Writers of Pro Football Prospectus 2008

30 Oct 2010

Matt Schaub Could Be Big Against Colts

This week's MNF feature looks at other times that a running back runs for over 200 yards against a division rival -- and what has happened in the second meeting. Hint: It is good news for Matt Schaub fantasy owners.

Posted by: Tom Gower on 30 Oct 2010

9 comments, Last at 03 Nov 2010, 10:34am by Tom Gower


by Andrew Potter :: Sat, 10/30/2010 - 4:42pm

It's Insider-only, but the usual ESPN label isn't on the front page and there's no indication of it being Insider-only here. Any chance that could be added please?

by Joseph :: Sat, 10/30/2010 - 5:53pm

Um, it's written on the big blue header right above the ad.

by Andrew Potter :: Sun, 10/31/2010 - 1:57am

Usually with ESPN Insider articles they're clearly indicated in the header and on the FO homepage link. This one is neither, which caused me to think "oh, great, the Monday Night feature's not Insider-only this week".

The big blue header says it's published by ESPN, but nothing about it being Insider-only. It's not a big deal, but the usual indications of Insider-only content aren't present.

by Basilicus :: Sun, 10/31/2010 - 12:03pm

Agreed. Not a big deal either way, but it's nice not to get one's hopes up.

by James19 (not verified) :: Sat, 10/30/2010 - 6:45pm

Footballoutsiders hits a new low with this article. A 10 game sample size over multiple years means absolutely nothing especially when you took out the 2 passing outliers. Essentially you took 20% of an already small sample to make it more unreliable. How about telling us something new that we don't know about that we can't simply figure on our own like regression to the mean. You mean to tell me that guys who rush for 200 yards don't do it a 2nd time. That is shocking to me. Who would have guessed that it is difficult to repeat a rare occurrence. This is similar to the bullshit 370 "curse". All it is is a small sample size that takes out a huge chunk of its sample (Dickerson and Smith) and says that they regress to the mean. That doesn't tell anybody anything they don't know already.

by SFC B (not verified) :: Mon, 11/01/2010 - 8:57am

Did someone at FO run over someone's puppy?

I don't always read entire comment threads, but lately I've noticed an increase in the number of insulting, even abusive, comments from folks who disagree with something written here. Has this always been the case and I just missed it, or has FO's increased exposure through the NYT and ESPN brought in more people whose oxes are being gored?

by RichC (not verified) :: Mon, 11/01/2010 - 11:17am

He didn't put it out in the best manner, but absolutely everything he said is true.

by morrongiello :: Tue, 11/02/2010 - 6:12pm

I love how everyone who criticizes FO's methodology or even just a particular article is suspected of having an agenda. C'mon, man, isn't it fair to criticize? It sparks a lot more discussion than a verbal beejay.

by Tom Gower :: Wed, 11/03/2010 - 10:34am

A slight re-statement of the points of the article:
1. Backs who run for 200 yards in the first game run for fewer yards in the second game. This is accurately classified as a "no duh" result.
2. Normally, when a player runs for 200 yards, his team has an unexceptional to bad passing performance by DVOA terms. I thought this trend was non-obvious. There have been, however, exceptions to that, which I had to acknowledge.
3. When the player's team has an unexceptional passing performance during his 200 yard game, that passing performance dramatically improves for the second game. This strikes me as understandable but not necessarily obvious.
4. Overall team offensive DVOA is approximately the same the second game as it was in the 200-yard rushing game. Given that we normally think of particularly individual performances as the result of outlying events unlikely to recur, I thought this result was non-obvious.

Whether or not I communicated those points effectively I will leave up to the reader to decide. As far as small sample size trend pieces go, I thought this piece came out reasonably well, though obviously if you don't believe we should ever write that kind of article, you probably won't agree.