The Statistics (and Physics) of Wide Receivers in the NFL (Pt. 2)

What did I find when evaluating Dr. Smith’s data on NFL receivers?

First a little about the data itself. It consists of performance data of  495 wide receivers in the NFL from 1999 - 2012.  It has information on the following performance characteristics from each player’s post college evaluation and NFL career:

Sammy Watkins runs the 40 yard dash in 4.34 seconds in the 2014 scouting combine (USAToday)
Watkins will play with the Buffalo Bills against the Miami Dolphins

Pre-career data from Scouting Combine or Pro Day
* Year drafted
* Player’s name
* College
* Height in inches
* Weight in lbs
* Wonderlic scores (very limited sample)
* 40yardsec (40 yd dash times from Pro Day)
* Benchpress (limited sample)
*Vertical leap in inches
*Broad jump in inches
*Shuttle time
* Cone time
            
Table 1
Characteristics of NFL Wide Receivers Evaluated at Combine/Pro-Day: 1999 - 2012

NFL Performance Indices
* Games (g)
* Games started (gs)
* Receptions (rec)
* Yards (yd)
* Yards per reception (yr)
* Touchdowns (td)
* Long reception (lng)
* Receptions per game (rg)

Table 2
Performance of NFL Wide Receivers: 1999 - 2012
    Variable |                  Obs        Mean              Std. Dev.                      Min        Max
-----------------+---------------------------------------------------------------------------

After reviewing Dr. Smith’s data, I then created some performance indices of my own.  In particular, I was interested in some  statistics that indicated scoring potential:

* TD per game (tdg) Obvious importance
* Yards per game (Ydg) set up runs for TD s and field goals)
* RECg (receptions per game)

The first two statistics are of extreme importance. The third gives an indication of productivity and potentially of ability to help one's team dominate the clock.

I also wanted to translate weight and 40 yard dash times into metric units  for reasons that will shortly become obvious.

* Speed (meters per sec or ms, as calculated from 40 yard dash time)
* Mass (weight in kilograms)

Table 3
Additional Performance Indexes of NFL Wide Receivers: 1999 - 2012

Additionally Dr. Smith designated 18 receivers in his database as “elite.”  He classified these by having at least four receptions in each game and one touchdown scored in every three games for a total of 32 games. Here is a summary of Smith’s elite receivers:
   

Table 4
 Elite NFL Wide Receivers: 1999 - 2012: Characteristics & Performance

There is plenty of well-considered evidence that the 40 dash time matters a lot to successful receivers. This is a nice example:

What-a-strong-40-yard-dash-means-for-elite-wide-receivers

What new can we add from further evaluation of Dr. Smith’s data?

Let's start with a non-sequitur.

I did a t-test of means of the elite receivers vs. the other receivers who play more than a single game. Get this:  it revealed essentially no difference at all in their average time at the 40 yard dash!

Okay, I was happy to forget college statistics, what does all that mean? 

That there is about a 64% chance that there is effectively no difference in the speed of elite receivers and non-elite receivers.  Not statistically significant.

Elite receivers are not necessarily faster. What's with that?

So, if it is not speed that separates elite receivers from those less successful, then what is it?

In the next column, I’ll show that speed is important, but not by itself. There is a large interaction that may surprise you. And together with speed, it does make a difference.
(No, it is not the quarterback or college...)

Next time...