AUTHOR'S NOTE: Sorry for the delay in getting this one done. I've been moving into my new apartment for the past week, so that's kind of eaten up the bulk of my spare time. Hopefully, it's worth the wait. Oh, and, be prepared. This is by far the longest of the review posts. Don't worry, though. The remaining ones (OL, DF7, and secondary) will be about half as long.
Welcome back for the 7th installment of my 2009 Niner stat review. If you missed Parts 1-6, you can read them here, here, here, here, here, and here.
Today in Part 7, I'll take a look at the WR and TE positions. The obvious star of this group in 2009 was Vernon Davis, but Michael Crabtree's and Josh Morgan's 2009 seasons - and what they mean for the 49ers' near future -- are more intriguing to me as a discussion topic. One reason why it's more intriguing is because, whereas we just witnessed VD's breakout season, we're still uncertain about the if and when of Crabs' and/or Morgan's breakout. In addition, given what I've said thus far in this series about win improvements being related to pass OFF improvements, a breakout of Crabs and/or Morgan in 2010 would likely have a big impact on the 49ers' playoff chances.
So, today's post will go something like this. First, I'll present the efficiency and yardage stats for each receiver. Then, I'll try to give you an idea -statistically speaking, of course - about the likelihood of a Crabs outbreak in 2010 (and a Morgan one as well, of course...it's just that "Crabs and/or Morgan outbreak" doesn't sound as good as "Crabs outbreak"...if anything, Captain Morgan comes before the Crabs).
After the jump, an epidemic of Crabs and VD (and Morgan)...
First, some housekeeping. In this post, I'm going to do present the stats a little differently. Specifically, the rankings you'll see in the WR tables are each WR's ranking among NFL receivers on the same level of the depth chart. In other words, Crabs was the 49ers' #1 WR in 2009, so I'll show you his rankings among #1 WRs; not all WRs. Obviously, I'm doing this to compare apples to apples as much as possible. In a similar vein, stats and rankings for the average playoff team will also be specific to WRs at the same level of the depth chart. Because of these changes, the only efficient way to display the tables is one receiver at a time; so that's what I'm going to do.
MICHAEL CRABTREE
Below are the receiving yardage and efficiency stats for Crabtree in 2009, alongside those of the average playoff #1 WR and the #1 WR for the Super Bowl champions (Top-quartile performance in bold; Bottom-quartile performance in italics):
|
WR |
DVOA |
Rk |
DYAR/T |
Rk |
Yds/T |
Rk |
EYds/T |
Rk |
|
Michael Crabtree |
-20.0% |
29 |
-0.56 |
29 |
7.27 |
22 |
5.37 |
29 |
|
Marques Colston |
25.3% |
4 |
3.09 |
4 |
10.04 |
4 |
11.19 |
3 |
|
Playoff #1 |
18.2% |
7.6 |
2.48 |
7.6 |
9.17 |
8.8 |
10.18 |
7.6 |
Clearly, the 49ers did not have a playoff-caliber #1 WR this season. Moreover, because these are per-target stats, Crabtree's holdout was not a factor in the rankings. Looking at specific stats, Crabtree's average target was about 4 yards less valuable to the Niners than Colston's was to the Saints (See DYAR/T). As bad as that value-related difference is, it pales in comparison to their efficiency-related differences (See DVOA and EYds/T). Specifically, Crabs' targets were over 45% less efficient than Colston's, which translates into Colston's EYds/T being twice that of Crabtree's. Furthermore, whereas Crabs' EYds-to-Yds/T difference was negative (-1.90), Colston's was positive (1.15).
So what is to blame for these differences? Easy answer: Crabs was a rookie. You don't need some thorough statistical analysis to know that the list of impact rookie WRs is very short, and doesn't even include the GOAT.
Another ironclad reason is that Crabs' played in a relatively inefficient and unproductive pass OFF overall. Obviously, a WR is going to be more productive and efficient playing in Sean Payton's OFF with Drew Brees at QB than playing in Jimmy Raye's OFF with Alex Smith at QB. A third, related reason is that the amount of identity turmoil exhibited by the Niners' OFF this season hurt Crabs' stats. Given that route-running, and knowledge of the OFF are 2 of the major obstacles for a rookie WR, it doesn't help when his team is switching back and forth between being the tortoise or the hare.
Finally, the fact of the matter is that, when put in proper context, Crabs' season probably wasn't as bad as FO's stats suggest. Being a rookie, and missing 5 games to start the season, it's a testament to Crabs that he was still able to become a competent starter right out of the gate. Furthermore, his 625 receiving yards might not be that great, but it prorates to 900 yards in 16-game season. You want another short list? How about rookie WRs who started at least 10 games and had 900 or more receiving yards? So, given proper context, Crabs' underwhelming 2009 stats are less important than his stats going forward. More on that later.
JOSH MORGAN
Below are the receiving yardage and efficiency stats for Morgan in 2009, alongside those of the average playoff #2 WR and the #2 WR for the Super Bowl champions (Top-quartile performance in bold; Bottom-quartile performance in italics):
|
WR |
DVOA |
Rk |
DYAR/T |
Rk |
Yds/T |
Rk |
EYds/T |
Rk |
|
Josh Morgan |
-18.6% |
28 |
-0.46 |
28 |
6.51 |
24 |
5.63 |
29 |
|
Devery Henderson |
10.7% |
6 |
1.78 |
7 |
9.69 |
2 |
8.83 |
8 |
|
Playoff #2 |
3.7% |
12.3 |
1.31 |
12.3 |
8.11 |
12.5 |
8.41 |
12.2 |
Pretty much the same story here. Morgan's stats were underwhelming, but it was his first year as a starter and his 2009 season was constantly beset by turmoil totally out of his control. Develop chemistry with a new QB at midseason? Check. Temporarily lose playing time midseason? Check. Learn a new position midseason? Check. Really, a major injury was the only thing Morgan's tumultuous 2009 season was missing. So we should cut the kid some slack. One thing that is legit about his stats though - and I can't resist beating this dead horse one last time - is that Morgan's terrible DVOA accurately reflects his penchant for coming up 1 yard short on 3rd-down receptions.
ISAAC BRUCE
I wasn't planning to review Isaac Bruce's stats when I set out to write this post. However, during the course of the research, I found something that was pretty interesting. Below are the receiving yardage and efficiency stats for Bruce in 2009, alongside those of the average playoff #3 WR and the #3 WR for the Super Bowl champions (Top-third performance in bold; Bottom-third performance in italics):
|
WR |
DVOA |
Rk |
DYAR/T |
Rk |
Yds/T |
Rk |
EYds/T |
Rk |
|
Isaac Bruce |
-30.9% |
23 |
-1.47 |
23 |
5.39 |
21 |
4.37 |
23 |
|
Robert Meachem |
39.2% |
1 |
4.19 |
1 |
11.28 |
1 |
12.72 |
1 |
|
Playoff #3 |
12.7% |
8.3 |
2.01 |
8.6 |
8.47 |
8.4 |
9.33 |
8.4 |
What's interesting to me is how unequivocal this table is in suggesting the importance of a good #3 WR. The Saints had the best #3 WR in the NFL, and their season finished with a championship. Of course, NO is famous for their ability to spread the ball around among a number of WRs, so it's important that the #3 WR trend is also supported by playoff teams in general. It seems that, in today's NFL, a good #3 WR is increasingly important for overall team success. It may just be a coincidence, but the 49ers haven't had a #3 WR gain 500 or more receiving yards since JJ Stokes had 524 in 2000, and their #3 WR has averaged only 276 yards per season (!!!) since 2002 (aka their last playoff appearance).
VERNON DAVIS
Below are the receiving yardage and efficiency stats for VD in 2009, alongside those of the average playoff #1 TE and the #1 TE for the Super Bowl champions (Top-quartile performance in bold; Bottom-quartile performance in italics):
|
TE |
DVOA |
Rk |
DYAR/T |
Rk |
Yds/T |
Rk |
EYds/T |
Rk |
|
Vernon Davis |
8.8% |
19 |
1.02 |
19 |
7.48 |
18 |
7.53 |
21 |
|
Jeremy Shockey |
23.4% |
7 |
2.03 |
7 |
8.49 |
7 |
9.97 |
7 |
|
Playoff |
23.5% |
9.7 |
2.07 |
9.9 |
8.34 |
11.6 |
10.06 |
10.1 |
This table is probably the most surprising in this post. Essentially, it reflects a statistical reality of count stats like yardage and TDs, categories in which VD was at or near the top of the NFL. Namely, the amount a player amasses is dependent on the number of opportunities he has to amass. That's precisely the reason why I prefer per-play stats. So, to the NFL record book, which relies on count stats, VD's season was awesome statistically. But a good part of that awesomeness was simply the byproduct of being targeted 129 times this season. To put that in perspective, Crabs and Morgan had 167 targets combined. It's certainly plausible that each of them would have posted something like 78-965-13 if they were as much of a focus of the Niner OFF as VD was in 2009.
The real problem with VD's high number of targets, though, is what happened on about 40% of them. Namely, they fell incomplete. By definition, those 51 incompletions represented a VD target that (a) was unsuccessful for the purposes of DVOA, and (b) resulted in a gain of 0 Yds/T; both of which hurt his FO stats. Of course, this is just a roundabout way of saying that VD suffered from a 60.5% Catch Rate, which ranked #32 among TEs this season. For a comparison, consider that Shockey's Catch Rate was 71.6%.
It's easy to look at his abysmally low Catch Rate and say, "VD needs to stop dropping so many passes." However, it's important to remember that incompletions aren't only the fault of the receiver. Indeed, subtracting VD's league-leading 11 drops, we're still left with 40 incompletions that weren't the fault of his stone hands. Just so we're on the same page here, I'm saying that VD's getting screwed statistically by 40 passes for which lack of efficiency and yardage were more the fault of his QB. It's true that every receiver's DVOA, DYAR, and EYds are affected by non-drop incompletions. However, the specific case of VD illustrates that some receivers are more negatively affected by inaccurate QBs than others. If anything, I'd love to see what VD's advanced stats would be if FO only considered targets that resulted in completions or drops, and omitted those that fell incomplete because of the QB.
It turns out that, if you look at the full TE rankings, you actually find statistical evidence of an across-the-board QB effect. Here's a table, sorted by Catch Rate, showing rankings for the TEs with 100 or more Targets, which represents the VD-like members of TE-focused pass OFFs:
|
TE |
Catch Rate |
DVOA |
DYAR/T |
Yds/T |
EYds/T |
|
Jason Witten |
75.8% |
9 |
11 |
10 |
11 |
|
Dallas Clark |
75.2% |
8 |
8 |
9 |
8 |
|
Antonio Gates |
69.3% |
2 |
2 |
1 |
2 |
|
Brent Celek |
67.9% |
7 |
10 |
6 |
10 |
|
Tony Gonzalez |
61.9% |
17 |
16 |
29 |
15 |
|
Kellen Winslow |
61.1% |
16 |
18 |
23 |
19 |
|
Vernon Davis |
60.5% |
19 |
19 |
18 |
21 |
|
Greg Olsen |
55.6% |
32 |
32 |
38 |
31 |
As expected, the four 100-Target TEs with high catch rates have rankings that are basically in the Top 10 of each category, whereas the opposite is true for the other 4. Now, do you notice the one glaring difference between the 2 groups? The difference, of course, is that the high-catch-rate group had Pro-Bowl-caliber QBs throwing them the ball, whereas the low-catch-rate group didn't.*
So, the moral of the story is that, rather than telling VD he needs to stop dropping so many passes, it's actually more accurate to say that Alex Smith needs to stop throwing so many incompletions on passes intended for VD. If he does that, then VD's efficiency stats will improve considerably. And, given the efficiency levels exhibited by the average playoff TE, that would be extremely valuable for the 49ers' playoff hopes. At least we now know that VD can improve on his own: Compared to 2008, his DVOA improved 33.4%, his DYAR/T improved 2.12 yards, and his EYds/T improved by a whopping 4.57 yards.
BREAKOUT OF CRABS IN 2010?
It's clear that VD just had his breakout season (finally), and that VD's breakout helped the Niners' pass OFF actually begin to look like an NFL-caliber unit. Nevertheless, it would do wonders for the Niners to have one of their young starting WRs to join the breakout party in 2010. So let's finish this post up with a look at the likelihood that Crabs and/or Morgan will break out.
For those of you who play fantasy football, you're probably aware of the 3rd-year WR breakout theory, which says that a WR's breakout season (if he ever has one) usually occurs during his 3rd year in the league. If you're stat savvy, you're probably aware that "the 3rd-year WR breakout theory" is a misnomer. Previous research shows that it should actually be called the 2nd-to-5th-year WR breakout theory. Here's a graph that illustrates the point:
As you can see, the first major jump in 1,000-yard seasons actually coincides with 2nd-year WRs, and the peak occurs in Year 4, not Year 3. In fact, 5th-year WRs also post more 1,000-yard seasons than do 3rd-year WRs. So Year 3 isn't some magic number the way a lot of media outlets and pundits make it seem. What does seem relatively accurate, though, is that Year 3 typically begins a WRs peak years, which last until Year 6 or 7. Indeed, nearly 60% of all 1,000-yard seasons by a WR since 1977 were achieved by WRs in their 3rd through 7th seasons (288 of 487).
Going into 2010, the Niners' have 1 starting WR going into his 2nd year and the other going into his 3rd year. Given what I just presented, that makes Crabs and Morgan breakout candidates for 2010. However, it's not enough to know when WRs generally break out. For our purposes, we want to know how likely the Niners' WRs specifically are going to break out in 2010.
The available research to date suggests that at least one of them will. Tony Sans Nicolas and Matt Waldman of Fantasy Football Today did the earliest study on the topic and found the following using data from 1983-2003:
- 85% of breakout WRs did so between Years 2-5 in their careers.
- 81% had at least 41 receptions during the previous season.
- 78% had at least 2 TDs during the previous season.
- 71% had at least 400 receiving yards during the previous season.
- On average, 5 WRs break out each season.
As important as it was, 1 flaw in their research was that they seem to have gotten cause and effect backwards. Specifically, they limited their sample to WRs that did break out, and then looked at what stats breakout WRs put up during the season prior to their breakout. The problem with this is that it tells you everything about breakout WRs, but nothing about non-breakout WRs; and the idea is to differentiate between the 2 groups, not to simply describe 1 group. It's not surprising then that, from 2004-2008, Waldman's accuracy rate for prediction was only around 50%, which, although still pretty good, isn't anything like the 70-80% probabilities found in the analysis. In short, it's much easier to explain the past by only analyzing the successes than it is to predict the future by analyzing both successes and failures.
Luckily, Phil Dussault of Ultimate Fantasy Football Strategy put the horse back in front of the cart, and actually determined breakout probabilities looking forward. Rather than looking only at breakout WRs, his sample included 543 2nd-to-5th-year WRs from 1990-2008 who met a specific criteria of production (at least 300 receiving yards & 10 games in consecutive seasons), regardless of whether or not they broke out. After statistically differentiating between the breakouts and non-breakouts, here's what he found:
- Breakout probabilities differ depending on whether a WR in his 2nd, 3rd, 4th, or 5th year.
- 3rd-year WRs actually have the lowest breakout probability among WRs in Years 2-5 of their careers.
- 43.5% of rookie WRs who had less than 70 catches, 1,000 receiving yards, 7 TDs, and a 0.120 TD-per-reception ratio ended up breaking out the following season.
- 33.7% of 2nd-year WRs who had less than 70 catches, 1,000 receiving yards, a 0.150 TD-per-reception ratio, and averaged less than 17 yards per catch ended up breaking out the following season.
- In general, 42.5% of 1st-to-4th-year WRs who scored 100-200 fantasy points, but averaged less than 0.060 TDs per reception, broke out the following season.
Obviously, situational factors like depth chart status, injury, and QB-caliber play an important role. Waldman also identified having a starting-caliber fantasy WR on the other side of the field and/or a starting-caliber fantasy RB as informative. Nevertheless, full-blown statistical studies incorporating these factors have yet to be done, so all we can go on as of right now is the starting point that Sans Nicolas, Waldman, and Dussault have provided.
Given the general similarities between the findings of the 2 studies, I think it's best at this point to treat them as complementary. So, here are 2 tables - 1 for Crabs and 1 for Morgan - showing all of the statistical criteria that either explain or predict WR breakouts, and whether or not Crabs and/or Morgan meet them:
|
General Breakout WR Characteristic |
N |
Crabtree |
2010 Breakout? |
|
Player Yr Within Range N |
2-5 |
2 |
Yes |
|
Recs in Prev Season > N |
40 |
48 |
Yes |
|
Rec Yds in Prev Season > N |
399 |
625 |
Yes |
|
TDs in Prev Season > N |
1 |
2 |
Yes |
|
TDs per Rec in Prev Season < N |
0.060 |
0.042 |
Yes |
|
FF Pts in Prev Season Within Range N |
100-200 |
122.5 |
Yes |
|
2nd-Year Breakout WR Characteristic |
N |
Crabtree |
2010 Breakout? |
|
Recs in Prev Season < N |
70 |
48 |
Yes |
|
Rec Yds in Prev Season < N |
1,000 |
625 |
Yes |
|
TDs in Prev Season < N |
7 |
2 |
Yes |
|
TDs per Rec in Prev Season < N |
0.120 |
0.04 |
Yes |
|
General Breakout WR Characteristic |
N |
Morgan |
2010 Breakout? |
|
Player Yr Within Range N |
2-5 |
3 |
Yes |
|
Recs in Prev Season > N |
40 |
52 |
Yes |
|
Rec Yds in Prev Season > N |
399 |
527 |
Yes |
|
TDs in Prev Season > N |
1 |
3 |
Yes |
|
TDs per Rec in Prev Season < N |
0.060 |
0.058 |
Yes |
|
FF Pts in Prev Season Within Range N |
100-200 |
122.7 |
Yes |
|
3rd-Year Breakout WR Characteristic |
N |
Morgan |
2010 Breakout? |
|
Recs in Prev Season < N |
70 |
52 |
Yes |
|
Rec Yds in Prev Season < N |
1,000 |
625 |
Yes |
|
TDs in Prev Season < N |
9 |
2 |
Yes |
|
Rec Yds per Catch in Prev Season < N |
17 |
10.13 |
Yes |
|
TDs per Rec in Prev Season < N |
0.150 |
0.058 |
Yes |
Well, well. I think we have what's called unanimous consent here. When looking only at their 2009 stats, it's clear that both Crabs and Morgan fit the profile of a breakout WR in 2010. That's definitely a good thing for the Niners. Among the non-statistical factors, it seems like depth chart status, starting-caliber fantasy RB, and health all work in their favor. So, if everything stays the same as it is right now, the only thing that might stand in the way of Crabs and Morgan is - you guessed it - Alex Smith. If Smith improves enough to become a Top 15 QB in 2010, Crabs and/or Morgan are likely to break out. If he doesn't, then their favorable statistical profile won't matter nearly as much.
BOTTOM LINE
Just to recap, VD was awesome according to traditional stats, but merely average according to FO's stats; largely due to the fact that Alex Smith enjoyed throwing incompletions in VD's direction. Crabs and Morgan were pretty bad statistically speaking, but there were several mitigating circumstances working against them. Finally, the 49ers essentially played the entire season without a #3 WR.
So, based on the stats I've presented in Part 7 of the season review, here are the things the Niners need to do at the WR position in order to seriously contend in 2010:
- Continue to develop Alex Smith, especially when it comes to accuracy
- Set it and forget it: Put Crabs and Morgan on the field from Day 1, and leave them there the whole season
- Find a competent #3 WR and get him the ball frequently
The theme is clear. If the 49ers are going to make the playoffs next season, then their pass OFF must improve. Furthermore, the continued development of Alex Smith is the most likely catalyst for improvement.
* Remember, Gonzalez played 2 games with Chris Redman at QB, and 3 more with an injured Matt Ryan at QB. Also, although Jay Cutler might be considered a Pro-Bowl-caliber QB based on his last season with the Broncos, he couldn't hit the broad side of a barn in his first season with the Bears.
**DVOA, DYAR, and EYds statistics used to produce this article were provided by Football Outsiders.