Yesterday, I looked at the 34 quarterbacks to throw 1,000 passes from 2011 to 2017. In that group, there were four quarterbacks who stood out with low interception rates and high sack rates. All four are black, which is pretty unlikely to happen by random chance given that there were just six black quarterbacks in the study.

But that was a small sample. Today, we expand the group to look at all quarterbacks with at least 1,000 career pass attempts who were active in 2002 or later. For each quarterback, I looked at their INT% and sack% in each season, and measured those rates relative to league average. I have plotted the career grades on the graph below.

On the X-Axis is interception rate relative to league average; it spans from -1.5% (which means 1.5% worse than league average: i.e., really bad) to +1.5% (really good). Tom Brady and Aaron Rodgers are the two at just over +1.0%, while Tommy Maddox is at the far left at -1.26% (he has the worst era-adjusted interception rate since the merger). On the Y-Axis is sack rate relative to league average: at the very bottom, of course, is David Carr, at -4.7%; at the very top are Peyton Manning (+3.3%) and … old Bears QB Jim Miller (+3.0%). ^[1]Miller played from ’95 to ’02, so he just makes it into the study. To be clear, if a player was active in ’02, his entire career was included, which is just another reason why we … Continue reading

So for purposes of this chart, you want to be up and to the right, which indicates better than average sack rates and better than average interception rates. There are 22 black quarterbacks who threw 1,000 passes. I have plotted them in red, while all other data points are in blue. Take a look:

It’s hard to make any definitive statements from that graph above; in some ways it looks like a random dispersion, but it’s also true that just over half (12) of the 22 red dots are in the bottom right quadrant, which means high sack rate/low interception rate. However, here’s one noteworthy takeaway:

• 12 of the 21 quarterbacks in the bottom right are black (57%)

• 4 of the 20 quarterbacks in the bottom left are black (20%)

• 4 of the 21 quarterbacks in the top left are black (19%)

• 2 of the 28 quarterbacks in the top right are black (7%)

So while the data doesn’t tell a clear story, the splits above are pretty extreme. Perhaps we can learn more if we look at the 10 quarterbacks who were not in the bottom right?

Top Right (Low sack rate, Low interception rate): Byron Leftwich and Steve McNair are the only two players of this group, which is of course the best place to be. McNair was an excellent quarterback who almost always had a good interception rate.  Early in his career, his sack rate was about average, but beginning at age 29, he became really good at avoiding sacks, too (he also significantly cut down on his running, due to injuries, beginning at age 30).  Leftwich was always immobile and was famous for his sloooooow release: given that combination, his above-average sack rate and interception rate spoke more to his ability to read defenses more than anything else. My hunch is people who remember Leftwich the player but not his stats will be surprised to see him here

Top Left (Low sack rate, High interception rate): This is the gun-slinger group, who are the opposites of the Taylor/Kaepernick/Griffin/Wilson group we identified yesterday.  So who are they? Kordell Stewart, Jameis Winston, Josh Freeman, and Vince Young.  Winston and Young fit the brash, confident, Favrian gunslinger mold: both were top-3 draft picks following outstanding careers as superstars at national programs where they won a championship.  Being confident gunslingers made sense.  Freeman and Stewart are harder to explain.  Freeman had the makings of being a terrific young quarterback: he had above-average ANY/A seasons at the ages of 22 and 24, and his poor interception rate probably spoke more to playing on bad teams than anything else.  Stewart was not a very accurate passer, so his poor INT rate isn’t a surprise.  To me, his low sack rate sure is, especially relative to how often he ran.

Bottom Left (High sack rate, high interception rate): This is the bad category, although remember we are only look at two metrics here. The quartet: Daunte Culpepper, Rodney Peete, Tarvaris Jackson, and Cam Newton.  An interesting group, with maybe not too much in common.  Newton is the standout here, but even during his magical 2015 season, his sack rate was still below-average and his interception rate was nothing special.  Newton runs so much and provides so much value as a runner that his sack rate will never be great, and he throws deep too much (and with below-average receivers) that he’s going to have a high interception rate, too.  Remember that Newton leads all active quarterbacks with a 12.5 yards/completion average, which offsets the poor sack and interception rate numbers (and also is, I presume, correlated with them).   Culpepper was similar physically, and was also a much better quarterback before his ACL injury; he had a 3.1% interception rate and 7.6% sack rate from ’99 to ’04, and 4.0% and 11.1% rates from ’05 through ’09.  He also had a great yards/attempt average, which offset the bad rates here. Jackson was not a very good quarterback, while Peete also had a strong (enough) yards per attempt average to offset his sack and INT rate issues.

What Does It Mean?

I looked at 90 quarterbacks who played in the NFL since 2002 and had at least 1,000 career pass attempts. There were 22 black quarterbacks and 76 white quarterbacks, ^[2]The other two are Mark Sanchez and Marcus Mariota. yet black quarterbacks dominate the extremes when it comes to low interception rate and high sack rate. Combining the two metrics, 7 of the 9 quarterbacks with the most extreme high sack rate/low INT rates were black quarterbacks: the other two were David Carr (boosted entirely by his historically bad sack rate, part of him playing on the expansion Texans) and Steve Beuerlein, who was also neutral on INTs but horrific on sacks.

If you use the PFR passing index numbers, and look at quarterbacks with INT+ ratings of 107+ and Sack+ ratings of 93 or lower, there are 9 quarterbacks who meet those thresholds since 1988. Six of them are black.

And on the other end of the spectrum, of the 15 QBs who identify most as low sack/high INT guys, just one — Freeman — is black.

So what does it all mean? Can we conclude anything for sure based on the data? The sample sizes are small, but the results are too extreme for their not to be something going on here.

I don’t presume to know exactly what the data tells us, or what is the best way to measure some of these things (or other things I’m not even thinking about). So let’s open source this, so you can have the raw data and run your own studies. The full data set below: in the “INT% vs. LgAvg” column, a positive number means a QB was good at avoiding interceptions. In the “Sk% vs. LgAvg” column, a positive number means a QB was good at avoiding sacks. The same goes for INT Z-Score and Sack Z-Score. The final column — the “Net” column — is INT score minus Sack Z-Score, which means being good at avoiding INTs and bad at avoiding sacks will give you a high grade here (which is why Taylor, Kaepernick, and Griffin are in the top four).

So what is the takeaway here (other than Peyton Manning being the least black quarterback ever)? As Lisk wrote years ago, black quarterbacks do seem to have better than average interception rates. This may in fact be purposeful and part of a strategy to avoid scrutiny and/or getting benched. The high sack rate is even more extreme than the low interception rate, which may reflect this (also, sacks are simply more common events).

So is there something here in the data? My hunch is yes, or else I wouldn’t have published this study. And we can look no further than someone not in the study: Ravens rookie Lamar Jackson. At some point in the next two years, he’ll be given a chance to lead the Baltimore offense. He was the 32nd pick in the draft and a controversial one at that: some teams reportedly wanted him to convert to wide receiver. In other words, he won’t have the security of a Winston or Newton. And my suspicion is that Jackson will end up with a better than average interception rate, and a worse than average sack rate.