Quick hits on targets data (1 Viewer)

[Chase Stuart]

Here are some league-wide target stats for WRs the past six seasons:

Rec/Tar Yard/Tar Yar/Rec2007 58.1% 7.6 13.12006 55.9% 7.6 13.52005 56.0% 7.5 13.42004 58.5% 8.2 14.12003 56.7% 7.6 13.42002 58.6% 7.8 13.3avg 57.3% 7.7 13.5Here are the numbers for all WRs with a certain number of targets:

Rec/Tar	Yard/Tar	Yar/Rec150+	 58.4%	  8.0		 13.6100-149  57.9%	  7.9		 13.560-100   56.6%	  7.7		 13.6<60	  56.4%	  7.4		 13.0

Wide receivers that averaged over 10 yards per target, in any season, minimum 40 targets:

yar/tar	tar	name13.5	   55	Devery Henderson12.3	   49	Dennis Northcutt12.0	   70	Lee Evans11.9	   44	Kelly Campbell11.8	   59	Plaxico Burress11.7	   45	James McKnight11.3	   51	Anthony Gonzalez11.2	   75	Santana Moss11.1	   61	Joe Horn11.1	   85	Santonio Holmes11.1	  134	Santana Moss11.0	   84	Greg Jennings10.9	   99	Ashley Lelie10.9	   99	Brandon Stokley10.8	   43	Hank Baskett10.8	   48	Patrick Crayton10.7	   63	Todd Pinkston10.7	  113	Reggie Wayne10.6	  102	Eddie Kennison10.6	  129	Torry Holt10.5	   48	Drew Bennett10.5	   60	Reggie Williams10.5	   92	Quincy Morgan10.5	   76	Johnnie Morton10.4	  150	Steve Smith10.3	   98	Joey Galloway10.3	  116	Michael Clayton10.3	   98	Nate Burleson10.2	   70	Reggie Wayne10.2	  108	Eddie Kennison10.1	   68	Jerry Porter10.1	  133	Amani Toomer10.0	  138	Javon Walker10.0	   68	Ronald Curry

Correlation coefficient of yards per target in Year N, to yards per target in Year N+1, assuming at least 40 targets in each season: 0.12.Correlation coefficient of receptions per target in Year N, to receptions per target in Year N+1, assuming at least 40 targets in each season: 0.379.

Correlation coefficient of targets per game in Year N, to targets per game in Year N+1, assuming at least 40 targets in each season: 0.659.

Correlation coefficient of targets per game in Year N, to targets per game in Year N+1, assuming at least 80 targets in each season: 0.505.

Correlation coefficient of targets per game in Year N, to targets per game in Year N+1, assuming at least 100 targets in each season: 0.348.

 

[Tanner9919]

not sure we need to go to THAT extent for fantasy football and I have NO IDEA what you just said, but it sure looks impressive..

in English, you mean to say that the NFL is going to swing back to a RB-heavy league in 2008, just like it did in 2005 after the other 'year of the WR' , i.e., 2004?

Not many WR's will top 100 receptions this season is what you mean to imply...

I agree..Wes Welker won't be a top 20 WR in 2008, but Earnest Graham will have a great season and

someone will chase 2000 yards rushing and 8 or more RB's will score double digit TD's...

 

[Banger]

I'm not a stat geek but at what % is the correlation coefficient considered reliable? (if that question makes sense) If I'm reading the numbers correctly it seems that the only stat that has any reliability is the fact that a player that get 40+ targets is likely to get near that many the following season.

 

[Chase Stuart]

I'm not a stat geek but at what % is the correlation coefficient considered reliable? (if that question makes sense) If I'm reading the numbers correctly it seems that the only stat that has any reliability is the fact that a player that get 40+ targets is likely to get near that many the following season.

The reason the correlation coefficient is high for players with 40+ targets each year is probably because guys that get around 40-50 targets are mediocre, fringe players that will average about that the next season, too. And that's not very interesting for fantasy purposes. The more interesting question concerns targets for the stud players, and those aren't consistent from year to year.At no point is the CC "reliable" -- it's simply a sliding scale where the closer to 1 (or negative 1) the number is, the more reliable it is; the closer to 0, the less reliable.

Yards per target, group-wide, seems very unreliable. It's possible that deep threats, and slot WRs have more consistent YPT numbers from year to year (I'd suspect that they do, although only slightly more than the full sample), but in general, noting that a player has a very high or very low yards per target ratio probably has little predictive value.

Receptions per target has more predictive value. There is something there, although not that much.

 

[Biabreakable]

I'm not a stat geek but at what % is the correlation coefficient considered reliable? (if that question makes sense) If I'm reading the numbers correctly it seems that the only stat that has any reliability is the fact that a player that get 40+ targets is likely to get near that many the following season.

The reason the correlation coefficient is high for players with 40+ targets each year is probably because guys that get around 40-50 targets are mediocre, fringe players that will average about that the next season, too. And that's not very interesting for fantasy purposes. The more interesting question concerns targets for the stud players, and those aren't consistent from year to year.At no point is the CC "reliable" -- it's simply a sliding scale where the closer to 1 (or negative 1) the number is, the more reliable it is; the closer to 0, the less reliable.

Yards per target, group-wide, seems very unreliable. It's possible that deep threats, and slot WRs have more consistent YPT numbers from year to year (I'd suspect that they do, although only slightly more than the full sample), but in general, noting that a player has a very high or very low yards per target ratio probably has little predictive value.

Receptions per target has more predictive value. There is something there, although not that much.

Nice to see that you have come to this conclushion. I think the catch percentage for WRs is what your looking for coupled with targets of course.

The yards/target is really not as useful. And as the players listed above shows it is the RAC WR/Speed demons who benifit most from looking at it this way with very few exceptions.

The balanced WRs who get high targets will not fair as well because of the defensive attention they get and the possession catches they will make to move the chains.. not always being involved in big plays.

That Randy Moss is not even listed in the YPT just shows how insignificant it really is compared to other measurements.

 

[Chase Stuart]

I'm not a stat geek but at what % is the correlation coefficient considered reliable? (if that question makes sense) If I'm reading the numbers correctly it seems that the only stat that has any reliability is the fact that a player that get 40+ targets is likely to get near that many the following season.

The reason the correlation coefficient is high for players with 40+ targets each year is probably because guys that get around 40-50 targets are mediocre, fringe players that will average about that the next season, too. And that's not very interesting for fantasy purposes. The more interesting question concerns targets for the stud players, and those aren't consistent from year to year.At no point is the CC "reliable" -- it's simply a sliding scale where the closer to 1 (or negative 1) the number is, the more reliable it is; the closer to 0, the less reliable.

Yards per target, group-wide, seems very unreliable. It's possible that deep threats, and slot WRs have more consistent YPT numbers from year to year (I'd suspect that they do, although only slightly more than the full sample), but in general, noting that a player has a very high or very low yards per target ratio probably has little predictive value.

Receptions per target has more predictive value. There is something there, although not that much.

Nice to see that you have come to this conclushion. I think the catch percentage for WRs is what your looking for coupled with targets of course.

The yards/target is really not as useful. And as the players listed above shows it is the RAC WR/Speed demons who benifit most from looking at it this way with very few exceptions.

The balanced WRs who get high targets will not fair as well because of the defensive attention they get and the possession catches they will make to move the chains.. not always being involved in big plays.

That Randy Moss is not even listed in the YPT just shows how insignificant it really is compared to other measurements.

I don't think yards per target is very "useful", even though I'm not sure how I want to define useful. I used to think it was very "useful", but I've since done a 180 on that. It started when I began to recognize that a target that does not result in a reception isn't a negative for that wide receiver, and is arguably a positive. Thus, something like receiving yards X targets might be a better indicator of something that receiving yards / targets.Catch percentage is a bit trickier, because it's not entirely clear to me how valuable catches are. A catch for zero yards is not valuable at all, unless you're trying to kill the clock. And looking at catch percentage without checking out yards per reception seems silly. And that brings us back to the yards per target problem. (And even with receptions per target, you still have target in the denominator).

 

[Dinsy Ejotuz]

Hmmm... I wonder what the correlation coefficient is on a WRs completion percentage differential from his QBs completion percentage to WRs from season to season across, say, three years?

Year 1: WR catch % - QB completion % to WRs

Year 2: WR catch % - QB completion % to WRs

Year 3: WR catch % - QB completion % to WRs

For this to be a meaningful stat someone has to consistently deviate from the QB's completion percentage to WRs. I did this a couple years ago and IIRC got pretty much nowhere.

 

[Doug Drinen]

I don't think yards per target is very "useful", even though I'm not sure how I want to define useful. I used to think it was very "useful", but I've since done a 180 on that. It started when I began to recognize that a target that does not result in a reception isn't a negative for that wide receiver, and is arguably a positive. Thus, something like receiving yards X targets might be a better indicator of something that receiving yards / targets.

A couple of niche stats that I like to help me spot good WRs are Yards-per-target times targets, and also yards-per-catch times receptions.

 

[Dinsy Ejotuz]

I don't think yards per target is very "useful", even though I'm not sure how I want to define useful. I used to think it was very "useful", but I've since done a 180 on that. It started when I began to recognize that a target that does not result in a reception isn't a negative for that wide receiver, and is arguably a positive. Thus, something like receiving yards X targets might be a better indicator of something that receiving yards / targets.

Yards-per-catch times receptions.

Isn't that just yardage?

 

[Doug Drinen]

I don't think yards per target is very "useful", even though I'm not sure how I want to define useful. I used to think it was very "useful", but I've since done a 180 on that. It started when I began to recognize that a target that does not result in a reception isn't a negative for that wide receiver, and is arguably a positive. Thus, something like receiving yards X targets might be a better indicator of something that receiving yards / targets.

Yards-per-catch times receptions.

Isn't that just yardage?

 

[Dinsy Ejotuz]

I don't think yards per target is very "useful", even though I'm not sure how I want to define useful. I used to think it was very "useful", but I've since done a 180 on that. It started when I began to recognize that a target that does not result in a reception isn't a negative for that wide receiver, and is arguably a positive. Thus, something like receiving yards X targets might be a better indicator of something that receiving yards / targets.

Yards-per-catch times receptions.

Isn't that just yardage?

And somewhere my 6th grade math teacher is smiling.

 

[gianmarco]

I don't think yards per target is very "useful", even though I'm not sure how I want to define useful. I used to think it was very "useful", but I've since done a 180 on that. It started when I began to recognize that a target that does not result in a reception isn't a negative for that wide receiver, and is arguably a positive. Thus, something like receiving yards X targets might be a better indicator of something that receiving yards / targets.

Yards-per-catch times receptions.

Isn't that just yardage?

As is this:

Yards-per-target times targets