Writers of Pro Football Prospectus 2008

23 Dec 2009

does adjusting a WR's stats based on his team's total pass-attempts favor certain systems?

I don't know if it's their idea originally, but PFR uses a metric known as Adjusted Catch Yards (ACY) for pass-catching stats. (SKIP TO SECOND PARAGRAPH IF YOU KNOW ABOUT ACY ALREADY) It consists of receiving yards + (20*recTD) + (6*receptions) -- to give value to guys who may be red-zone "specialists" (or boost guys who are good in yards AND get TDs) and the classic "possession receiver". What they also do is, instead of using yards/catch, they use ACY/team-attempts (only for those games the WR participated in of course). The logic is to give more credit to guys who play on run-first teams.

I think it's a cool stat regardless, as it gives more depth than just each of those stats gives alone. Originally I thought it was cool that it gave some statistical "proof" to the notion CAR's Steven Smith was even better than his good #s showed, because they ran the ball a lot(and not to mention Jake Delhomme as QB :p). However based on 2008 ACY/att, Steve Smith was far and away the leader -- it wasn't even close.

The top 30 WR for ACY/att in 2008 can be found on this PFR blog post
http://www.pro-football-reference.com/blog/?p=1345

All of the top names are commonly thought of as the best WRs, but I question why Larry Fitzgerald, despite being 5th, is considerably lower than Steve Smith. You could argue that he's not as good as his hyped image. You could say that the stat is a reflection of production, not talent (though I would argue that we try to use stats to quantify how good a player is, and that I am wondering if this particular one could do a better job).

To zero in on my point here: Does ACY/att favor guys in high-run/low-pass systems too much? (ie: more than the "evening out" it's supposed to give them).

In the games those 2 WR played in, CAR threw 362 passes, to AZ's 630. In run heavy offenses, if the team can snag a WR that could be called a "star", they generally lean heavily on him when they occasionally do throw the ball. Offenses like AZ's, which pass much more & often out of 3&4 WR formations, generally have 2 or more "quality" WRs (out of necessity for the system, really) and because they bring 3+ on the field, they "spread it around" considerably more than the run-teams.

My gut feeling (and so I ask to see if there would be a way to use statistics to sort this out, or if it just has to be thought out) is that throwing more passes already is a disadvantage for ACY/att, and the fact that such a system almost dictates you spread it out (and that a run-system often results in leaning on 1 WR) creates an even bigger disadvantage to a WR when being measured by ACY/att.

For example, of those 362 attempts by CAR, Smith was targeted 123 times, or ~33.98% of them. Of AZ's 630 attempts, Fitzgerald was targeted 146 times, or 23.17%. If Smith played in a spread-'em-out offense though, I do not think he would have the same % of balls his way (it would be 214 targets of 630 passes. Marvin Harrison saw 34.69% of IND's 591 '02 attempts, but that was a historic year). In other words, there's a limit to how many balls a guy is going to see.

On top of the system, surrounding talent has got to play a part in ball distribution, as Steve Smith has always been much more talented than whoever that place across from him -- meaning they'll be forced to go his way more, while Larry plays with fellow Pro-Bowler Boldin, and a very good Steve Breaston. (This is all in addition to a 3-4WR offense requiring you have depth at receiver, as not all teams that have run such an offense, has had as good of duos as Boldin & Fitz).

Personally I hope I have some validity here as my own Chargers' Hall of Fame WR Charlie Joiner does not grade out particularly well in this metric. That is, I hope he can use the fact that he played on the field with Kellen Winslow (the favorite target at least in terms of rec/game) and John Jefferson/Wes Chandler, very good reasons to spread it around, as somewhat of an excuse.

What is everybody's thoughts on this notion?

Posted by: lightsout85 on 23 Dec 2009

0 replies

Login or register to post comments