Adam Steele is back for another guest post, with today’s work being initially published at The GridFe. You can view all of Adam’s posts here. As always, we thank him for contributing.

---

Yesterday, I outlined my methodology for measuring QB defensive support, and looked at the best and worst seasons using that metric. Today’s post will explore the career numbers for the 107 quarterbacks in my study.

I originally planned to simply tally the seasonal numbers without adjustment, but that presented a problem: The top of the support leaderboard was disproportionately filled with QB’s who were exceptional at avoiding interceptions. This makes sense because throwing picks generally makes it harder to prevent the opponent from scoring, although I expected the effect to be minimal enough that it wouldn’t really make a difference. Well, it does make a difference.

To counter the interception issue, I added three point of defensive support for every marginal INT thrown in the regular season. Marginal INT’s are calculated by taking the difference between a given QB’s INT% and the league average rate ^[1]Technically a rolling three year average calculated by Bryan Frye, then multiplying that difference by the number of pass attempts. I know from expected points models that the typical INT is worth roughly -3 EPA, so each marginal INT makes the defense look three points worse than it actually is. As an example, Len Dawson threw -31 marginal INT’s during his career, which “saved” 93 points for his defense. Dawson’s career support is then adjusted by -93 points.

A more rigorous study would also adjust for playoff interceptions and lost fumbles, but I don’t currently have the means for easily compiling the marginal version of those statistics. Again you’ll just have to use common sense and make mental adjustments when necessary.

Results

By deafult, the table below is sorted by career defensive support, but you can also sort by support per game, adjusted points allowed per game, and Conventional Wisdom score. The latter is included to give perspective on how the defensive support a QB receives affects his public perception.

Much like the single season lists, the top of the career list is populated by quarterbacks who won a lot of rings. Otto Graham famously reached the championship game in all 10 of his professional seasons, but the accomplishment is less impressive knowing that Graham was essentially gifted a touchdown head start in every game of his career by his incredible defenses. Steelers greats Ben Roethlisberger and Terry Bradshaw owe a significant amount of their jewelry to the ferocious defenses they shared a locker room with.

Perhaps the most telling example of great defenses boosting a quarterback’s legacy is the reputation of Joe Flacco. Despite career passing statistics that fall below league average, Flacco is generally viewed as a top 50 QB of all time. While his 2012 playoff run and Super Bowl MVP are the shining beacons for his legacy, the foundation for Flacco’s reputation as a winner was built over the previous four seasons. Flacco posted a 44-20 record over his first four years and made the playoffs each season despite posting a middling ANY/A+ of 102. That’s because he was carried to wins by the dominant Baltimore defense, gifting him +422 points of support from 2008-2011. With all those victories in the bank, Flacco was dubbed a winner, a clutch QB, and better than Matt Ryan. But he wasn’t a “winner”, he just had a great defense behind him.

In order to fully grasp the relationship between defensive support and QB reputation, I think it helps to see the data presented in a visually intuitive manner. Below is a chart plotting career defensive support against Conventional Wisdom score:

The correlation between the two variables is a pretty strong .42, which indicates that a QB’s reputation is significantly affected by the strength of his defenses. The guys on the left were dealt a major disadvantage by having to overcome poor defenses, and you can see the effect by virtue of how low most of their CW scores are. A few elite QB’s made it to Canton despite bad defenses holding them back, but Fouts, Brees, Tarkenton, Jurgensen, Tittle, Namath, Moon, and Marino have only two championships on their resumes combined. Meanwhile, the likes of Graham, Brady, Bradshaw, and Starr have 21 total rings on their fingers. It’s well within reason to think that any foursome from the former group could also win 21 championships if they had the defensive support enjoyed by the latter group.

Speaking of foursomes, the same hypothesis applies to the Big 4 of the current era. Tom Brady has more rings and SB appearances than Peyton Manning, Drew Brees, and Aaron Rodgers combined. But it’s not because Brady is a better QB than the other three (he’s about even in my opinion), it’s because he’s had consistently good defenses his entire career while the others have most certainly not. In the chart you can see Brady all alone in the top right corner with +622 support, while Manning, Rodgers, and especially Brees fall on the negative side of the ledger. Imagine if Brees had an extra 1,118 points in his pocket (the difference between his support and Brady’s)…don’t you think he might have a few extra rings to show off?