by Michael Smith (2003 numbers added by Aaron Schatz)

Few certainties exist in the world of football, but this is one of them: You could assemble an offensive line with Anthony Munoz, Forrest Gregg, Joe DeLamielleure, Larry Allen, and Mike Webster, all in their primes, but if you put Rob Johnson behind them, he'd still find a way to get sacked.

Most observers think sacks are a function of the battle at the line of scrimmage.  When the offensive line wins the battle, the quarterback is able to throw his pass; when the defensive line wins, he isn't.

But what about the quarterback? Isn't he equally responsible for getting rid of the ball before the defense reaches him?  Last Monday, Football Outsiders took a look at how often a team took a sack based not on total sacks but on sack rate, the percentage of pass attempts that resulted in sacks.  Well, the best way to determine whether the quarterback or line is responsible is to see if different quarterbacks playing behind the same line have similar sack rates.  So I've looked at the 30 teams in the past three full seasons with more than one quarterback with 100 passes.  And I found that sack rates vary widely even when quarterbacks are playing behind the same line.  My feeling, based on these statistics, is that the quarterback is more important than the line.  And the Lions' quarterback, Joey Harrington, might be the best in history at avoiding sacks.  Then again, looking at his passing numbers, maybe he's just the quickest in history at throwing the ball away.

And then there's Johnson.  If you stayed up late enough for the season's final Monday night game, you saw him get sacked against the Packers. In 2000 he was sacked on 16% of his pass attempts.  Playing behind the same line, Doug Flutie was sacked on 4.3% of his attempts.  But Flutie is known for his smarts and his mobility.  Could it be that Johnson was playing behind a line whose weaknesses were masked by Flutie's skills?  Nope.  The next year, splitting time with Alex Van Pelt (who is known neither for smarts nor mobility), Johnson was sacked at more than three times the rate of his fellow quarterback.

I should acknowledge that these statistics have some weaknesses.  I haven't adjusted for situation or defense faced, and I haven't examined how much the personnel on these offensive lines changed as the season wore on.  Nonetheless, the statistics reveal that when quarterbacks are sacked, they can blame themselves as much as their linemen.

(Aaron's note: After Mike sent me this article, I decided to hold off on publishing it until I could run the 2003 data. The 2003 numbers include both sack rate and adjusted sack rate, which takes into account down, distance, and opponent.)

The following tables include all teams that had two or three quarterbacks with 100 pass attempts each, plus both 2003 Jaguar quarterbacks just for the heck of it.