What if Passer Ratings Were Used for WR's?

[T-RO]

Very nice. And it doesn't surprise me about Miles. He's crazy good with the YAC.

And to think there were some idiots on this forum recently talking about trading him.

 

[NextGenBoys]

Great research man

 

[NinePointOh]

It's neat and all, but to offer some constructive criticism, it's not as useful of a stat as everyone seems to think.

Most interceptions and most incompletions are either the quarterback's fault or well-defended plays, not mistakes by the receiver. What the stat basically shows is which quarterbacks threw the fewest interceptions and the most yards. Surprise surprise, those quarterbacks were Brett Favre, Philip Rivers, Tony Romo, Aaron Rodgers, and Drew Brees.

But does that mean guys like Andre Johnson, Larry Fitzgerald, Brandon Marshall, and Reggie Wayne were worse because their quarterbacks threw more interceptions last year?

It is conceivable to create some kind of useful "receiver rating" composite statistic, but the parallel to the quarterback rating seems forced here because some of the stats tell you much more about the quarterback than about the receiver.

 

[DallasEast]

Austin DAL      81-120 69% 1,320 11.0ypa 11td 2int 127.8
Rice MIN        84-120 70% 1,322 11.0ypa  8td 1int 125.1
Colston NO      70-102 69% 1,074 10.5ypa  9td 2int 124.4
VJackson SD     68-105 65% 1,167 11.1ypa  9td 2int 123.0
Welker NE      123-152 80% 1,348  8.8ypa  4td 1int 109.4
 
Bolden ARI      85-118 72% 1,029  8.7ypa  5td 2int 105.5 
AJohnson HOU   101-152 66% 1,569 10.3ypa  9td 6int 103.8
Driver GB       70-109 64% 1,061  9.7ypa  6td 3int 103.0
Marshall DEN   101-149 68% 1,120  7.5ypa 10td 4int 101.1
DJackson PHI    63-112 56% 1,168 10.4ypa  9td 5int 100.6
 
Fitzgerald ARI  97-147 66% 1,092  7.4ypa 13td 6int 100.5
Ward PIT        95-129 74% 1,167  9.0ypa  6td 5int 100.5
Smith NYG      107-151 71% 1,220  8.1ypa  7td 4int  99.2
Jennings GB     68-107 64% 1,113 10.4ypa  4td 3int  99.2
Moss NE         83-136 61% 1,264  9.3ypa 13td 8int  99.0
 
Sims-Walker JAX 63-100 63%   869  8.7ypa  7td 4int  97.5
Ochocinco CIN   72-124 58% 1,047  8.4ypa  9td 4int  96.4
Manningham NYG   57-96 59%   825  8.6ypa  5td 2int  96.1
Holmes PIT      79-127 62% 1,248  9.8ypa  5td 4int  94.9
Mason BAL       73-128 57% 1,028  8.0ypa  7td 2int  94.8

catches-targets, catch%, yards, yards per target
interceptions on passes targeted to that receiver
receiver rating

You had to be targeted at least 6 times per game to qualify.

Miles was targeted 120 times, and caught 81 balls.

He averaged 11 yards for every pass thrown his way, whether it was caught or not.

He scored 11 touchdowns, and 2 of the passes intended for him were picked off.

His receiver rating led the NFL last year.

Nice post, percyhoward. VERY nice.

 

[InmanRoshi]

He's going to be even more potent in the slot in nickel with the entire field to work with, a clean release off the line and most likely going up against lesser caliber defensive backs.

It is conceivable to create some kind of useful "receiver rating" composite statistic, but the parallel to the quarterback rating seems forced here because some of the stats tell you much more about the quarterback than about the receiver.

Or maybe a little of both ... the relationship between the quarterback and the WR and their chemistry together. There's no doubt Romo and Austin share a very unique chemistry.

 

[percyhoward]

It's neat and all, but to offer some constructive criticism, it's not as useful of a stat as everyone seems to think.

Most interceptions and most incompletions are either the quarterback's fault or well-defended plays, not mistakes by the receiver. What the stat basically shows is which quarterbacks threw the fewest interceptions and the most yards. Surprise surprise, those quarterbacks were Brett Favre, Philip Rivers, Tony Romo, Aaron Rodgers, and Drew Brees.

But does that mean guys like Andre Johnson, Larry Fitzgerald, Brandon Marshall, and Reggie Wayne were worse because their quarterbacks threw more interceptions last year?

It is conceivable to create some kind of useful "receiver rating" composite statistic, but the parallel to the quarterback rating seems forced here because some of the stats tell you much more about the quarterback than about the receiver.

I appreciate the thoughtful criticism.

You mention Johnson, Fitzgerald, Marshall, and Wayne. For all practical purposes, the first three all made the top 10 on my list. Fitzgerald only missed it by a tenth of a point. Looking at the number of targets for all 4 players, I submit that their high number of targets had just as much to do with "their" high interception totals as their quarterbacks did. After all, in 3 of the 4 cases, you're talking about Pro Bowl QB's.

A wide receiver is not completely without a role in the number of picks his QB will throw. It's the old familiar TD/INT ratio. The more TD oasses your QB throws, the more INT you're willing to put up with. Same is true for the guys on the other end of that pass. It has to be. The risk/reward factor is working heavily in the case of those 4 wide receivers you mentioned. In all but one case--Wayne's--the reward of a potential TD outweighed the risk of a interception.

When multiple receivers on the same team have bad TD/INT ratios, it's obviously a case of their QB throwing a lot of picks overall. That's why Carolina's Steve Smith doesn't make the list. His ratio was 7td/11int. But the WR who lined up on the other end from him, Muhammad, had a ratio of 1td/5int. Obviously the quarterback's fault.

Johnson, Fitzgerald, Marshall, and Wayne don't fit this pattern at all, because their QB's weren't spraying INT's all over the field to different receivers. Their high INT numbers were the result of being targeted so many--and in Wayne's case, too many--times.

Peyton Manning threw 5 more interceptions in 2009 (16) than he averaged over the previous 6 years (11). If you look at the Colts' TD/INT ratios by target, this rise in INT's is no mystery.
Clark 10td/2int
Collie 7td/2int
Garcon 4td/2int
Wayne 10td/10int
So, does Wayne not being in the top 20 in receiver rating reflect having a quarterback who threw too many interceptions? Of course not. Does Wayne not being there have anything at all to do with Manning? Yes, it does. Manning tried to force way too many passes into his #1 WR.

Receiver rating does more than a simple list of receptions, yards, and TD's, because it shows the risk/reward factor that, over a given season, made some WR's (like Fitzgerald and Johnson) more valuable than others (Wayne).

 

[NinePointOh]

Or maybe a little of both ... the relationship between the quarterback and the WR and their chemistry together.

That's probably pretty accurate, except I still don't know what the inclusion of interceptions when targeting a receiver is supposed to tell you. That has very little to do with the WR or their chemistry. It almost always has to do with either the quarterback's mistake or a particularly good play by a defender, which is why it makes a lot more sense in a quarterback stat than a receiver stat.

 

[percyhoward]

Or maybe a little of both ... the relationship between the quarterback and the WR and their chemistry together. There's no doubt Romo and Austin share a very unique chemistry.

I agree. I see it as the classic "it takes two" statistic.

All this receiver rating does is measure the same five variables measured in passer rating. All five of which require a targeted receiver about 95% of the time.

If it stimulates informed discussion about the game, it probably isn't that bad a stat.

 

[THUMPER]

That's probably pretty accurate, except I still don't know what the inclusion of interceptions when targeting a receiver is supposed to tell you. That has very little to do with the WR or their chemistry. It almost always has to do with either the quarterback's mistake or a particularly good play by a defender, which is why it makes a lot more sense in a quarterback stat than a receiver stat.

While I agree that INTs are usually on the QB or DB a WR can have a large effect on that number as well.

If he runs the wrong route or breaks a different direction than his QB thought he would then it is on the WR not the QB. WRs who use their bodies well to shield the DB away from the ball (Mushin Muhammad was great at this a few years ago) or fight for the ball can cut down on the number of INTs.

Terrell Owens was horrible at using his body to shield the defender away from the ball nor did he fight for the ball much. Considering his size and strength I always wondered why he was so bad at this. Irvin was one that was outstanding at it and it helped Aikman keep his INTs low.

So a WR can affect the number of INTs in the following ways:

1. Route running
2. Body position
3. Fighting for the ball

A last one is with tipped passes. A WR that can catch well with his hands typically doesn't have a lot of tipped balls that go off his hands. Guys with "soft" hands do a better job of not allowing the ball to get away from them than guys who catch with their body/arms more. The ball tends to bounce off more often in those cases and you can see a few INTs result from that.

 

[percyhoward]

1. Route running
2. Body position
3. Fighting for the ball
4. Tipped passes

Excellent explanation of the "it takes two" factor.

Because of the above reasons (and because of QB decisions to force a ball to a specific receiver), receivers can have a large impact on unsuccessful attempts on passes intended for them--both with interceptions and also just incompletions. There are players who carry their low catch percentages with them from team to team, quarterback to quarterback.

The more passes you fail to catch, the more your yards per target drops, while your yards per catch stays unaffected. If yards per attempt is a more useful measure than yards per completion for quarterback production, then it follows that yards per target would be more useful than yards per reception for receiver production.

 

[NinePointOh]

You mention Johnson, Fitzgerald, Marshall, and Wayne. For all practical purposes, the first three all made the top 10 on my list. Fitzgerald only missed it by a tenth of a point.

That's true, but that's a necessary-but-not-sufficient condition. Any composite statistic that purports to rank receivers according to a large number of near-exhaustive factors had better put receivers near the top 10 who most would agree are the best at their position, or else nobody would take it seriously. That doesn't mean it's without flaws.

Looking at the number of targets for all 4 players, I submit that their high number of targets had just as much to do with "their" high interception totals as their quarterbacks did. After all, in 3 of the 4 cases, you're talking about Pro Bowl QB's.

They're Pro Bowl QBs (which is why the receivers still rank highly) who still threw a lot of interceptions (which is why the receivers don't rank as highly as those who were thrown to by Brees, Romo, and Favre). Schaub, Manning, and Warner were all in the top 11 of the league in terms of interceptions thrown last year.

And yes, the number of targets had a lot to do with it. When you have a quarterback who throws a lot of interceptions, he's probably going to throw them while aiming at his top receiver. But the question is, is that really the receiver's fault?

A wide receiver is not completely without a role in the number of picks his QB will throw.

I didn't say he plays no role, I said his role is much, much smaller than the role of the QB and the defender. To the point where I wonder whether it's worth including in the first place, even if the receiver has some part in it. For example, a shortstop plays an important role in a no-hitter, but we don't give him credit. We calculate a shortstop's defensive stats, but only after the scorekeeper excludes those plays that he shouldn't be expected to make -- a line drive over his head doesn't count against him.

It's the old familiar TD/INT ratio. The more TD oasses your QB throws, the more INT you're willing to put up with. Same is true for the guys on the other end of that pass. It has to be. The risk/reward factor is working heavily in the case of those 4 wide receivers you mentioned. In all but one case--Wayne's--the reward of a potential TD outweighed the risk of a interception.

I'm not sure what you mean by "outweighed." Fitzgerald is ranked lower than several receivers who caught far fewer touchdowns because Kurt Warner threw just a couple more interceptions when looking in his direction. The problem, since I assume you're using the same QB rating formula, is that even if you choose to include interceptions for a receiver, there's absolutely no justification for weighting it equally as heavily as you would for a quarterback. That's how you end up with Mario Manningham, who was mediocre in virtually every other category, ranked in the top 20.

When multiple receivers on the same team have bad TD/INT ratios, it's obviously a case of their QB throwing a lot of picks overall. That's why Carolina's Steve Smith doesn't make the list. His ratio was 7td/11int. But the WR who lined up on the other end from him, Muhammad, had a ratio of 1td/5int. Obviously the quarterback's fault.

Johnson, Fitzgerald, Marshall, and Wayne don't fit this pattern at all, because their QB's weren't spraying INT's all over the field to different receivers. Their high INT numbers were the result of being targeted so many--and in Wayne's case, too many--times.

Peyton Manning threw 5 more interceptions in 2009 (16) than he averaged over the previous 6 years (11). If you look at the Colts' TD/INT ratios by target, this rise in INT's is no mystery.
Clark 10td/2int
Collie 7td/2int
Garcon 4td/2int
Wayne 10td/10int
So, does Wayne not being in the top 20 in receiver rating reflect having a quarterback who threw too many interceptions? Of course not. Does Wayne not being there have anything at all to do with Manning? Yes, it does. Manning tried to force way too many passes into his #1 WR.

I'm not sure why that's supposed to be an indictment of Wayne, but I can also tell you that's not an accurate account of the interceptions in question. Since Wayne was your example, I went back and re-watched the NFL.com highlights of all ten interceptions, and only three (against Houston, Baltimore, and arguably Jacksonville -- though that was a jump ball in the end zone) can be considered "forced" to Wayne or in any way indicative of Wayne's "riskiness" as a target. I also counted five that were tipped or batted near the line of scrimmage, or where Manning's arm was hit as he threw (two against Denver, one against Tennessee, one against Houston, and one against Arizona) and two (against Seattle and New England) that were thrown very poorly by Manning. If Wayne is your prototype example of a receiver the stat would classify as "risky", I don't think the tape validates your assertion. If anything, I'd say Tony Ugoh is the risky player.

Receiver rating does more than a simple list of receptions, yards, and TD's, because it shows the risk/reward factor that, over a given season, made some WR's (like Fitzgerald and Johnson) more valuable than others (Wayne).

Again, that's only if you accept that when a quarterback throws an interception while aiming for a certain receiver, it's because that receiver is "risky." That's very rarely the case, and Wayne is a great example.

 

[NinePointOh]

While I agree that INTs are usually on the QB or DB a WR can have a large effect on that number as well.

If he runs the wrong route or breaks a different direction than his QB thought he would then it is on the WR not the QB. WRs who use their bodies well to shield the DB away from the ball (Mushin Muhammad was great at this a few years ago) or fight for the ball can cut down on the number of INTs.

Terrell Owens was horrible at using his body to shield the defender away from the ball nor did he fight for the ball much. Considering his size and strength I always wondered why he was so bad at this. Irvin was one that was outstanding at it and it helped Aikman keep his INTs low.

So a WR can affect the number of INTs in the following ways:

1. Route running
2. Body position
3. Fighting for the ball

A last one is with tipped passes. A WR that can catch well with his hands typically doesn't have a lot of tipped balls that go off his hands. Guys with "soft" hands do a better job of not allowing the ball to get away from them than guys who catch with their body/arms more. The ball tends to bounce off more often in those cases and you can see a few INTs result from that.

Again, I don't dispute that WRs can have some impact on the number of interceptions. However, the percentage of interceptions where the receiver is at fault in any noticeable way is so low that it may not be worth including in the formula in the first place. At the very least, you'd need to either (a) actually watch the play and judge whether the receiver had anything to do with it, or (b) weight interceptions far less heavily than you do in the QB rating formula.

 

[percyhoward]

At the end of the day, targets to which receivers helped their teams the most?

Picture #1
Fitzgerald 13 td
Wayne 10 td
Marshall 10 td
Johnson 9 td

Picture #2
Fitzgerald 13 td/6 int
Wayne 10 td/10 int
Marshall 10 td/4 int
Johnson 9 td/6 int

Picture #2 gives you a clearer, better idea.

It's information you can use for a million purposes, and playing the blame game is certainly one of them.

 

[NinePointOh]

At the end of the day, targets to which receivers helped their teams the most?

Picture #1
Fitzgerald 13 td
Wayne 10 td
Marshall 10 td
Johnson 9 td

Picture #2
Fitzgerald 13 td/6 int
Wayne 10 td/10 int
Marshall 10 td/4 int
Johnson 9 td/6 int

Picture #2 gives you a clearer, better idea.

It's information you can use for a million purposes, and playing the blame game is certainly one of them.

That doesn't mean the blame is justified or that the information tells you what you think it does.

And when blaming receivers for all interceptions, it more than likely isn't and doesn't. I've only watched the interceptions targeting Wayne so far, but it's certainly possible that one or more of Johnson, Fitzgerald, or Marshall were personally responsible for more interceptions than Wayne was. It wouldn't take much.

I wouldn't be so quick to claim that picture two is either clearer or better.

 

[percyhoward]

Yards per Reception, 2009
Miles Austin 16.3
Roy Williams 15.7

Yards per Target, 2009
Miles Austin 11.0
Roy Williams 7.1

Every incompletion is a wasted down.

Which is the clearer, better picture?

How much of the difference is attributable to Romo?

 

[NinePointOh]

Yards per Reception, 2009
Miles Austin 16.3
Roy Williams 15.7

Yards per Target, 2009
Miles Austin 11.0
Roy Williams 7.1

Every incompletion is a wasted down.

Which is the clearer, better picture?

How much of the difference is attributable to Romo?

Wait, why are you changing the subject now?

I was saying the inclusion of interceptions isn't very useful, not yards per target. That stat does have its own weaknesses, but they aren't nearly as problematic as blaming receivers for every interception.

 

[percyhoward]

The only reason we were talking about interceptions is because you contend that interceptions are primarily the fault of the QB. But then, that must true for every WR targeted anyway, not just the ones with high interception rates. It must also be true with incompletions.

If Miles catches 69% of the passes targeted to him, and Roy catches 45%, and the same QB was targeting both WR's, you could conclude that the difference probably had more to do with the receiver than the passer. From the raw list, you could then examine each player and even each play to see if it checked out, but it's a safe general assumption. That is, without ending up writing an entire book!

To large degree, the blaming is in the eye of the beholder, and we could go on back and forth. To see that the list is thought-provoking is a good thing.

 

[NinePointOh]

The only reason we were talking about interceptions is because you contend that interceptions are primarily the fault of the QB. But then, that must true for every WR targeted anyway, not just the ones with high interception rates.

Exactly. Which means the difference between a receiver who has a high interception rate and a receiver who has a low interception rate is far more likely to be related to the quarterback than the "riskiness" of the receiver. That's why it doesn't make much sense to differentiate receivers that way (and especially not to weight that factor exactly as heavily as you'd weight it for a quarterback), particularly receivers on two different teams.

It must also be true with incompletions.

It is true, but to a far lesser extent. Your equating the two doesn't make much sense, though. A completion requires the receiver to be involved. An interception usually doesn't involve him at all, which is precisely the issue -- the throw is usually off the mark, or tipped before it reaches him, or thrown when it shouldn't have been.

If Miles catches 69% of the passes targeted to him, and Roy catches 45%, and the same QB was targeting both WR's, you could conclude that the difference probably had more to do with the receiver than the passer.

The problem with interceptions is far bigger than the problem with completion percentage, but what you're saying here isn't really true, either. Some receivers are asked to run higher percentage routes than other receivers. Some receivers are covered by better defenders than other receivers. And then there's just plain random variation in any probabilistic event. Miles vs. Roy might be a clear cut example, but most comparisons won't be so obvious.

But more importantly, that point is only applicable as long as you're holding the passer constant. When you begin to compare receivers on different teams, however (as you do in the original post), the argument breaks down. Roddy White caught just 56% of the passes where he was targeted, and Marques Colston caught 69%. Is that all about the receiver, or might it have something to do with the fact that Matt Ryan completed 58% of all his passes while Drew Brees completed 70%?

From the raw list, you could then examine each player and even each play to see if it checked out, but it's a safe general assumption. That is, without ending up writing an entire book!

It's an absolutely unsafe assumption unless you control for the passer, and even then you wouldn't eliminate the problem entirely.

To large degree, the blaming is in the eye of the beholder, and we could go on back and forth.

I'd wager that almost any knowledgeable football observer would conclude that interceptions have a lot more to do with the quarterback than the receiver. The question, then, is why would you weight interceptions equally as heavily for a receiver as you would for a quarterback? I suspect that the reason is to achieve the parallel to the QB rating, but in that case the similarity is just as forced as any of Peyton Manning's throws to Reggie Wayne.

To see that the list is thought-provoking is a good thing.

Good, I'm glad.

 

[baj1dallas]

I predict that the top 5 all fall below #5 in the ranking next year, although all will stay in the top 15.

 

[Cowboy Brian]

I predict that the top 5 all fall below #5 in the ranking next year, although all will stay in the top 15.

#1 will stay exactly where he is. The rest doesnt matter 1 bit