Writers of Pro Football Prospectus 2008

09 Sep 2009

An Outsider's Perspective on the Redskins

As part of the Washington Post's Redskins preview section, I did a Q & A on the issues that kept Washington's offense from succeeding down the stretch, and how the team projects in 2009.

Posted by: Doug Farrar on 09 Sep 2009

8 comments, Last at 10 Sep 2009, 9:07am by perly


by crack (not verified) :: Wed, 09/09/2009 - 12:09pm

Is it just me or does the first response dodge the question of why the redskins offense fell off a cliff and actually answer the question of the different ways to show that it did fall off a cliff?

by perly :: Wed, 09/09/2009 - 12:27pm

It's not the most complete picture they've drawn, and there's a weird mix of FO and conventional stats. Chris Samuels and Randy Thomas got hurt mid-season, the running back depth was so sketchy that Shaun Alexander got game reps, and the receivers weren't good enough to pick up the slack.

That said, I'd be genuinely surprised to learn that Washington's offensive DVOA in their Week 8 win @ DET was worse than in their Week 9 loss to PIT, and I'm assuming the Week 1-9/10-17 split is similarly off.

by Jeremy Billones :: Wed, 09/09/2009 - 3:01pm

They only beat DET 25-17, kicking 4 field goals (and missing a 2 point conversion) and allowing 21-35-223 through the air. In contrast, in the 23-6 loss to PIT they held them to 29-64 rushing and 12-27-179 passing, while putting up a 15-60 rushing of their own.

Once you adjust up for the PIT defense, and down for the DET offense and defense, I can believe their DVOA was better in the loss.

by perly :: Wed, 09/09/2009 - 3:56pm

The article only talks about offensive DVOA, though, so the defensive performances shouldn't offset. But those are some cherry-picked numbers (not that we're not all guilty of that from time to time). 22 of CP's 51 rushing yards came on a single carry. Campbell threw 2 interceptions and was sacked 7 times. And it was the only week they didn't score a TD.

I suppose it's possible with the opponent adjustment that it wasn't their worst offensive DVOA performance of the year, but it was certainly the most painful to watch. Their performance in Detroit wasn't awful; it was just disappointing.

by Jeremy Billones :: Wed, 09/09/2009 - 5:24pm

"22 of CP's 51 rushing yards came on a single carry."

You can say something similar about most rushing perfomances; it's the same principle that says the winning baseball team will score more runs in one inning than the other team does in the rest of the game. (Against DET, it was 31 of his 126 yards -- 23 for 95 are good stats, but not against DET.)

Detroit's rushing yards allowed (ordered): 318, 292, 264, 211, 182, 181, 157, 154, 150, 135, 135, 133, 123, 116, 106, 97, for a 172 average. Washington had one of the 135s, below average.

Detroit's passing yards allowed (ordered): 532, 484, 474, 456, 447, 439, 425, 421, 404, 392, 384, 370, 362, 320, 301, 255, for a 404 average. Washington had the 439, a good game. The total yards work out to average, but on a percentage basis the running was more bad than the passing was good.

For Pittsburgh running: 122, 117, 112, 106, 103, 95, 84, 83, 75, 66, 65, 62, 60, 53, 43, 38, average of 80. Washington had 60, very bad.

For Pittsburgh passing: 228, 205, 199, 195, 194, 175, 165, 161, 159, 155, 152, 145, 140, 128, 90, 20(!), average of 157. Washington had 161.

So, on raw numbers, it does look like Washington played worse against PIT -- below average rushing both times, above average passing both times, each worse against PIT. There is a second-order adjustment to DVOA, but both PIT and DET faced a 7.7% schedule in 2008 and I'm too lazy to look up all 32 opposing offenses.

by perly :: Thu, 09/10/2009 - 9:07am

PFR's of limited use for this sort of thing, because the aggregate totals don't tell you much about a game.

It won't tell you that the Skins got something like 75% of their yardage in the last 20 minutes of the game, down by 3 scores. They had 3 possessions in the red zone, ending in one interception and two turnovers on downs (at the 12 and, dear god, the 1). They started two possessions inside of Pittsburgh territory and settled for field goals because they couldn't get a first down. In their first 10 drives, the Redskins gained something like 50 total yards and 2 first downs. If you didn't watch the game (I was in the awkward company of Steeler fans), take a look at the gamebook. It was that bad.

And DET was Jason Campbell's best game of the season. DVOA may disagree, but if it thinks 23/28 for 328 yards and a TD is the worst *anything* (even allowing for the two fumbles and substandard opposition), it's just wrong.

by tomdrees :: Wed, 09/09/2009 - 2:27pm

It's weird that there were no projections of improvement for the young wide receivers, especially considering the preseason that Marko Mitchell had. I know he's a seventh round rookie dark horse who just barely clawed his way onto the roster, but he's been a bright spot.

A common observation about the 2008 Skins has been "they failed in the second half of '08 because their rookie WRs didn't step up production as the O-line and running backs got injured." I don't have numbers to back this up, but I'd venture to say that for a West Coast timing-based possession receiver to produce as a rookie in newly installed offense that even veterans are struggling with is fairly rare, especially if that rookie is not on the field due to injury. I understand that: 1. considering how Campbell typically throws an accurate ball, a 42% catch rate is atrocious, and 2. health is a skill, but rookie WRs also have a history of under-producing. I find it very odd that not even one projection allows for significant improvement in the passing game.

Also, on a completely unrelated note: I'm somewhat new to the site, but I haven't heard anything explaining the significance of Variance. What does it mean? Is it an indicator of anything at all? It's hard to reconcile the fact that the Skins had the least Variance in the league with the media (FO included) consensus that the story of the 2008 Redskins was "A tale of two seasons." What explains this? Was it the defense surging ahead as the offense faltered?

I'm hopeful about the Skins this year, and think the improved pass rush is going to cause turnovers (they could hardly cause fewer turnovers), which will in turn get more reps for the passing game to find a rhythm in, which will ease the burden on the O-line. Fingers crossed. As you can tell I'm a total homer, so this is all with a grain of salt, but I am curious.

by JasonK :: Wed, 09/09/2009 - 4:27pm

I'm guessing that the FO system doesn't project much improvement from the young WRs because there just isn't enough of a sample size upon which to base any kind of projection. Sure, there are lots of WRs who didn't do anything in their rookie seasons, and who developed into better-than-replacement-level players, but there are also lots of WRs (even highly drafted ones) who didn't do anything in their rookie seasons and continued to not do anything of note for the rest of their careers. The objective variables that FO uses don't have an entry for "coaches said he looked good in camp" or the like. Where there is scant data, that's what gets entered into the system, and if these guys not being able to push a lousy starting WR like Randel-El out of some playing time as rookies is a sign of anything, it's not a sign of anything good.