clock menu more-arrow no yes mobile

Filed under:

Quantitative Analysis of the Greatest QBs in NFL History

WolfPackSteelersFan continues to impress with his exhaustively thorough research and ideas for interpreting the statistical greatness of players by position in different eras. I think one would be hard pressed to find all this nifty and informative data consolidated in one place on the internet. BehindtheSteelCurtain is blessed to have him in its community. Click through for some oustanding data compiling, objective interpretation, and accessible writing. Enjoy. -Blitz-


So far I've analyzed the top RBs in NFL History here (Part 1) and here (Part 2), and I've looked at the top WRs in NFL History. So, I have decided to turn my attention to the top QBs in NFL history. 

For each of these analyses, I have gotten my raw data from's leader boards. If you take a look at that site, you will see that there are many more statistics for the QBs than for RBs or WRs. So, I had a lot more data to work with and more players to analyze. As with the WRs analysis, I started by pulling the data for players in the top 10 of every statistical category, which gave me a total of 63 QBs. But then, once I had done that, I saw that there were players that clearly should not be included in the analysis. Take, for example, Jim Miller, who played from 1995 to 2002. He played a total of 8 years, but only appeared in 37 of a possible 128 games, starting 27 of them. So, I decided I had to weed out some players.

I filtered out any QB who played since 1980 with fewer than 2000 yards per season. I figured that if they played in the modern era and had production that low, they likely spent a significant amount of time as a backup. This eliminated players such as Rich Gannon, Mark Rypien and Neal O'Donnell, all of whom started Super Bowls (Rypien winning his). Despite the SB experience, each were backups for as much as half of their careers. It also eliminated some promising new starters: Philip Rivers, Tony Romo, and David Garrard. In these cases, they have not been starters long enough to establish themselves as one of the historical greats. Overall, the weeding out process eliminated 13 of players.

At that point, I began looking at the data and realized that it would be necessary to break this analysis up into two parts: Modern Era (ME) and Old School (OS). Since the passing games were so different between these two eras, it seemed to me that comparing players from the 40s, 50s and 60s with those from the 80s and 90s would not really be an accurate comparison. The older players had much lower completion percentages and yards/season or game, but also much higher yards/attempt than the Modern Era QBs. So I broke up the list into players that played before 1978 and since 1978. This seemed to be the logical break point since that is the year that major rules changes began to aid the passing game. However, I decided to include Johnny Unitas and Fran Tarkenton in the Modern Era analysis because, of those QBs that played their entire career before the rule changes, they were the only two in the top 11 in career passing yards.

As with the previous analyses, I have tried to look at these players in terms of what they did in a per game and per attempt basis. But, because of the fact that there were so many other statistics available, I also included yards per season and career completion percentage in the analysis. I thought these were important to include because QBs are so often judged in a given year based on these statistics. Below are the categories that I used to analyze these players:

  • Yards/Season
  • Completions/Game
  • Completion Percentage
  • Yards/Game
  • Touchdowns/Game
  • Touchdown Percentage (TDs/Attempt x 100)
  • Yards/Attempt
  • Yards/Completion
  • Passer Rating
  • Net Yards/Attempt (Incorporates Sack Yards lost) - Not used for OS
  • Sack Percentage (Sacks/Attempt x 100) - Not used for OS
  • Sacks/Game - Not used for OS
  • Interceptions/Game
  • Interception Percentage (Interceptions/Attempt x 100)

As noted, I did not use the sack statistics to analyze the Old School players, because they did not keep sack statistics during the careers of most of those players. Also, as with the Fumbles/Game in previous analyses, I ranked those with fewer sacks and interceptions higher.

Enough introduction; let's get into the analysis. First, we will review the Old School players, then the Modern Era.

The table below shows the Old School players, with their statistics.


Click to Enlarge

As with the previous analyses, I sorted these players and ranked them in each category. The results are shown in the table below.


Click to Enlarge

And, here is the table showing the aggregate ranking of players that were analyzed. The rankings were calculated by taking summing the rankings of each player and dividing by the number of categories (in this analysis, that number is 11).


What do you know? Another old Cleveland Brown at the top of the list! I guess that will make kwoop and bereadawg happy. Man, they were loaded their first 20 years! Otto Graham is in the top 6 in every category except Interceptions/Game. Fran Tarkenton is ranked number 1 in the most categories (5), but he's hurt by his relatively low ranking in Touchdown Percentage, Yards/Attempt, Yards/Completion, and Interceptions/Game (all 17 or lower). I expected Unitas and Tarkenton to dominate this analysis, but I was surprised. Given that Otto Graham also won several championships, I guess it's safe to say that he was the greatest quarterback of his era.

Now that we've looked at the Old School, let's take a look at the data for Modern Era QBs. This will, of course follow the same pattern as the Old School players. Here is the statistical data for the Modern Era players.


Click to Enlarge

The Modern Era analysis contained the sack statistics that I mentioned were left out of the Old School analysis. That game me 14 categories, which made the table too long to be able to read, even by clicking it to enlarge the image. So, I have broken it down into two images, which are shown below.

Modernqbsortedrank1_medium  Modernqbsortedrank2_medium
Click Either to Enlarge

And, again, the table with the aggregate ranking of the Modern Era Quarterbacks is shown below.


For the Modern Era analysis, Peyton Manning is the pretty clear number 1. He in the top 3 in 9 of the 14 categories, top 5 in 11 of the 14 categories, and he's in the top half of 2 of the other 3 categories. His lowest ranking is 14, for yards/completion. By comparison, Johnny Unitas is ranked number 1 in more categories (4), but is in the bottom 5 in half of the categories.  Dan Marino, ranked second in this analysis, is in the top 5 in 6 of the 14 categories, and he is also in the top half of 12 of the categories. Carson Palmer is ranked in the top 5 in 5 of the 14 categories and the top half in 11 of the categories.

After reviewing these two analyses, Otto Graham and Peyton Manning actually have a greater gap in overall ranking than Jim Brown and his next closest competitor, Ladainian Tomlinson. Since Graham's career, and those of his era, are all over, we can say more surely that he was the greatest QB of his time. Peyton Manning is still playing, and therefore may drop in production and hurt his rankings. But, up to this point in his career, he has clearly been the most productive QB of the Modern Era, at least in the regular season.

Again, here are a few points of interest that I wanted to highlight before wrapping up this post:

  • Our own Big Ben ranked number 1 in two categories, and was in the top 10 in 8 of the categories. Unfortunately, as we know, he's been sacked a lot, and his ranking in the sack categories reflect that. But, overall, as a player entering the prime of his career, he stacks up favorably with some of the historic greats. His overall average ranking puts him at number 9 in the Modern Era analysis, with an average ranking of 12.21. As we've discussed here before, he looks like he could be a future HOFer.
  • Just a few highlights to differentiate the game before 1978 and after:
    • None of the OS QBs had more than 15 completions/game, most less than 10.
    • Only 4 in the ME analysis had fewer than 15 completions/game. Tarkenton and Unitas, I included just to see how they measured up. The other two were Dave Krieg, who started many years, but also was a back for a few years, and Jim Hart, who probably should have been in the old school analysis, since he played 13 of his 19 years before 1978.
    • Most of OS QBs had a completion percentage under 55 (all but 8).
    • Over half of the ME QBs had a completion percentage over 60.
    • No OS QB averaged more than 200 yards/game, most less than 150.
    • Over 3/4 of the ME QBs averaged more than 200 yards/game, a handful over 250.
    • All but 3 OS QBs average more than 13 yards/completion, more than half average more than 14 yard/completion.
    • All but 3 ME QBs average less than 13 yards/completion.
    • Most OS QBs had passer ratings below 80.
    • All but 6 ME QBs had passer ratings above 80, with 7 of them above 90.
    • Nearly half of OS average less than 1 TD/game, but nearly half also had a TD percentage of 6 or greater.
    • All of the ME QBs average more than 1 TD/game, but none have a TD percentage as high as 6.
    • So, the game went from few passes, usually deep bombs, to a lot of passes, generally shorter and higher percentage, which most of us already knew, I guess.

In some ways, this analysis is also an analysis of the transition that the NFL made from a primarily running league to a primarily passing league. As I already said, I really kind of expected Johnny Unitas and Fran Tarkenton to dominate the Old School analysis. Mainly, this was because they were so high on the all time passing yardage list. However, considering how well they did in both analyses, it's safe to say that both were certainly great QBs. They also played during a type of transition period, and it's likely that their passing and the excitement that it could create is why the league wanted to become more of a passing league. However, of the one QB that dominated either analysis, Otto Graham has by far the most championships. Unitas won 3, Tarkenton 0, and Manning 1, so far. Depending on the source, Otto Graham has won 7 or 8 football championships, and I believe he had a little help from a guy named Marion Motley. So, then, just as now, an overall balanced team is more likely to win a championship than the team with Superman at QB.

Well, I think that's it for this one. What do you guys think? Any other data that should have been included?