November 10, 2010

Quick & Dirty Roughing The Passer Study

Posted by Derek

The following tables and graphs prove nothing, definitively answer no questions, and should really only be a stepping stone to a larger study.  So long as we're all clear on that, let's move on.

Research question:  Why do the Eagles seem to get jobbed out of some pretty obvious roughing the passer penalties?

Hypothesis:  NFL referees have different standards for black quarterbacks and non-black quarterbacks in terms of what constitutes "roughing the passer."  

Just one more reason to go with Kevin Kolb.

Study methodology:  Pull 2009 penalty data from FO play-by-play spreadsheets.  Code by QB race.  Control for variations in playing time by comparing RTPs to QB pass attempts and sacks (on the theory that maybe players who are more frequently sacked will receive more RTP penalties).

Before we get to the stuff people really care about, some summary tables:

2009RTPs

Sort of obvious why Eagles fans think they're might be an issue.  Note in passing that Jacksonville, Philadelphia, Tennessee and Washington were all teams where a black quarterback took the majority of the snaps last year.  Hmmmm ...

Buffalo is the clear outlier.  Oh, and New England.  Totally didn't see that coming.  In fact, let's look by quarterback:

2009RTPsbyQB

I really did not expect that!  Not in a league that is so careful to ensure the rulebook is applied fairly and consistently.

And now the graphic everyone really wants to see:

2009RTPsbyEth

I ran the numbers two ways, depending upon how you think we should handle Mark Sanchez (and to much less extent Matt Gutierrez).

Certainly some interesting results there.  Black QBs have 16.7% of the pass attempts, 19.9% of the sacks taken, but only 11.3% of the roughing penalties.  If you want to make the numbers balance by adding roughing penalties, here's what you get:

2009RTPsbyEthCorrected

Think Donovan maybe deserved one or two of those 4.6 (or 7.7) extra penalties last year?  I certainly do.

So here's where I repeat what I wrote at the top:

The [preceding] tables and graphs prove nothing, definitively answer no questions, and should really only be a stepping stone to a larger study.  So long as we're all clear on that, let's move on.

Kinda looks like that larger study might be worth doing now, no?

Comments

« Update From The NFL Office: The Right Kind Of Wrong | Main | League In Backfill Mode Now »



Copyright 2010 IgglesBlog. All rights reserved.