SB Nation - Blake May

Targets and Catches (Part Two, a.k.a. Will Franklin's Redemption)

Bill C. — Wed, 09 Dec 2009 18:00:17 -0000

Image via The Trib's G.J. McCarthy

You may have noticed that, somewhere around 1pm yesterday, the title of yesterday's "Targets and Catches" post changed to "Targets and Catches (Part One)." That's because I figured out the next direction I wanted to take the concept of targets and couldn't wait to get started on it.

As discussed yesterday, the concept of targets (catches + incomplete passes intended for a given receiver) versus catches is a pretty good one, despite the likely flaws in the ways target data is collected (dropped/tipped passes that become interceptions don't have target data attached, not every scorekeeper probably records targets the same way, etc.). It allows us to look into a receiver's reliability in ways that simpler measures do not, and once national data is collected on this (offseason project #12,424), it might give us a way to measure a quarterback's performance independent of the quality of his receivers (hey Uribe Auction, here's how we get to Receiver Independent Passing Stats, or RIPS...I love it when a plan comes together).

But simply looking at a players Catch% (Catches per Target) does a disservice to those who are likely to run deeper, lower-percentage routes. How about we do this as well: create a Yards Per Target measure. Yards Per Catch is a good way to measure a receiver's explosiveness, and wouldn't the incorporation of target data create something that shows us both explosiveness and reliability? Let's try that.

Below are tables for 2006-09 giving us the following information:

Catches
Yards
Catch% (Catches divided by Targets)
Yards Per Target

2006

2006 Yards Per Target
Name	Pos.	Catches	Yards	Catch%	Yds/Target
Chase Coffman	TE	58	638	65.2%	7.2
Martin Rucker	TE	53	511	77.9%	7.5
Will Franklin	WR	48	829	57.8%	10.0
Jared Perry	WR	37	429	67.3%	7.8
Brad Ekwerekwu	WR	32	430	64.0%	8.6
Tommy Saunders	WR	25	271	73.5%	8.0
Danario Alexander	WR	15	251	71.4%	12.0
Jason Ray	WR	6	50	40.0%	3.3
Tony Temple	RB	6	34	85.7%	4.9
Earl Goldsmith	RB	5	59	100.0%	11.8
Greg Bracey	WR	3	79	50.0%	13.2
Marcus Woods	RB	2	11	40.0%	2.2
Jimmy Jackson	RB	1	-2	50.0%	-1.0
Andrew Hoskins	WR	0	0	0.0%	0.0
TOTAL	ALL	291	3,590	66.0%	8.1

Here's where we start to see the value of Will Franklin. The 'Copta had an awful catch rate this season, but as we see here, a) he made up for it with his explosiveness, and b) being a deep threat means you're probably seeing more low-percentage passes. Meanwhile, tight ends running high-percentage routes will have good catch rates, but their Yards Per Target figures show that while it's good to have high-percentage targets, there isn't necessarily a lot of explosiveness there.

Two other interesting things to note:

1) Danario was very clearly showing his big-play potential early on. It really is a damn shame that he ended up having to fight so many injuries, but on the bright side, if he hadn't, he'd have almost certainly gone pro last year.

2) Clearly we should have been passing to Earl Goldsmith more often. Ahem.

2007

Image via the inestimable Bill Carter

2007 Yards Per Target
Name	Pos.	Catches	Yards	Catch%	Yds/Target
Martin Rucker	TE	84	834	77.8%	7.7
Jeremy Maclin	WR	80	1,055	72.1%	9.5
Chase Coffman	TE	52	531	80.0%	8.2
Will Franklin	WR	49	709	64.5%	9.3
Tommy Saunders	WR	41	397	74.5%	7.2
Danario Alexander	WR	37	417	77.1%	8.7
Jared Perry	WR	13	152	65.0%	7.6
Derrick Washington	RB	10	70	71.4%	5.0
Jason Ray	WR	8	61	80.0%	6.1
Tony Temple	RB	7	68	77.8%	7.6
Jimmy Jackson	RB	6	30	100.0%	5.0
Earl Goldsmith	RB	4	20	66.7%	3.3
Greg Bracey	WR	3	53	33.3%	5.9
Blake May	WR	0	0	0.0%	0.0
TOTAL	ALL	394	4,397	73.2%	8.2

And here's where the numbers even out for Senor Goldsmith.

Even though his YPT dipped in 2007, Franklin was still a decent weapon here, averaging over 9.0 per target (which appears to be a pretty good number) and, as we must not forget, becoming one helluva downfield blocker for Maclin. In all, Chase Coffman was the most reliable target this season, but obviously the WRs were more dangerous, even though they didn't really thrust too many nunchuks.

(Non-RMN regulars just read and re-read that last sentence 5-6 times, then gave up on figuring it out what it meant. Here's the explanation.)

2008

Bill Carter again

2008 Yards Per Target
Name	Pos.	Catches	Yards	Catch%	Yds/Target
Jeremy Maclin	WR	102	1,260	75.6%	9.3
Chase Coffman	TE	90	987	88.2%	9.7
Tommy Saunders	WR	72	833	78.3%	9.1
Jared Perry	WR	41	567	68.3%	9.5
Derrick Washington	RB	29	277	76.3%	7.3
Danario Alexander	WR	26	329	60.5%	7.7
Andrew Jones	TE	20	146	83.3%	6.1
Jerrell Jackson	WR	9	98	69.2%	7.5
Earl Goldsmith	WR	5	58	83.3%	9.7
Michael Egnew	TE	4	22	50.0%	2.8
Jimmy Jackson	RB	3	26	75.0%	6.5
Wes Kemp	WR	1	15	50.0%	7.5
Jon Gissinger	TE	1	10	100.0%	10.0
Brandon Gerau	WR	0	0	0.0%	0.0
Forrest Shock	WR	0	0	0.0%	0.0
TOTAL	ALL	403	4,628	76.0%	8.7

Here's where Chase Coffman became the best tight end we'll ever see. He was so reliable that throwing to him actually resulted in more yards per target than Jeremy Maclin, All-American. Here's also where Jared Perry, a.k.a. Will Franklin 2.0, began to show value. He was a reasonably reliable option his first couple of years (obviously targeted much more in 2006 than 2007, when Maclin emerged), but while he still couldn't hit 70% from a Catch% perspective, his big-play ability went up a notch, and he went from averaging 7-8 yards per target to over 9. Meanwhile, Jerrell Jackson registered with numbers very similar to Perry's freshman numbers, albeit on a small sample size.

2009

Yup, Bill Carter.

2009 Yards Per Target
Name	Pos.	Catches	Yards	Catch%	Yds/Target
Danario Alexander	WR	107	1,644	73.8%	11.3
Jared Perry	WR	46	695	53.5%	8.1
Jerrell Jackson	WR	37	458	60.7%	7.5
Wes Kemp	WR	23	335	52.3%	7.6
Derrick Washington	RB	23	99	69.7%	3.0
Andrew Jones	TE	7	39	58.3%	3.3
Rolandis Woodland	WR	4	23	66.7%	3.8
Michael Egnew	TE	3	25	60.0%	5.0
De'Vion Moore	RB	3	13	100.0%	4.3
Kendial Lawrence	RB	2	42	100.0%	21.0
Brandon Gerau	WR	2	19	66.7%	6.3
T.J. Moe	WR	2	8	40.0%	1.6
Gilbert Moye	RB	1	5	50.0%	2.5
TOTAL	ALL	260	3,405	63.9%	8.4

And here's where we find out just how amazing Danario Alexander has been this season (as if there were any doubt). Looking only at receivers who were targeted at least ten times in a given season, Alexander's 11.3 per-target average is topped only by ... 2006 Danario Alexander, who was targeted only 21 times. For ~~Domino Axelrod~~ Danario to be both this reliable and this explosive has just been amazing.

Here are the overall averages per position:

WR: 67.3% catch rate, 8.7 yards per target
TE: 77.2% catch rate, 8.2 yards per target
RB: 75.0% catch rate, 5.5 yards per target

So for the last four seasons, here are your top ten in terms of Yards Per Target (with a minimum of ten targets):

Danario Alexander (2006): 12.0 per target (on 21 targets)
Danario Alexander (2009): 11.3 (145)
Will Franklin (2006): 10.0 (83)
Chase Coffman (2008): 9.7 (102)
Jeremy Maclin (2007): 9.5 (111)
Jared Perry (2008): 9.5 (60)
Jeremy Maclin (2008): 9.3 (135)
Will Franklin (2007): 9.3 (76)
Tommy Saunders (2008): 9.1 (92)
Danario Alexander (2007): 8.7 (48)

Will Franklin, you are absolved.

Danario Alexander, you officially have a case for having been the best receiver in Mizzou history.

Targets and Catches (Part One)

Bill C. — Tue, 08 Dec 2009 18:00:15 -0000

One of the measures I am looking to track more in-depth in the future, in terms of my play-by-play data, is the notion of catches versus targets. In the play-by-play data exists a line that I have yet allowed to go unrecorded: "Pass incomplete to _____." Obviously I record that the pass is incomplete, but I have not really paid attention to who the pass was intended for. It's a rather objective measure -- does everybody record who a pass was intended for the same way? If the QB is throwing the ball away and it goes 20 feet over a certain receiver's head, does it count as an incompletion "to" that receiver? Does every score-keeper even keep track of this? Plus, if a ball is tipped by a receiver and intercepted (e.g. Jerrell Jackson in the Oklahoma State game), the "intended for" bit isn't recorded. Clearly this is a piece of data that is rather inconsistent.

That said, this is also one of those measures where you figure that the inconsistencies balance out over time. Even if it doesn't paint a perfect picture, it probably tells an interesting story. Therefore, let's take a look at how our view of a receiver's (or quarterback's) performance changes when adding in "intended for" data.

The categories below are:

Catches: receptions
Targets: receptions + incomplete passes intended for that receiver (as recorded in the play-by-play data). In general, you can think of those incomplete passes as either being a) dropped or b) too well-covered to be completed. If your QB is accurate, you will obviously catch more of your targeted passes, but if you struggle to get open and clear space from the guy covering you, a lot more of your targeted passes are going to fall incomplete.
Catch%: the percentage of "targets" a receiver catches
Target%: the percentage of a team's overall targets aimed at a specific receiver

Without further ado, let's take a look at the last four years of Mizzou "target" data and see what it tells us.

2006

Image via Phoebe Sexton / Columbia Missourian

2006 Targets and Catches
Name	Catches	Targets	Catch%	Target%
Chase Coffman	58	89	65.2%	20.2%
Will Franklin	48	83	57.8%	18.8%
Martin Rucker	53	68	77.9%	15.4%
Jared Perry	37	55	67.3%	12.5%
Brad Ekwerekwu	32	50	64.0%	11.3%
Tommy Saunders	25	34	73.5%	7.7%
Danario Alexander	15	21	71.4%	4.8%
Jason Ray	6	15	40.0%	3.4%
Tony Temple	6	7	85.7%	1.6%
Greg Bracey	3	6	50.0%	1.4%
Marcus Woods	2	5	40.0%	1.1%
Earl Goldsmith	5	5	100.0%	1.1%
Jimmy Jackson	1	2	50.0%	0.5%
Andrew Hoskins	0	1	0.0%	0.2%
TOTAL	291	441	66.0%	100.0%

WRs: 60.1% of targets
TEs: 35.6%
RBs: 4.3%

Before looking at ensuing seasons, it is hard to get a grasp on what these numbers actually mean, but a couple of things should stick out pretty quickly. First, thanks to a midseason bout with the dropsies, Chase Coffman's numbers were only decent, not great. Martin Rucker was a rather automatic target at TE, but Coffman wasn't amazing. Meanwhile, if your #1 WR is only catching 58% of the passes thrown at him, that is going to do a number on your completion percentage.

Anyway, before we try to interpret this too much, let's take a look at how 2007 unfolded.

2007

Image via ESPN.com

2007 Targets and Catches
Name	Catches	Targets	Catch%	Target%
Jeremy Maclin	80	111	72.1%	20.6%
Martin Rucker	84	108	77.8%	20.1%
Will Franklin	49	76	64.5%	14.1%
Chase Coffman	52	65	80.0%	12.1%
Tommy Saunders	41	55	74.5%	10.2%
Danario Alexander	37	48	77.1%	8.9%
Jared Perry	13	20	65.0%	3.7%
Derrick Washington	10	14	71.4%	2.6%
Jason Ray	8	10	80.0%	1.9%
Greg Bracey	3	9	33.3%	1.7%
Tony Temple	7	9	77.8%	1.7%
Earl Goldsmith	4	6	66.7%	1.1%
Jimmy Jackson	6	6	100.0%	1.1%
Blake May	0	1	0.0%	0.2%
TOTAL	394	538	73.2%	100.0%

WRs: 63.9% of targets
TEs: 32.2%
RBs: 3.9%

Despite a deep, experienced cast of receivers (Rucker and Franklin were seniors, Coffman and Saunders were juniors), Jeremy Maclin came out of nowhere to become Chase Daniel's most trusted receiver. His 72% Catch% was much higher than Will Franklin's 58% from the season before. Meanwhile, MotherRucker stayed right around 78% while Chase Coffman improved dramatically, from 65% to 80%. Franklin did improve to 65%, but he was still a far less reliable target than others around him. Of course, he obviously ran deeper routes than guys like Coffman, Rucker or Saunders, so that might have something to do with it. Then again, were his routes any different than those of Maclin? Maclin still outshone him in that regard.

With Maclin and an extra year of experience for Chase Daniel, the overall Catch% rose from 66% to 73% in 2007. This, of course, tells us nothing about how many passes Daniel chose to throw away (which would just go in the play-by-play as "pass incomplete" with no "intended for" information), but it does show that Daniel was completing one more pass for roughly every 14 catchable balls in 2007, and that could make a difference of 2-3 completions per game.

2008

Image via Chris Lee / Post-Dispatch

2008 Targets and Catches
Name	Catches	Targets	Catch%	Target%
Jeremy Maclin	102	135	75.6%	25.5%
Chase Coffman	90	102	88.2%	19.2%
Tommy Saunders	72	92	78.3%	17.4%
Jared Perry	41	60	68.3%	11.3%
Danario Alexander	26	43	60.5%	8.1%
Derrick Washington	29	38	76.3%	7.2%
Andrew Jones	20	24	83.3%	4.5%
Jerrell Jackson	9	13	69.2%	2.5%
Michael Egnew	4	8	50.0%	1.5%
Earl Goldsmith	5	6	83.3%	1.1%
Jimmy Jackson	3	4	75.0%	0.8%
Wes Kemp	1	2	50.0%	0.4%
Jon Gissinger	1	1	100.0%	0.2%
Brandon Gerau	0	1	0.0%	0.2%
Forrest Shock	0	1	0.0%	0.2%
TOTAL	403	530	76.0%	100.0%

WRs: 66.6% of targets
TEs: 25.5%
RBs: 7.9%

In 2008, Missouri lost Martin Rucker and his 78% catch rate but still improved as a team, from 73% to 76%. Danario Alexander aside (this was the year he was recovering from knee injuries and was clearly not as confident/aggressive a receiver), just about everybody showed improvement. Coffman improved to a sickening 88% (that means that eight of every nine balls that were thrown at him and were even remotely catchable, were caught...yeah, he deserved the Mackey Award), Saunders improved to 78%, which is the highest you'll see from a highly-targeted receiver, and even Maclin improved to 76%. Having a ton of reps with the same quarterback helped a ton, I'm sure, as did the fact that said quarterback (Daniel) was as automatic as a computer in his reads, especially over the first half of the season.

So what happens when you lose your quarterback and your top three targets? As shocking as it may sound, your numbers go down!

2009

Image via the one, the only Bill Carter

2009 Targets and Catches
Name	Catches	Targets	Catch%	Target%
Danario Alexander	107	145	73.8%	35.6%
Jared Perry	46	86	53.5%	21.1%
Jerrell Jackson	37	61	60.7%	15.0%
Wes Kemp	23	44	52.3%	10.8%
Derrick Washington	23	33	69.7%	8.1%
Andrew Jones	7	12	58.3%	2.9%
Rolandis Woodland	4	6	66.7%	1.5%
Michael Egnew	3	5	60.0%	1.2%
T.J. Moe	2	5	40.0%	1.2%
De'Vion Moore	3	3	100.0%	0.7%
Brandon Gerau	2	3	66.7%	0.7%
Kendial Lawrence	2	2	100.0%	0.5%
Gilbert Moye	1	2	50.0%	0.5%
TOTAL	260	407	63.9%	100.0%

WRs: 86.0% of targets
TEs: 4.2%
RBs: 9.8%

So here's where I start to get encouraged about Blaine Gabbert's prospects moving forward. With a relatively similar cast of characters in the receiving corps (Alexander was clearly better than any receiver in 2006, but Gabbert was without anyone nearing the reliability of Rucker, so they cancel out) and the propensity for throwing many more deep balls than Daniel, the Catch% for Gabbert's sophomore season was almost identical to Daniel's.

Now, unless one of our redshirting freshmen is suddenly capable of pulling a Jeremy Maclin impersonation (doubtful, to say the very least), Gabbert will head into 2010 without his top two targets, which is obviously something Daniel didn't have to worry about heading into 2007. However, while losing Alexander is a huge concern, losing Perry simply might not be too big a deal. Jackson has done a lovely Perry impersonation in recent weeks, and besides, Perry's 54% catch rate is simply horrendous. We know that Perry had a decent amount of bad luck this season -- how many times did Gabbert throw too high for him in the Oklahoma State game? -- but 54% is too low to blame simply on bad luck. Over four seasons, Perry never topped 70%, and while Jackson's 61% is nothing impressive, it is still better than what Perry managed this season. With another year (at least) of getting used to catching Gabbert's fastballs (and another year of Gabbert learning to keep the ball down as much as possible), I would expect Jackson's catch rate to be closer to the 65-70% range next year.

However, is that good enough? Gabbert can be expected to improve a decent amount next year, but ... who becomes the #1 target? And if it's Jackson, can Mizzou succeed at a high level with a #1 WR target catching only 65-70%, especially when there is suddenly no ultra-reliable tight end option to fall back on in times where a completion absolutely must be made?

Speaking of the tight end position, I guess this is the year in which we officially learned that WRs and TEs are basically exactly the same in this system. Alexander's Target% of 35% this year was 10% higher than Maclin ever managed and came within 10% of matching that of Maclin and Coffman from 2008. Lining up in the tight end slot position quite often, Alexander was actually able to approximately compensate for the loss of both players this year and catch passes at a very high rate for a receiver. Meanwhile, thanks to Alexander being a WR, tight ends accounted for only 4% of targets this year, and that's only if you count Michael Egnew as a tight end. The position disappeared from Mizzou's repertoire, but the passing numbers stayed relatively successful thanks to Alexander. With Alexander gone in 2010, do we see guys like Andrew Jones or Egnew working their way back into Gabbert's sights, or will another big receiver like Wes Kemp move into that role?

I have a lot more questions here than answers, but answers weren't the purpose of this exercise. I wanted to see if target data was viable and interesting, and while I found it both of those things, I'm curious what you think.