Writers of Pro Football Prospectus 2008

23 Feb 2014

All Estimated DVOA Ratings, 1950-2013

You may remember Andreas Shepard's article with estimated DVOA ratings going back to 1950. Over on Chase Stuart's Football Perspective site, they've put together a couple of tables with all of the estimated ratings for all 1,638 teams. Note these are all estimated ratings, rather than a mix of estimated and actual ratings like we ran here at Football Outsiders; therefore the 1962 Packers are on top, not the 1991 Redskins.

Posted by: Aaron Schatz on 23 Feb 2014

11 comments, Last at 25 Feb 2014, 3:18pm by tuluse


by RickD :: Mon, 02/24/2014 - 1:15am

The Hogs would have rolled that tiny 1960s NFL defense.

Seriously, I don't know why there's any pretense of comparing teams across eras.

by tuluse :: Mon, 02/24/2014 - 2:16am

They're being compared to their peers, which is the only way you can compare these things. So the 62 Packers were really good compared to other pro football teams of the same year, as were the 91 Skins.

Moreover, this is a "just for fun" type of research, don't take it too seriously.

by Guido Merkens :: Mon, 02/24/2014 - 9:36am

I'm not sure if this applies to the "Estimated DVOA" methodology, but DVOA is normalized year-by-year, so there isn't any pretense of comparing teams to teams from a different era. This is just a question of which team was the most dominant within its own era.

by RickD :: Mon, 02/24/2014 - 5:13pm

You really think there's no such pretense here? Should I dig through all of FO's DVOA columns for phrases like "historically good" and "historically bad"?

by Andreas Shepard :: Mon, 02/24/2014 - 9:04pm

Yep, estimated DVOA is also normalized year-by-year. So when I say "the 1962 Packers have a better estimated DVOA than the 2013 Seahawks", I really mean "the 1962 Packers stood out more relative to the league in 1962 than the Seahawks stood out relative to the league in 2013". It does not mean "If you took the time machine back to 1962 and had Bart Starr and co. play against the modern Seahawks, they would be favored".

by nat :: Mon, 02/24/2014 - 12:05pm

DVOA tries to compare teams on how well they dominate their year with its own technologies, talent pool, rules, etc. That's why it uses a league average to normalize everything.

Personally, I think it fails at even that. Expansion means that a 1970 league average is 13th place in a league with all the best talent concentrated in 26 teams, while a 2013 league average is 16th place in a league with all the best talent spread out over six more teams. That's three places lower in a league with diluted talent.

There is a way to estimate the effect of each expansion. The other era-effects such as new technologies, rules, and schemes are and should be part of what we mean by adjusting for eras. But that's a longer discussion.

by Karl Cuba :: Mon, 02/24/2014 - 11:11am

Nicely done Mr Shepard.

by tomdrees :: Mon, 02/24/2014 - 1:29pm

In case anybody's wondering, I plugged the numbers into an excel file and it turns out the Dan Snyder Era Redskins average out, roughly, to watching the 1985 Seattle Seahawks over and over again. Sounds about right.

by Vincent Verhei :: Mon, 02/24/2014 - 3:17pm

This is an amazing and wonderful post for a ton of different reasons. Kudos!

by Jim C. :: Tue, 02/25/2014 - 2:41pm

Am I reading this thing right that the 1974 Redskins were the 25th best team of all time by estimated DVOA? This was the team that had Deacon Jones handling place-kicks at the end of the season.

by tuluse :: Tue, 02/25/2014 - 3:18pm

This is really cool, I made a stacked bar chart of the Bears throughout history. I believe in 65 years, the Bears have only fielded teams above average in all 3 phases of the game 4 times: 1965, 1984, 1985, and 1988.