Wide Receivers and Further Exploring the Limitations of Statistics

by John Morgan on Feb 15, 2010 6:45 AM PST

I have explored this quite a bit, but short of making it a crusade, I want to invite you to share your opinion with one of the most respected statisticians working in football, Brian Burke. He asks:

Here's a question. I've been applying the Win Probability model and Expected Points model to individual player performance lately. For QBs, it's fairly straightforward--we add up all the successful events like completions and touchdowns, and subtract out all the unsuccessful events like incompletions and turnovers. But what about receivers? Should we count incompletions and interceptions against them?

People know my opinion about applying stats to specific players. I would argue even applying stats to quarterbacks is futile, as its attempting to create a universal measure across completely different offenses with different abilities and different schemes, and especially difficult because unlike baseball, a single in the ninth inning of a blowout is not as easy as a single in the first inning of a scoreless game. A completion when a team has nothing to lose by allowing it, say in the final five minutes of a thirty point blow out, is much easier than a completion in the last minute of the third in a one score game. The achievement is superficially the same, but the game state and game conditions are completely different. And the game state and conditions impact the pass rush and coverage schemes, along with the players that populate those schemes, their effort and their decision making.

Regarding Brian's specific question, I think it's better to include all information and produce a stat that measures "targets to that receiver" rather than the receiver himself. That measure is not possible with the information currently provided by the NFL, and truly, is not maybe possible given the information we will ever have.

What do you think?

Tweet Comment 13 comments | 0 recs |

Do you like this story?

Comments

How would one determine, for example,

if a particular interception was due to the ball careening off a receivers’ hands, or if the ball was thrown into double coverage, or if the receiver ran the wrong route, or any number of other scenarios. I guess if there were a way to assign responsibility for the interception and incompletion (because many of the same questions apply) then I’d have no problem counting them against receivers. All that said, I’m not nearly as invested in stats as others.

I'm gonna go calm submissive on your ass.

by Dukeshire on Feb 15, 2010 8:00 AM PST reply actions

Subjective observation would seem to be the most effective way, wouldn't it?

Game charting, for example.

Over a career, I’m almost not concerned at all about not being able to accurately assign credit, blame, and impact on every play. All players share the same affliction of being being accurately measured, it balances out some and over time the numbers bear the significant things out.

Over one season, though, or a game, there’s just such few instances to work with that you have to be so careful. Only 16 games, often only 75 snaps, give or take, in a game. A pitcher can easily toss over 3,000 pitches in a season.

by jacobstevens on Feb 15, 2010 10:38 AM PST up reply actions

How is 'targets to that receiver' different from the available stat: "Targets"?

Andre Johnson led the league in 2009 with 171 targets . Or, 170, depending on who you ask. Or is it a matter of these available target stats being unofficial, or, simply that the ‘Target’ stats are at least partially inaccurate?

by Misfit74 on Feb 15, 2010 10:02 AM PST reply actions

They aren't. Discrepancy in stats is usually just that some sites remove hail marys and other non representational plays.

by John Morgan on Feb 15, 2010 10:07 AM PST up reply actions

I wouldn't mind a super-modified 'Catch-Rate' stat.

Receptions/Targets is useful, but I think it could be constructed to be more accurate. What are your thoughts on Catch-Rate?

by Misfit74 on Feb 15, 2010 10:12 AM PST up reply actions

If you're going to do it,

if you’re going to measure the impact of a WR, it’s more accurate to measure their impact in whole, than to subtract any potential part of their impact for the sake of reducing un-isolated variables like the QBs’ impact, or that of the coordinator, line, running game, defense, schedule or overall team.

I dunno if “impact in whole” is the best description, but for lack of a better one. Incompletions and turnovers, the receiver can have an impact on. I would not be terribly interested in a metric that left that out. A metric that can’t fully isolate and account for teammates, well, that’s football stats, I can live with that.

by jacobstevens on Feb 15, 2010 10:33 AM PST reply actions

I pretty much agree with the comment you made over there

The one thing I wonder about is including interceptions. For starters, if I had to guess what percentage of interceptions were on the wide receivers I think it’d be a fairly small number, maybe 25% at most. It’s always seemed to me that blaming a pick on the receiver is a way for an announcer to look smart or chide a diva receiver when in reality nobody outside the team has any definitive knowledge about the route or timing on any given play. I tend to think that receiver interceptions are overblown.

The other problem I have with including picks is that there is a third element involved. I haven’t seen any mention of the defender being another variable. Sometimes defenders just make a great play, or a better play, and theres not much you can do about it. Should a receiver be penalized because Darrelle Revis is really good? It seems like on top of measuring both the quarterback and the receiver (and the line and the play call and game situation….) you’re also measuring the defenders ability.

by Nate Dogg on Feb 15, 2010 2:03 PM PST reply actions

You sure are

But you are measuring the corner, pass rush and overall coverage scheme when you attempt to quantify quarterback or receiver. It’s the same thing that led many to think D.J. Hackett was such an exceptional receiver. The guy was rarely playing anything but fourth wide receiver and as such was competing against dime backs and safeties. Stats, what can you say? They might provide more bad information than good.

by John Morgan on Feb 15, 2010 3:07 PM PST up reply actions

Lately it sure has felt like they provide more bad info than good.

I will say this for stats: I imagine if my football knowledge was only grown on scouting, I’d probably have a pretty limited view and understanding. Built on platitudes and truthiness. I feel better for having gone all Aikman Efficiency Ratings, Curse of 370, etc., then come back and lived to tell about it.

by jacobstevens on Feb 15, 2010 4:18 PM PST up reply actions

Some Interceptions should be given to recievers

It just seems unfair when a QB throws the ball to a reciever who tips it, and then it’s picked off, because most of the time it’s a clearly catchable ball, and the QB doesn’t deserve to be given an interception on a decent pass. Which is why passes like that should be (I think) recorded as reciever INT’s. Agree?

by why's-guy on Feb 15, 2010 9:53 PM PST reply actions