Writers of Pro Football Prospectus 2008

16 Sep 2009

Quarterback Accuracy: Beyond the Ratings

Doug Farrar will be writing two articles a week this year for the Washington Post; here's one on re-examining quarterback rating.

Posted by: Bill Barnwell on 16 Sep 2009

20 comments, Last at 17 Sep 2009, 1:23pm by tuluse


by Bobman :: Wed, 09/16/2009 - 1:52pm

Ooooh, Doug, why so short a piece? Just when it was getting interesting. Oh well, it was nice, but hopefully you can expand on it in future weeks. (I was hoping for examples, such as a 14 yard completion on 3rd and 15 from your own 20 gives more benefit to the standard system than the FO system, but in reality does not help the team all that much. Therefore, it should not give the QB more benefit than a converted 3rd and 4 when the team is trailing by 2 in the final few minutes....

by NoraDaddy :: Wed, 09/16/2009 - 3:01pm

This link seems like the real article. The page linked in the header seems to be a shortened version of it.


by Key19 :: Wed, 09/16/2009 - 1:59pm

I thought that the most disappointing part (besides the length of the piece) was that it didn't even address the initial question, which was "what does the Philip Rivers having the highest QB rating in the league last season really mean?" The piece followed up merely with an explanation of what DVOA and DYAR are, then pretty much just highlighted that Drew Brees is accurate and so is Jason Campbell.

by White Rose Duelist :: Wed, 09/16/2009 - 2:06pm

They need to give you some more column inches.

by Q (not verified) :: Wed, 09/16/2009 - 2:20pm

To address Bobman's point, I still wonder if DVOA gives enough value to a 14 yard gain on 3rd and 15. 14 yards of field position is hugely important in football. For example compare the expected points on a drive for a team starting at the 26 as opposed to the 40.

by DrewTS (not verified) :: Wed, 09/16/2009 - 2:45pm

I think DVOA still gives a decent amount of credit for a 14 yard gain on 3rd and 15. It just acknowledges that a 4 yard gain on 3rd and 3 is more valuable, because it is. Hell, I'd rather have the ball than any amount of field position.

I remember there was a study done somewhere that figured out a point after which you were so far from the goal that it was actually better to be on defense than offense. But factoring in my Colts-fan adjustment, I figure that line is about 20 yards behind the goal line.

by greybeard :: Thu, 09/17/2009 - 12:53am

4 yards on 3rd and 3 is more valuable than 14 yards on 3rd down based on what happens next. If sustaining the drive does not result in 10 more yards than you were better off with gaining 14 yards. If you were behind and did not want to loose time but had to punt after the next 3 downs, 14 yards on 3rd and 15 is better.

Besides whether 4 yards on 3rd and 3 is better than 14 yards on 3rd and 15 for the team is beyond the point. This is a rating for the QB. Which one is harder, getting 14 yards on 3rd and 15 or 4 yards on 3rd and 3? I don't have any statistics to back me up but I suspect that the former is harder. Accomplishing something harder should give you more awards.

The QB rating and DVOA both indirectly reward 4 yards on 3rd and 3. Qb rating favors touchdown, so obviously the QB has a better chance of throwing a TD when he gets a first down. Same for DVOA. But the DVOA rewards it twice: once for sustaining the drive, and indirectly for the TD when sustaining the drive results in that.

by vinyltoupee (not verified) :: Wed, 09/16/2009 - 3:18pm
by The Guy You Don't Want to Hear (not verified) :: Wed, 09/16/2009 - 4:04pm

Should we call it football sabermetrics or "safermetrics?"

by jebmak :: Thu, 09/17/2009 - 10:26am

I prefer the "A Cafellas".

by Will Allen (not verified) :: Wed, 09/16/2009 - 4:09pm

I thought it interesting that the qbs who were right behind Canmpbell in suffering from dropped passes are considered Pro Bowl or even HOF caliber. I have no idea what that means, if anything.

by Bowl Game Anomaly :: Wed, 09/16/2009 - 10:36pm

I noticed that too. Could it be an effect of volume (i.e., if you have a ton of attempts you will have a ton of drops)?

by NRG :: Wed, 09/16/2009 - 11:01pm

I suspect that explains Brees' presence on the list, at least. God knows the Saints throw a lot.

by Doug Farrar :: Wed, 09/16/2009 - 9:58pm

Here's what happened -- the Post informed me yesterday that they'd like an excerpt of Smarter Stats for the paper every Wednesday, "and can you write it right now, please?" (Obvious answer: "Sure!") That's the shorter blurb. The longer one (written a day early to link up with the paper) is the regular internet column.

by Bowl Game Anomaly :: Wed, 09/16/2009 - 10:37pm

Makes sense, but please link to the longer version in the future.

by Independent George :: Thu, 09/17/2009 - 10:14am

Hey, with the price of pixels the way they are, I can understand why one might prefer to link the blurb.

by Doug Farrar :: Thu, 09/17/2009 - 10:25am

Link fixed. Since it's football season, we get a volume discount on pixels.

by CWS (not verified) :: Thu, 09/17/2009 - 11:44am

In his article on receiving plus/minus Bill found that the percentage of dropped passes was more dependent on the QB and scheme than on WRs. I know that there probably needs to be more research into this but doesn't that mean that we can't just assume that drops are something that the QB is not responsible for?

by TGT2 (not verified) :: Thu, 09/17/2009 - 1:13pm

The league average for Quarterback Accuracy was 82.6 percent in 2008, putting 17.3 percent of the average quarterback's incompletions on his own head

I don't think the proper words are being used there. It's not 17.3 percent of the incompletions are on the QB; it's 17.3 percent of the passes are incomplete due to the QB. That's a pretty important difference. It's misstated multiple times farther down in the article as well.

I got the point of the article and I liked it, but that word choice was more than confusing. It was flat out wrong.

by tuluse :: Thu, 09/17/2009 - 1:23pm

I don't think I understand the difference, can you clarify?