns
Making seven title games in the 1960s, you knew it had to be the Browns, but it’s hard to overstate how dominant they were. Cleveland won the same number of titles as the Lions (3), but that was on the low end for what they might have won. There was about a 40% chance the Browns would have won 5 or more titles in the decade. This jives with Chase’s research that Paul Brown’s teams underachieved in the NFL playoffs, even if such a statement seems weird to see. In any event, the amazing thing to see (even if it’s partially a function of the era) is 4.12 expected titles, more than twice as many as the Browns’ nearest competitor.

The What-Might-Have-Been Team: Cleveland Browns
Four title game losses. In the NFL playoffs, Otto Graham “only” posted a 4-3 record. And during the Jim Brown era, Cleveland lost to an inferior Giants team in the 1958 play-in game, missing the chance to play the Colts for the title. Even controlling for era, Cleveland could have been remembered more by the casual fan had they won the four or five that was within their grasp. If we have to go with a no-title team, the Bears are the best choice. Chicago was slightly better than the Giants in 1956 and lost a division playoff to the great Rams team in 1950.

The Unlikely Winner: Detroit Lions
The Lions also won three titles in the 1950s, despite having just 1.2 expected wins. Then Detroit traded Bobby Layne and the franchise has been too bad to get lucky.

1960s

Team0123456E(Wins)
GB0.66.422.334.425.99.21.23.109
CLT14.339.733.811.11.2001.453
NYG33.945.818.22.10000.884
CLE37.245.3161.50000.817
DAL42.841.913.61.60000.741
OAK46.5458.10.40000.624
CHI42.657.4000000.574
PHI53.546.5000000.465
KC59.834.15.80.30000.465
MIN61.438.20.400000.39
RAM77.820.91.200000.234
NYJ85.714.10.200000.144
BUF92.77.3000000.073
OIL97.52.5000000.025
NO1000000000
CRD1000000000
ATL1000000000
DET1000000000
SF1000000000
WAS1000000000
PIT1000000000

The True Team of the Decade: Green Bay Packers
Utterly dominant in 1961 and 1962, the Packers continued to be very good later in the decade when they won three of their titles. Green Bay had a total of 3.11 expected titles, a number the Pack overachieved due to their relatively fortunate titles in 1965 and 1967 [Chase note: On the other hand, had Paul Hornung not produced the worst kicking season ever in 1964, Green Bay would have had a chance for yet another title.] In 1965, the Packers won a playoff game against a possibly better Colts team in overtime.  As Chase wrote about last December, the Colts were forced to play running back Tom Matte at quarterback due to injuries. The Packers were playing their backup, too, after an early injury to Bart Starr, but at least Zeke Bratkowski was a legitimate quarterback.

The What-Might-Have-Been Team: Baltimore Colts
There’s something a little weird about putting the Colts here, since the team won titles in ’58, ’59, and ’70, but it’s also easy to forget how good Baltimore was for the ten-year stretch when they didn’t win. The team’s 1.45 expected wins are the highest of a team that didn’t win a title in any decade, although that doesn’t seem to impact the legacy of that Colts quarterback. Baltimore had only a 14% chance to emerge from the ’60s without a title, and yet that’s exactly what happened. In addition to the loss to the Jets in Super Bowl III and the aforementioned loss to the Packers in 1965, the Colts also lost to an inferior Browns team in the 1964 title game. [5]Interesting thought experiment: Imagine the Colts won their two expected titles in the 1960s. The debate about the greatest QB would probably be much more clear-cut. Not that it should be, but it … Continue reading

1970s

Team0123456E(Wins)
DAL11.229.632.318.96.51.40.21.849
PIT1533.330.815.44.60.80.11.639
OAK30.539.921.86.51.10.101.082
MIA3142.921.14.60.5001.007
RAM3640.118.64.60.70.100.942
MIN45.638.413.32.40.3000.734
WAS58.934.56.10.40000.481
BAL63.230.75.50.50000.433
SF66.82940.20000.375
DEN73.324.71.900000.286
SD75.524.6000000.245
DET82.817.2000000.172
CIN86.612.80.600000.14
NE86.6130.400000.138
PHI91.48.50.100000.088
GB91.78.3000000.083
KC92.77.3000000.073
STL93.960.100000.062
CHI95.54.5000000.046
CLE96.13.9000000.04
TB96.13.9000000.039
HOU96.83.2000000.032
BUF99.10.9000000.009
ATL99.50.5000000.005
NO1000000000
NYJ1000000000
NYG1000000000

The New True Team of the Decade: Dallas Cowboys
Using estimated DVOA, we get a new true TOD for the 1970s. Dallas edges out Pittsburgh with 1.85 expected wins in the decade. The Cowboys are rewarded for consistent excellence throughout the decade, jumping from 1.38 expected titles when using the Simple Rating System. Pittsburgh still has about the same number of expected titles as with the SRS, now 1.64.

The New What-Might-Have-Been Team: Los Angeles Rams
DVOA likes the Rams more than the SRS did, and it likes the Vikings a fair bit less. The Rams had only a 36% chance of winning the zero titles in the decade that they did.

1980s

Team0123456E(Wins)
SF9.127.132.921.17.81.70.21.973
CHI25.644.823.95.20.5001.101
CLE4639.312.71.90.1000.709
MIA46.638.912.41.90.1000.7
WAS49.840.49.10.70000.608
CIN57.83660.20000.486
DEN60.932.560.50000.461
SD62.832.14.90.20000.425
PHI62.333.24.40.10000.423
OAK64.730.14.90.30000.409
DAL66.529.14.10.20000.381
MIN71.125.63.10.20000.323
BUF7224.930.10000.313
LAR73.723.22.90.20000.295
NYJ77.2211.80.10000.247
NYG77.820.51.700000.239
ATL79.520.20.300000.209
SEA88.6110.400000.117
NE89.410.40.200000.108
PIT90.29.40.300000.101
HOU927.70.200000.082
TB92.27.70.200000.08
NO92.37.7000000.077
BAL/IND94.85.2000000.052
GB964000000.04
DET97.22.8000000.028
STL/PHO99.20.8000000.008
KC99.40.6000000.006

The True Team of the Decade: San Francisco 49ers
The 49ers continue to be the clear true TOD and emerge considerably stronger under DVOA than they did under SRS.

The New What-Might-Have-Been Team: Cleveland Browns
Using DVOA, the Browns edge out the Dolphins as the team from the ’80s with the most expected titles among the group that didn’t win a Super Bowl. I have the Browns now with only a 46% chance of winning no Super Bowls in the 1980s.

1990s

Team0123456E(Wins)
SF18.335.529.1133.50.60.11.497
DAL25.14025.38.11.30.101.209
WAS27.665.46.70.20000.796
PIT4240.514.72.60.2000.786
GB42.743.812.41.10000.72
KC47.939.111.41.50.1000.667
DEN47.240.411.21.10000.665
BUF58.633.37.30.80000.504
RAM/STL61.938.1000000.381
JAX69.7291.300000.316
NYG72.426.41.200000.288
MIN76.422.21.400000.251
OAK7722.20.800000.238
HOU78.520.11.400000.23
MIA79.1191.80.10000.228
NYJ81.518.5000000.185
PHI84.614.70.700000.161
ATL87.312.50.200000.129
SD88.710.90.400000.117
NE90.790.300000.095
TB90.69.30.100000.095
CAR90.79.3000000.093
NO91.97.90.200000.082
CHI94.55.40.100000.055
DET94.85.10.100000.053
IND94.95.1000000.052
TEN95.54.5000000.045
SEA96.63.4000000.034
CLE97.52.5000000.025
CIN99.80.2000000.002
PHO/ARI1000000000
BAL1000000000

The True Team of the Decade: San Francisco 49ers
The Niners stay on top with DVOA, but the gap between them and Dallas gets much closer. Now there’s only a 25% chance that Dallas would have earned zero rings.

The New What-Might-Have-Been Team: Pittsburgh Steelers
This is perhaps the single biggest change from using DVOA rather than the SRS. Now, the Steelers jump over the Bills and have the highest number of expected championships among teams with no titles in the ’90s. Pittsburgh had only a 42% chance of getting shut out in the decade and a 17% chance of winning multiple championships.

2000s

Team0123456E(Wins)
NE22.343.5267.110.101.211
PHI28.239.523.27.51.40.201.15
PIT41.341.814.62.10.1000.78
IND45.939.112.72.10.2000.717
TEN48.540.79.90.80000.632
OAK56.4376.30.30000.505
BAL61.132.16.20.60000.463
TB62.433.14.30.20000.423
SD6332.44.40.20000.419
NYG64.3314.50.20000.407
GB69.126.640.30000.355
CHI68.328.730.10000.348
STL69.228.62.200000.33
SEA70.427.71.800000.315
DAL75.622.320.10000.265
DEN75.223.51.300000.262
KC74.9250.200000.253
NO78.620.31.100000.226
CAR81.917.30.800000.19
NYJ84.314.7100000.167
MIN8514.50.500000.155
MIA87.911.70.400000.124
SF92.870.100000.073
ATL94.55.40.100000.056
WAS94.55.5000000.055
JAX95.74.3000000.044
CIN95.94.1000000.041
ARI97.32.6000000.027
CLE99.20.8000000.008
BUF1000000000
DET1000000000
HOU1000000000

The True Team of the Decade: New England Patriots
The Patriots still come out on top, but with DVOA it becomes a very close call over the Eagles. New England’s expected wins drop just a little in comparison to what we get when using SRS. As difficult as it may be to remember, both teams reached the same number (five) of conference championship games in the ’00s.

The What-Might-Have-Been Team: Philadelphia Eagles
On the other hand, DVOA causes the Eagles to rise and the Colts to drop. In fact, the Eagles now become the only team that won nothing and still mounted a real challenge for the top spot in expected wins in a decade. With DVOA, Philadelphia’s consistent excellence gave them a 72% chance of winning at least one title in the decade, along with a 32% chance of multiple titles. The 2000s Eagles, according to DVOA, are the best team to win no championships in a decade in the post-merger NFL.

Coming soon: I’ll break down the most unlikely winners and losers in NFL history. Going back to 1950, I identify the team that should have won the most titles overall, the teams that most outperformed their expectations, and the Super-Bowl-less team with the lowest chance (only 12%) of being in that sorry state. Each of these might not be the teams you expect. At least two of them were surprises to me.

References

References
1 I also thank Bryan Frye for sharing SRS ratings going back to 1941. Predictions for the 1950s and 1960s using SRS generally look pretty similar to what you get with DVOA. I’m happy to share those results upon request.
2 The standard deviation of DVOA is about three times that of SRS, so I took the DVOA difference and divided by three to create the point spread.
3 Chase note: For those interested in reading a more quantitative approach to comparing the two leagues, Jason Lisk’s series of articles comes highly recommended.
4 If we look at 1970, the SRS for the former AFL teams averaged about seven points below the average for the NFL teams that year. We could use the actual point spread, of course, but those lines were likely biased towards the NFL, although it’s possible that the true spread was somewhere between 7 and 14. Anyway, it doesn’t make much difference if we use a slightly higher adjustment for the first four Super Bowls.
5 Interesting thought experiment: Imagine the Colts won their two expected titles in the 1960s. The debate about the greatest QB would probably be much more clear-cut. Not that it should be, but it probably would be. Still, over the whole Johnny Unitas era, the Colts won almost exactly their expected number of titles.
{ 17 comments }