In this week's Varsity Numbers, Bill Connelly revisits some measures and concepts: Adjusted Scores, Covariance, and momentum (or whatever you choose to call it).
20 Sep 2013
by Bill Connelly
After a wonderful Week 3 in the college football season, we've got a bit of a slow Saturday for Week 4. Friday night's Boise State-Fresno State game should be intriguing, and Stanford-Arizona State could define a good portion of the Pac-12 race, but most highly ranked teams are playing either FCS opponents or semi-competitive mid-majors (or, in LSU's case, Auburn), and there aren't nearly as many marquee story lines at play.
With that in mind, I thought it would be a good time to tackle something I intended to do in the offseason: receiver profiling. Marty Couvillan of cfbstats.com opened up his database last year and shares updated 2013 data on a weekly basis; part of that data is general height/weight/position/experience data that I've been meaning to play with for quite a while. So let's do that.
Below is the most general of general data. It is an attempt to profile receivers and tight ends based on height, weight, and experience. It isn't a shot at pronouncing that 6-foot-2 receivers are better than 6-foot-4 receivers; it's just a way for you to understand the baseline at play when your team has a receiver of a certain type. Sounds fun, yes?
(Note: The data below represents any FBS player who caught at least one pass in 2012.)
There were 912 players who caught a pass in 2012 and were listed as a wide receiver (or split end, or flanker). One-third of them were between 6-foot-0 and 6-foot-2, and another sixth of them were either 5-foot-11 or 6-foot-3. In terms of weight, 20 percent of them were under 180 pounds, 47 percent were between 180 and 199 pounds, 28 percent were between 200 and 219, and five percent were 220 or bigger. None of this should be a surprise. But let's start generalizing.
The basic components of any receiver's stats are catch rate and yards per catch. In general, receivers have a catch rate around 62 percent and average between 12 and 13 yards per catch. But what if we break that out into ranges of height and weight?
|Ht Range||Catch Rate||Yds/Catch||Yds/Target||Pct of Total WRs|
|5'5" to 5'9"||64.2%||10.4||6.6||12.2%|
|5'10" to 6'4"||61.8%||12.6||7.8||84.1%|
|6'5" to 6'8"||57.3%||14.0||8.0||3.7%|
Smaller receivers are often used in slotback roles or, in the case of somebody like Ace Sanders, frequently lined up wide and fed short passes. The result is that, on average, these receivers have a higher catch rate but lower yards per catch. A lot of them are used as extensions of the running game.
Meanwhile, taller receivers are more likely of the "get downfield and post up" variety; quarterbacks send them lower-percentage passes with higher potential. In terms of 2013 draftees, a good example of this type of player might be Tennessee's Justin Hunter; he's only 6-foot-4, so he technically falls into the middle range, but he caught just 55 percent of his passes in 2012 but averaged 14.8 yards per catch.
Most receivers, however, are in the middle range. This is an enormous range, encompassing 85 percent of all receivers, but it was difficult to distinguish between them. Receivers who stand 5-foot-10 averaged a 63 percent catch rate and 12.3 yards per catch, 6-foot-1 receivers averaged 62 percent and 12.6, and 6-foot-3 receivers averaged 61 percent and 11.8. They all fall pretty close to this "62 percent and 12.6 per catch" range.
You can probably guess, then, that weight ranges will probably follow a similar trend since, you know, small players are likely to weight less and big players are likely to weight more.
|Wt Range||Catch Rate||Yds/Catch||Yds/Target||Pct of Total WRs|
Though we tend to equate "bigger" with "slower" a receiver with good size is more likely to be able to establish good position to catch longer passes. It could be hard to lob a pass into Tavon Austin, right? Even if he beats his man, the corner might still be able to stay close enough to bat a ball down before it reaches him. A smaller receiver must either establish a lot of space between himself or possess an unholy vertical leap to make up the difference. Great ones do, but average ones are not trusted to do so.
From the tight end position, the height distribution was minimal. Only eight of the 312 tight ends who caught a pass in 2012 (2.6 percent) were shorter than 6-foot-2, and only 11 (3.5 percent) were taller than 6-foot-6. A solid 28 percent were either 6-foot-2 or 6-foot-3, 51 percent were 6-foot-4 or 6-foot-5, and 14 percent were 6-foot-6. From a weight range, nine percent were lighter than 230 pounds, 48 percent were between 230 and 249, 41 percent were between 250 and 269, and three percent were 270 or heavier. Again, this all likely fits with your preconceptions for the position.
So are tight ends used differently based on height or weight?
|Ht Range||Catch Rate||Yds/Catch||Yds/Target||Pct of Total TEs|
|6'0" to 6'2"||62.9%||9.8||6.2||13.1%|
|6'4" to 6'6"||64.7%||10.6||6.8||65.5%|
|6'7" to 6'8"||51.3%||12.9||6.6||3.5%|
When I get the opportunity, I want to run the data for a bigger sample than just the 2012 season. I assume that when I do so, I won't find a magical sweet spot at 6-foot-3, where players are quite a bit better than those one inch taller or shorter. But with this single year of data, 6-foot-3 was the height to be. Meanwhile, one could say that there really might be a defined disadvantage in being a shorter tight end.
|Wt Range||Catch Rate||Yds/Catch||Yds/Target||Pct of Total TEs|
From a weight perspective, there are almost no distinguishing characteristics. Really big tight ends are probably blockers first and don't necessarily have the hands that receiver-first types in the 230-250 range probably do. But on average, anybody under 270 seems to average about the same on a per-target basis.
I also wanted to look at how teams use different positions in the passing game. Here's a look at the position of teams' No. 1 targets, No. 2 targets, etc., in 2012. (This type of data is damaged a bit by injuries, of course. If a team's go-to receiver gets hurt in the fourth game and ends up with the fifth-most targets on the, he's the No. 5 target. This is a "perfect vs. good" situation here.)
|Target No.||WR||RB||TE/HB||FB||Grand Total|
Last exercise: Let's break these samples out by year of experience. Now, obviously the best way to do this would be to track the same pool of players from freshman to senior season -- this is looking at players from each class from a single season, so the odds are good that only the best freshmen will play.
|WR||Catch Rate||Yds/Catch||Yds/Target||Pct of Total|
It's difficult to draw too much from this, other than the fact that older receivers seem to possibly improve more from a yards-per-catch standpoint than from a catch rate standpoint. Your hands are your hands no matter how experienced you are, I guess.
|TE||Catch Rate||Yds/Catch||Yds/Target||Pct of Total|
Almost no distinguishable characteristics here.
What about if we only look at the most leaned-on players? Here's the same data looking only at players who were among the top three targets on their team. (That means the tight end sample is pretty small).
|WR (top-3 target)||Catch Rate||Yds/Catch||Yds/Target||Pct of Total|
Again, there is no difference in catch rate, but the per-catch yardage improves a bit.
|TE (top-3 target)||Catch Rate||Yds/Catch||Yds/Target||Pct of Total|
As rare as it is to find a go-to tight end, it's even rarer to find one young than a junior. But if you're throwing to one, they're probably pretty good. The only freshman tight ends in the sample: Memphis' Alan Cross (26 targets, 23 catches, 301 yards), Penn State's Kyle Carter (52 targets, 36 catches, 453 yards), and UNLV's Jake Phillips (45 targets, 23 catches, 226 yards). Phillips was underwhelming, but the other two were quite successful.
4 comments, Last at 29 Sep 2013, 10:56am by Jocuri Gratuite