Comparing Manning and Unitas

It is bad form to try and catch up a friend mid-conversation.

I’m going to do it anyway.

Over at the Football Outsiders, a discussion has been raging in the comments section of the Mike Tanier piece on Peyton Manning and John Unitas. It is notoriously difficult to compare quarterbacks across eras. Football has changed and evolved considerably since the 1950s, and we can’t simply take today’s stats and proclaim every quarterback in the league a legend just because they surpassed the performance of a Hall of Famer who played 50 years ago.

That’s why the complex field of era adjustments is necessary.

At this point, I would encourage you to go back, read the discussion that has taken place, and rejoin me here. Don’t worry. I’ll wait.

 

By now, you know that the issue centers around what constitutes a valid comparison. Unitas played in a smaller league than Peyton Manning does. One reader strongly believes that most era adjustments that have been done by sites like ProFootball Reference have introduced ‘grade inflation’ into the equation. The idea is that there are more QBs playing now than then, therefore a valid comparison between Manning and Unitas should compare Unitas to the league average for his day and Manning to only the average of the top 16 QBs of his day.

 

Obviously, I disagree.

To resolve the debate, however, I’ve run the numbers to compare each quarterback to the league top five of his day.  That is to say I did the following:

From 1956-1973, I took the stats for each of the top five quarterbacks not named Unitas in the NFL in the following categories:

Completion Percentage

Touchdown Percentage

Interception Percentage

Yards Per Attempt

Passer Rating

I then ran those numbers to produce a mythical ‘Average Top 5 Quarterback from 1956-1973′ in each category. Note for the record, that for years 1966-1968 I included AFL quarterbacks as well.

Then I went and did the same process from 1998-2010. I took the stats for the top five QBs not named Peyton Manning, and built a profile. Now we can compare Manning and Unitas against a mythical elite performer in each category.

Here are the results:

Comp % TD % INT % YPA Rating
Top 5 1956-1973 56.7% 7.0% 3.8% 8.3 84.2
Unitas Career 54.6% 5.6% 4.9% 7.8 78.2
% of Top 5 96.3% 80% 128.9% 93.97% 92.9%
Top 5 1998-2010 65.5% 6.2% 1.7% 8.1 99.2
Manning Career 64.9% 5.5% 2.7% 7.6 94.9
% of Top 5 99.1% 88.7% 158.8% 93.8% 95.7%

As you can see, if you only compare Unitas and Manning to the top five quarterbacks of their day, Manning fares better than Unitas. Manning leads in three categories and is essentially tied in YPA.  When factoring in Manning’s massive advantages in terms of longevity and volume, there’s no question of his greatness.

There’s a lot that’s fascinating about that chart. When you look at the difference in how the top 5 in the league performed in each era, you can get a sense of how difficult it is to compare players from different eras.  This is why I favor a ‘tier system’ for quarterback evaluation rather than a ranking system.

I don’t know that this chart settles the Manning/Unitas debate, but I hope it sheds some light on it.

Quantcast