Writers of Pro Football Prospectus 2008

28 Oct 2010

ESPN: Fantasy Matchups 2010, Week 8

Bench Drew Brees and start Jon Kitna? Really? Yes, really, plus more matchup advice.

Posted by: Bill Barnwell on 28 Oct 2010

12 comments, Last at 01 Nov 2010, 10:10am by morrongiello


by morrongiello :: Thu, 10/28/2010 - 10:10pm

"The Steelers are an elite defense by any stretch of the imagination"

I must have hit the wrong buttons on my calculator but last time I checked, Steelers opponents are completing passes at a rate of 66%. Yes? No?

And it's not like they've seen a procession of hotshots, either. Matt Ryan, Kerry Collins / Vince Young, Josh Freeman, Joe Flacco, Colt McCoy, Chad Henne. Pretty mixed bag, wouldn't you say?

by SammyG (not verified) :: Fri, 10/29/2010 - 8:35am

Since you are posting on the Football Outsiders web site, you ought to be aware that DVOA has the Steelers defense ranked 3rd overall. That includes opponent adjustments, albeit not at full strength, so 3rd overall seems like it should qualify as elite. They rank 6th against the pass, too, so it could be that opponents are checking down and completing passes that don't go anywhere. Also, completion percentage has probably never been used as the defining statistic of a defense.

by ChaosOnion :: Fri, 10/29/2010 - 9:09am

The only data point I am interested in right now is the 3 points PIT-def scored for me last week in a 7 point loss.

by tally :: Fri, 10/29/2010 - 10:25am

Which has nothing to do with Brees' expected output. This article doesn't touch on the Steelers as a fantasy defense, only as a matchup for opposing skill position players.

by morrongiello :: Sat, 10/30/2010 - 2:36pm

I know that their system of crunching data ranks the Steelers defense as 3rd overall. No need to get... DEFENSIVE!

Ok, so what you're saying does make a little sense. A defense could never be considered great while allowing an above average percentage of passes to be completed unless those passes didn't really end up getting the other team a first down, or significant yardage, or points.

But I guess my question is, who cares what the end result of those passes were? Who cares whether it got the other team 8 yards on a 3rd and 20 or 8 yards on a 2nd and 5? It got them 8 yards. Does a defense ever really WANT to give up 8 yards?

Sorry, I know you're not the cult leader around here, but I'd be grateful if you or anyone wanted to discuss...

by Eddo :: Sat, 10/30/2010 - 2:46pm

"But I guess my question is, who cares what the end result of those passes were? Who cares whether it got the other team 8 yards on a 3rd and 20 or 8 yards on a 2nd and 5? It got them 8 yards. Does a defense ever really WANT to give up 8 yards?"

In an extremely literal sense, no, they don't want to give up eight yards. However, defenses are perfectly willing to give up an eight-yard completion on third-and-20; that's why they play three- or four-deep zone coverage in such a situation. Any defensive coordinator would glady concede eight yards with 20 to go, when it means the other team will wind up punting.

by morrongiello :: Sat, 10/30/2010 - 6:05pm

Yeah, that makes perfect sense. And if it's 2nd and 10 then a defense doesn't want to give up the 8 yards because it's considered a "success" (a certain percentage of what's needed for a new set of downs), right?

So does the same apply for rushing plays? Because whereas a completion is a yes-or-no thing, the running back has the ball in his hands most of the time, so it would seem like defenses are just literally trying to hold him to as little yardage as possible.

by Felden (not verified) :: Sun, 10/31/2010 - 1:09pm

Basically, yes.

I can't remember the exact numbers, but I think in 'normal' situations (i.e., not in the red zone, not in the two minute drill, etc), a 'successful' play for the offense is 40% of the yards to go on 1st down (i.e 4 yards on 1st and 10), 60% on 2nd down (i.e. 4 yards on 2nd and 6), and 100% on 3rd or 4th down. Less than that, and it's a defensive win.

Not that it's binary; a 4 yard loss is ALWAYS better than a 4 yard gain, it's just a matter of how MUCH better, depending on the situation.

by Jayson Werth (not verified) :: Sat, 10/30/2010 - 1:42am

I just got INsider for the purpose of this column, so help me out fellas! Should I cut D. Amendola or P. Crayton for M. Sims-Walker this week and his matchup against DAL?

by morrongiello :: Sun, 10/31/2010 - 11:30pm

So how did benching Drew Brees work out for everyone this week? 77% for 305 yards.

by Jayson Werth (not verified) :: Mon, 11/01/2010 - 1:22am

You can harp on the Drew Brees prediction if you'd like, but this week's data was right on the money with several players, including David Garrard, Mike Sims-Walker, and Jamaal Charles.

by morrongiello :: Mon, 11/01/2010 - 10:10am

I agree, and I give them credit for the work that they do every week. It's a great help to me and a fascinating read.

To me, the only thing that really stuck out was this Steelers' defense. I don't understand defense as well as these guys but if a team is awarding its opponents a high rate of catches it tells me, "gee, this defense might not be as good as everyone else in the room thinks they are". This weekend, I was right. Next weekend, I'll be wrong. And so on, being right about the same number of times as I am about the outcome of a coin flip.

But, yes, thank you for your praises. It's really not necessary!