By Bill Barnwell, Football Outsiders – special to BSMW Patriots Game Day

The David Carr Conundrum

It’s hard to write a column about the Houston Texans without taking a look at the decision they made with regards to this year’s draft. Everyone who’s going to read this knows the story: they passed up Vince Young and Reggie Bush to stick with David Carr and Domanick Davis, and selected Mario Williams. At the time of the decision, I mostly agreed with this — running backs are fungible properties and I felt that Carr hadn’t been given a chance to shine. The only difference between the Texans’ decision and my own would have been the selection of D’Brickashaw Ferguson instead of Mario Williams. Both have struggled in their first year. Davis was placed on IR before the season with a knee injury that failed to heal; even so, the Texans rush attack is the 17th best in the league according to Football Outsiders’ DVOA metric, which measures their performance in each down, distance, and opponent situation versus that of the rest of the league. Perhaps more interestingly, their pass offense grades out as 13th — above-average! When you consider that Houston’s rush offense was 13th last year and its passing attack 31st, that appears to be a huge step forward.

Still, though, DVOA is just a metric and should be used in combination with observation and other data. If you’ve followed the Texans this season, what you’ve heard are increased rumblings, much like last season, about David Carr’s job security. Several weeks ago, he was benched halfway through a game after turning the ball over twice for Sage Rosenfels. Now, I don’t mean to disparage Mr. Rosenfels, but it’s Sage Rosenfels. This was not Gary Kubiak benching Carr so that he could get a rookie some game experience; he was benching his starting quarterback for his older, veteran backup. Speculation as to whether Kubiak was just living out some bizarre punitive fantasy involving a benching of John Elway is ridiculous, but Kubiak was clearly unhappy with the play of Carr. Further discussion of Carr’s impending dismissal has come out since then — do a Google News Search for Carr and most, if not all, of the articles will point the Texans “mistake” out.

The thing is, looking at Carr’s statistical line, it’s hard to reconcile that with what people are saying. He leads the NFL in completion percentage at 69.4%; Drew Brees, in second at 66.4%, is closer to eighth than first. While you can make a case that this is due to Houston’s offensive style, it’s some evidence that Carr is at least doing a decent job of performing within the system he plays in; it’s not as if he’s wildly inaccurate. Let’s then look at his yards per attempt, to see if he’s actually getting the ball anywhere:


As you can see, his numbers were going up from year-to-year at a pretty steady rate (the average yards per attempt for a starting QB, since the merger, is 6.88) — until last year, when his numbers went down to the levels of his rookie season. When I was compiling this data, I began to notice that Carr was one of the few quarterbacks that had this happen to him, and that this might say something about a quarterback’s professional feasibility.

I took every quarterback since the merger and eliminated all seasons where they didn’t throw at least fifteen passes a game and/or played fewer than eight games — since we’re looking at a rate stat, it’s okay if we include the strike years in this. From there, I simply subtracted their yards per attempt in their first season from their fourth, getting rid of the quarterbacks who hadn’t had four starting seasons in the process. Here are the quarterbacks who saw little to no change in their Yards per Attempt over those four seasons:


As you can see, it’s not a terribly impressive group — the names that you would find impressive, the Dan Marino, Mark Rypien, and Warren Moon’s, for example, all have significantly superior levels of yards per attempt than Carr to begin with. It’s simply hard to improve when you are already averaging 8+ yards per attempt. Terry Bradshaw looms as someone who improved significantly in his sixth year as a starter, but he appears to be one of the exceptions to the rule.

Here’s a list of the ten most and least improved quarterbacks when it comes to yards per attempt:


A player like Steve Young is a little misleading because he was playing in a different offense by the time he had his fourth season, but there’s nothing stopping the players on the bottom half of the list from going to a different offense and improving, either.

See that column to the right that I added? That’s the number of games each player played after his fourth starting season in the NFL. The players on the top of the list played 573 games after their fourth seasons; the players at the bottom of the list, 333. (The + symbols are for Donovan McNabb, Jake Plummer, Kurt Warner, and Aaron Brooks, all of whom are still playing in the NFL; if anyone wants to make a nifty prop bet with me, I’ll take Donovan McNabb having more starts than the other three combined for the rest of their respective NFL careers.)

The 20 guys most similar to and David Carr? Their games played numbers after starting season #4 are below.


They averaged 63 games played after their fourth season, nearly seven more than the players whose yards per attempt improved dramatically, but 30 more than those who saw their numbers decline dramatically.

So, then, it’s really hard to make a case that Carr’s yards per attempt stagnating after last season would be a dramatic point against him being a NFL-caliber quarterback.

What then, do we say about David Carr’s chances? Why is he so maligned? Interceptions? Can’t be. Starting quarterbacks since the merger throw an interception, on average, once every 30.93 throws. Carr threw an interception every 38.46 throws in 2005. Fumbles? Maybe, but is that Carr’s fault or the shellacking he takes? His offensive line, according to Football Outsiders Adjusted Line Yards metric, has ranked 32nd, 25th, 30th, 32nd, and this year 27th in protecting Carr from harmful DL waves. With that in mind, it’s hard to point out anything that Carr really does too incorrectly, and points even further to the benefits of having selected D’Brickashaw Ferguson in this year’s draft — even if Ferguson would’ve struggled in a similar manner in Houston, it would at least give David Carr a chance.


As you can see, Houston’s had the worst offensive line in football over the last five years for pass blocking. Note that the teams around them have also struggled to throw the ball effectively. Is that a coincidence? It might take a Pro Bowl performance by David Carr in Miami for the Texans to find out.