Let me say up front that I honestly do not think the Texans will take Ryan Mathews. Call me a subscriber to BFD's theory of a Mathews-related conspiracy, I guess, but it seems to me that any pro-Mathews talk that might trickle out of Reliant is more than likely leaked with some ulterior motive. That said, in the event that I am wrong -- and, really, what are the odds of THAT? -- I wanted to make one final argument against taking a running back in the first round (and, really, in the first four rounds).
First, some numbers. (Don't worry, Tim; I'll explain with words, too, so that you can understand what's up. There will not, however, be illustrations. So, if Vince Young is staring at his technamalogical box what connects to them there interwebz, he's kind of out of luck. Might I suggest this instead?)
| Pts/G | Plays | Yds/G | Yds/Pl | 1stD/G |
| 31.9 | 1,032 | 403.8 | 6.3 | 21.8 |
| 24.2 | 1,043 | 383.1 | 5.9 | 21.2 |
| Pts/G | Plays | Yds/G | Yds/Pl | 1stD/G |
| 28.9 | 1,047 | 410.7 | 6.3 | 22.1 |
| 22.9 | 1,019 | 382.1 | 6.0 | 21.2 |
More after the jump...
The first row of numbers in each of the tables above represents the NFL's best offense (in both cases, New Orleans) in 2009 and 2008, respectively. The second row of numbers represents the offense of your Houston Texans in the same season. There are a number of interesting nuggets buried in there (for example, New Orleans' Yds/G and total plays went down from '08 to '09, but their scoring increased), but I want to focus primarily on the Yds/G column.
In fact, let's throw in one more table for good measure:
| Year | Yds/G |
| 2009 | 403.8 |
| 2008 | 410.7 |
| 2007 | 411.2 |
| 2006 | 391.5 |
| 2005 | 387.0 |
| 2004 | 418.4 |
| 2003 | 393.4 |
| 2002 | 389.8 |
Those are, pretty obviously, the yards/game of the best offenses each season that the Houston Texans have fielded a team of professional football players (and David Carr). The average best offense during that span is 400.7 yards/game. So, what do we have?
In 2009, New Orleans had 20.7 more yards/game than did Houston.
In 2008, New Orleans had 28.6 more yards/game than did Houston.
Houston's most recent offense was only 17.9 yards/game worse than the average best offense in the NFL since 2002.
Houston's most recent offense was only 36.3 yards/game worse than the best single season offense since 2002 (which, somewhat oddly, was the Kansas City Chiefs in 2004.)
/kicking dead horse
I belabor the point of these differences in yards/game because I think it is important to look at potential draft choices in terms of what they are potentially worth to your team. (Think of it as an extension of D+ from this post.) Assuming we got a very good RB, the best-case scenario that you could reasonably expect is that he would improve our offense by about 36 yards/game. More likely, the improvement would be somewhat less than that number.
Perhaps the best way to look at the benefit a new RB would bring to this team is in economic terms -- what is Ryan Mathews' (or any RB's) marginal value? According to the Hidden Game of Football, Football Outsiders, and the like, it takes around 75 extra yards in a game to add an extra TD per game (I don't have an online cite for this, and my Hidden Game book is at home, but feel free to double-check me on the exact number). So, assuming a great back and our best-case scenario of 36 yards/game, you are talking about half a TD/game improvement (or, more accurately, about one extra TD every other game).
When you factor in the likely contract for the 20th pick -- 5-years, ~$15MM, $10MM guaranteed -- we actually have a very good example of the concept of diminishing marginal returns. For the non-econ types, DMR says that, where you have a system (say an NFL offense) with fixed inputs (say amount of minutes in a game) and variable inputs (talent of players individually), you reach a point beyond which each additional unit of the variable input (talent) will yield smaller and smaller increases in the output (yardage), which by extension means that each additional unit of output (yardage) is costing you more and more.
Now, the flipside of that 75 yards/game (or whatever the exact number is) is that preventing those yards will generally result in one less TD/game allowed. The Greatest Defense In Texans History(TM) allowed 324.9 yards/game; the league's best yards/game defense, the Jets, allowed 252.3. That's a difference of 72.6 yards/game. If we are giving the best-case scenario for Mathews of 36 yards, then it makes sense (somewhat) to say that the best-case scenario were we to draft an absolutely outstanding defensive player, would be an improvement of about 72 or 73 yards/game. (Not likely, I realize, but neither is Mathews' theoretical 36 yards/game). Given the same likely contract for the 20th pick, the best-case scenarios give you roughly twice as much output (yardage) for your dollar.
When you add to that the fact that The Greatest Defense In Texans History(TM) had some pretty obvious holes (and, based on the Zac Diles Fanboy Love of Kerns, et al, some less obvious ones), it seems reasonable to assume that plugging one of the defensive holes is more likely to yield a larger return than putting a new running back into one of the best passing offenses in football.
Besides, this whole analysis doesn't even factor in that last season, the league averaged more passing yards/game than in any season other than 1995. Theoretically, increasing the Texans' rushing total, to the extent it cut into the passing yards they would have had, would be handicapping them against opponents who are more than happy to take the passing yards. But that's a different post.
Do I think Ryan Mathews could be good in the Houston Texans' system? Sure. But not at that pricetag, and not when the money can be better spent on other positions.