Posted by Jason Lisk on February 12, 2010
If you haven't been keeping up with the comments to the previous post about amazing stats and context, well, shame on you, because there has been a lot of interesting discussion about a lot of things. One of the things that came up was a discussion of player talent with some references to Joe Namath. I'm going to just quote some comments from BSK, responding to JWL:
Are you we thinking of the same Joe Namath? He of the 173/220 TD/INT ratio? Of the 62-63-4 record as a starter? Of the career completion percentage of 50.1%? Of the career 65.5 passer rating? And while perhaps Namath would have had a better career despite his injury, the fact is the injury happened, his career numbers were pretty poor, and by no legitimate statistical measure can you say he was a legit HoF. Take away the story of Super Bowl III and his personality and it wouldn't even be up for discussion. And that is the problem with the HoF.
And later . . .
Now, what happened on the field obviously went better for Namath than (Bo) Jackson, but neither really did anything particularly exceptional when it's all said and done with. My point was that there are supremely talented players who, for one reason or another, don't live up to that talent (or the perception of their talent). Unfortunately, we cannot give them credit for what they did not actually accomplish. Namath gets no bonus points for what might have been had he not hurt his leg before he even got to the NFL, just like Jackson gets no bonus points for what might have been. To say that Namath was a top 5 QB and then say that my statistical demonstration of exactly why he wasn't is off the mark is laughable at best.
If you read the title of this post, you might correctly guess that I am going to try to show that by legitimate statistical measures, Namath was a legitimate Hall of Famer. While I am quoting BSK here, because he just happens to be the one making comments in a post this week, I don't think he is exactly in the minority. I see lots of comments about Joe Namath not being a Hall of Fame caliber quarterback, or talking about how bad his numbers were. Joe Posnanski wrote about Namath and his "shockingly bad" numbers on his blog two years ago.
I guess I should first point out that Chase Stuart wrote a series of posts on the Greatest Quarterbacks of All-Time last summer, and in the most recent version, Namath ranked #24 all-time. Now, Chase is a Jets fan, so perhaps you think that Chase, just because he sponsors Namath's player page at PFR, was cooking the books to make Namath look better. Actually, we had a lot of discussion behind the scenes about that series. Chase was contemplating including a completion percentage calculation as part of the updated formula, and I am actually the one who deterred him by showing him some numbers about teams with similar YPA's and different completion percentages, and the resulting win/loss and points scored. I should probably do a separate post on that this off-season, so I'll just say that for now, it didn't appear that including completion percentage would actually better measure value. Namath, as we know, had a relatively low completion percentage, so including that would have lowered him in the rankings.
So, we see that in what I would hope would pass as a legitimate statistical measure, Joe Namath ranks as a valid Hall of Famer, even without "the guarantee" and the New York media. So let's break down that a little further and discuss why some think he is not, and why I think he is.
1. When we cite things like quarterback rating, completion percentage, and interception ratio, we are going to find that they do not favor Namath. Of course, quarterback rating is over-reliant on completion percentage, and interception percentage also plays a big factor, so mentioning those things and also citing qb rating is redundant.
2. As we know, qb rating does not include sack percentage, though I argued a few months ago that it should. This also disfavors Namath when we cite qb rating, because he had a quick release, which is statistically confirmed by his extremely low sack percentage relative to his era.
3. I talked about quarterback personality types this summer and one of the traits I used was the Gambler trait. If you threw more interceptions and more incompletions because you were avoiding sacks, you were a Gambler in my book, and Namath was an extreme Gambler. Of course, this isn't necessarily bad for your point production and value, even though it is bad for your blessed qb rating. I actually wrote a modest proposal for a Kansas City area sports blog entitled Matt Cassel needs to throw MORE interceptions, where I discuss some of these things. Holding the ball and taking sacks can be as costly or more costly than throwing some interceptions by throwing the ball before you are ready. In Namath's case, we are underselling how good he was when we don't also cite his sack data. His effective completion percentage (completions divided by total passes plus sacks) ranks him much better, and in my opinion, more accurately provides a full picture of a quarterback's contribution.
4. Completion percentage is vastly over-rated. Again, I will probably have a separate post sometime this off-season. I also looked at quarterbacks with similar passer ratings, but different sub-ratings in the four categories, and you will probably be interested in the results as they relate to how frequently, say, a qb with a 90 rating that is dropped down by a bad completion percentage wins and scores, compared to one that is propped up by a good one.
5. Most people agree that yards per attempt is a better indicator of passing value, and Namath exceeded 8.0 yards per attempt in 1967 and 1968, and was at 7.0 or higher every year between ages 23 and 32. Using our advanced passing table which adjusts to league average, he was above average in that category in every one of those seasons. He was insanely above average in 1972 (over two standard deviations above the league average).
6. When we look at adjusted net yards per attempt, which does include his sack rate and his interception rate (but does not include completion percentage), we see a well above average quarterback for most of his career. We don't have reliable sack data for individual quarterbacks before 1969, but extrapolating his career sack rate after 1969 (combined with his completion percentage and interception rate) we can make a pretty good guess that he was also good at avoiding sacks before 1969. From 1969 forward, which would be after he won the Super Bowl and most think he stopped being a good quarterback, he was over a standard deviation better than the league in 1969, 1971 and 1972. He was above average in 1973 and 1974 as well. The only year he was average was in 1970, when he played in only 5 games. So, using adjusted net yards per attempt, rather than quarterback rating, we see that he was a well above average quarterback from ages 24 to 31.
7. His numbers need to be put in context of his era, which we can do with things like our Advanced Passing Table, as interception rates and sack rates and completion percentage were all much lower than they are today.
Of course, all of the above is why he ranks in the top 30 all-time on Chase's list, but I just wanted to spell out why that was, and why I disagree with assessments that selectively cite things like qb rating and completion percentage. I don't know how talented he was relative to other quarterbacks. I do know that he ranks in the top 30 by what I think is a pretty good objective measure, which takes into account rate stats and total attempts to derive value.
And he is ranked in the top 30 despite missing a substantial portion of what would be the prime years for a lot of quarterbacks (missing 28 games between ages 27 and 30). The one year he played almost a full season during that stretch (1972), he led the league in passing yards, touchdowns, yards per attempt, adjusted yards per attempt, net yards per attempt, and adjusted net yards per attempt. Oh, and he completed 50% of his passes, so he sucked. The Jets went 7-6 when he started that year, but it was because they ranked 19th out of 26 teams in points allowed, and not because they finished 2nd in points scored.
And he is ranked in the top 30 despite hanging around too long and playing broken down and on bad knees, and putting up awful numbers at the end of his career. We don't know what he would have been if he had stayed healthy (though it's not going out on a limb to say he would rank higher), but let's be clear. His career numbers were not "pretty poor", unless you worship at the Church of the Blessed Quarterback Rating, and ignore everything else.