Writers of Pro Football Prospectus 2008

02 Oct 2017

FO Mailbag: DVOA vs. Receiver Types

We haven't done one of these FO Mailbag posts in a while, but with a lot of people using our Defense vs. Types of Receivers numbers for building DFS lineups, I'm hoping this will help them out.

Nick Lucano: Absolutely love your site and I would be lost without it when playing NFL DFS.

With that being said this is not a complaint about anything being done wrong, more just a question about the DVOA ranking for Tennessee defense vs. WR1 and I am wondering if I missed something here.

When looking at your charts you have them as a ranking of 6th overall with a -26.9% DVOA. I am a little bit confused about this result. The Tennessee Titans have been continuing their trend of last season and being awful and getting torched vs. both WR1 and WR2 positions.

If you look at the three games so far this season, the proof is still there with the amount of targets and receptions and yards they are giving up to both WR1 and WR2.

So again, my question: Am I missing something as far as how DVOA was calculated? Because I dont see how they can be ranked to highly overall vs. WR1!

Aaron Schatz: So it turns out the issue here is one of "rate stat" vs. "total stat" and also the difference between what DVOA is trying to measure and what is useful for DFS purposes.

You are right that the Titans gave up a lot of yardage to WR1s in the first three weeks. But they also got targeted a TON. WR1 (Doug Baldwin, Amari Cooper, and Marqise Lee) were targeted 40 times in the first three Tennessee games. No other defense faced more than 33 targets to WR1 in the first three weeks.

The catch rate by those three receivers was 55 percent, less than the WR1 league average of 62 percent. But does that matter for DFS purposes? Nope! Also, DVOA accounts for the fact that the Titans defense picked off two passes intended for those three receivers, making them one of only four defenses with two or more picks against WR1 in Weeks 1-3. Does that matter for DFS purposes? Also nope!

However, the fact that DVOA is not measuring just blunt "yards and targets" averages is why the DVOA vs. types of receivers table also features yards per game and targets per game, which will both be adjusted for opponent starting after Week 4. You probably notice that through Week 3, the Titans are second in most passes per game to WR1 (12.2) and tied for eighth in most yards per game by WR1 (81.0). Those are the numbers that matter more for DFS purposes.

Of course, the Titans will be much worse once I update things on Tuesday afternoon. DeAndre Hopkins caught 10-of-12 passes on them for 107 yards, plus a 13-yard DPI. But guess what? The Titans also picked off a pass intended for Hopkins, their third interception on a pass intended for WR1 this season. Weird.

Posted by: Aaron Schatz on 02 Oct 2017

6 comments, Last at 05 Oct 2017, 9:16pm by Vincent Verhei

Comments

1
by Vincent Verhei :: Tue, 10/03/2017 - 1:06am

We also got a lot of questions on this regarding Arizona and Patrick Peterson -- specifically, why Arizona was ranked last in coverage against WR1s through Week 3, when Peterson appears to be having a fabulous year. Here is the response I gave to those folks. It's pretty similar to what Aaron wrote above:

---------

Teams are ranked according to DVOA on passes thrown to each category of receiver (DVOA on throws to No. 1s, DVOA on throws to No. 2s, etc.) In Arizona's case, coming into Week 3, opposing No. 1s are catching 80 percent of their targets, for nearly 10 yards per pass, with a pair of touchdowns. Those are all bad numbers.

However, those are all rate stats. While No. 1s are making the most of their opportunities against Arizona, as you noted, they aren't getting many opportunities in the first place -- 3.5 passes per game. You'll see on the Team Defense page that this is the lowest number in the league. And they're only giving up 32.7 yards per game to No. 1s, which is lower than anybody except Jacksonville.

This is something we sometimes see with A-plus corners -- their rate stats are worse than you'd expect because opponents only throw at them on the rare occasions when they are very clearly beaten. There's also the fact that we're only three games into the season, and some of this weirdness will sort itself out over time.

But we like to say that there is no one number that evaluates everything a player or team can do. That's why we put multiple numbers out there to paint a more vivid picture of what's going on in the NFL.

2
by Crunchums :: Tue, 10/03/2017 - 9:20am

"The Titans also picked off a pass intended for Watson" - I assume you mean Hopkins here?

4
by Aaron Schatz :: Tue, 10/03/2017 - 11:22am

Yes! I'll fix that.

3
by ALauff :: Tue, 10/03/2017 - 10:18am

Great information--thank you. I've been using that page for lineup decisions for years.

5
by mackodocious :: Thu, 10/05/2017 - 1:27pm

How do you determine who's considered a #1 wide receiver? Is it based on targets or where the wide receiver lines up? Like, is Larry Fitzgerald considered a number 1 wide receiver for this metric even though he's primarily in the slot?

6
by Vincent Verhei :: Thu, 10/05/2017 - 9:16pm

Again, because somebody else recently asked this, I will copy and paste:

------------

There's no magic formula in determining who is the No. 1 receiver for
each team. It can change from game to game based on injuries and other
factors. The biggest question is this: which receiver does the
opposing defense devote most of its attention to? There was one year
for Detroit where Golden Tate led the team in targets, receptions, and
yards, but by watching the games it was clear that Calvin Johnson was
really the WR1 and Tate the WR2. So we just take it one week at a time
and do the best we can.

------------

To answer your specific question: yes, Larry Fitzgerald is absolutely the WR1 for the Cardinals right now.