I know I promised an addendum to my previous post, but I think that was getting confusing so I'm starting a new one aimed at answering a specific series of comments.
The basic premise, as you might have guessed from the title even if you did not read the initial post, which is located here, is that despite the record number of receptions, and the large number of counting stats, this was possibly Jason Witten's worst year, not his best. Furthermore, since TEs tend to fall off after 30, this is quite possibly a sign of his decline.
One of the primary arguments against my thesis is that the Cowboys used Jason Witten differently in 2012 than they had in previous years. If this were the case, we should see some evidence in the play-by-play statistics.
Since this is statistic-driven, I should mention that I have discovered that statistics vary by source. My source is PFF. Unfortunately, their data only goes back to 2008. However, I choose to use them because they have more detailed play-by-play data than anyone else I have access to. Also, in order to get the largest sample I could, I included the playoff games in 2009.
Also, I am starting this post with a large table of data. I apologize but we cannot properly examine Jason's usage without this data.
|Behind LOS Targets||2||4||6||8||5||25|
|Behind LOS Tar. Pct.||1.4%||3.5%||4.9%||5.8%||4.2%||3.9%|
|Behind LOS Receptions||2||4||6||6||5||23|
|Behind LOS Rec. Pct.||1.8%||5.1%||6.4%||5.6%||6.2%||4.9%|
|Behind LOS YAC/rec||9.0||5.5||9.8||7.2||7.4||7.8|
|0-9 Yards Targets||103||73||86||97||69||428|
|0-9 Targets Pct.||70.5%||64.0%||69.9%||70.8%||58.0%||67.0%|
|0-9 Yards Receptions||84||53||70||82||53||342|
|0-9 Yards Rec. Pct.||76.4%||67.1%||74.5%||75.9%||65.4%||72.5%|
|Total Targets >9 Yards Pct.||71.9%||67.5%||74.8%||76.6%||62.2%||70.9%|
|10-19 Yards Targets||30||25||26||25||32||138|
|10-19 Targets Pct.||20.5%||21.9%||21.1%||18.2%||26.9%||21.6%|
|10-19 Yards Receptions||22||14||16||17||16||85|
|10-19 Yards Rec. Pct.||20.0%||17.7%||17.0%||15.7%||19.8%||18.0%|
|20+ Yards Targets||11||12||5||7||13||48|
|20+ Targets Pct.||7.5%||10.5%||4.1%||5.1%||10.9%||7.5%|
|20+ Yards Receptions||2||8||2||3||7||22|
|20+ Yards Rec. Pct.||1.8%||10.1%||2.1%||2.5%||8.6%||4.7%|
|Total Targets 10+ Yards Pct.||28.0%||32.4%||25.2%||23.3%||37.8%||29.1%|
|Pass Plays Witten Played||658||579||554||613||491||2895|
|Pass Plays Witten Blocked||71||61||91||92||50||365|
|Pass Plays Targeted Pct.||22.2%||19.7%||22.2%||22.3%||24.1%||22.1%|
|Pass Block Pct.||10.8%||10.5%||16.4%||15.0%||10.2%||12.6%|
|Yards per Pass Route||1.65||1.68||1.92||1.98||2.03||1.84|
Yeah, that's a lot of stuff. Sorry, there's no way around it.
There are 4 levels of throws. Behind the line, less than 10 yards, 10-19 yards, and 20+ yards. Again, these are the distances that the pass traveled in the air. Each level lists the targets, percentage of total targets were at that distance, receptions, percentage of receptions at that distance, and the Yards After the Catch per reception at each level. Because of the small samples of both the passes Behind the LOS and over 20+ yards, I have also included a row combining the target percentage of less than 10 yards and more than 10 yards.
At the bottom I list the number of passing plays that Witten was involved in, the number of passing plays he stayed in to block, and the percentage that they kept him. I also listed the percentage of pass plays that Witten was targeted and his yards per route run.
The primary argument with my thesis was that the Cowboys threw to Jason Witten deep less than in previous years because of the offensive line. Jason's efficiency, their theory goes, was hurt by a change in the game plan.
So let's look at the percentage of targets thrown to Jason at various distances. If their game plan changed, we should be able to see it in the data by the distance of the passes targeted to Jason. If they threw fewer deep passes to Jason in 2012 than in other years of his career, then their argument likely has some validity. If not, then we need to look at other causes.
Unfortunately, this data does not corroborate the argument that they used Jason Witten differently in 2012 than in his career. In fact, the data shows that 2012 is the single most average year in Jason's career from 2008-2012. In almost every category, the median number is the 2012 number.
They did throw fewer deep routes by percentage to Jason in 2012 than they had in 2011, but not by a ton. 28.0% of Jason's targets in 2012 were over 10 yards in the air. 32.4% of Jason's targets in 2011 were over 10 yards in the air. This is a significant, though not large, difference, and if this matched the rest of Jason's career than this would be a clear argument in favor of the change in the way Jason used affecting his efficiency.
However, this does not match his career. In these 5 years of this data, they basically had three game plans. In 2011 and 2008 they targeted Jason with more deep passes. In 2010 and 2009 they targeted Jason with fewer deep passes. 2012 is the average year between these two route trends.
His average of throws less than 10 yards for his career is 70.9%. His average in 2012 was 71.9%. Essentially the same. His average of throws over 20 yards for his career is 7.5%. His average in 2012 was 7.5%. Exactly the same. These numbers are also the median of the 5 years of data.
Despite these 3 different game plans, Jason's yards per catch, yards per target, and YAC per catch remained essentially the same whether they threw a lot to him deep such as in 2008 or threw a lot of short passes as in 2009 except for the dropoff in 2012.
So let's see how they used him overall. Why did he get more targets than ever? When Jason was in the game, what was his role?
Jason was left in to block 10.8% of the time. This is exactly in line with his 2011 and 2008 averages. Clearly, they wanted him out as a target.
More interesting is that while Jason was targeted 146 times, this is entirely a product of the number of passes attempted by the Cowboys as a whole. Of the pass plays the Cowboys attempted in 2012, 22.2% of the time they ended up throwing to Jason. This is exactly the same as in 2010, and basically the same as 2009 and his career average.
So, in other words, he basically was used as he normally was. He was thrown to basically as he normally was, at the average distance, and he blocked about the same amount of time.
Unfortunately, in 2012, he was less productive when they threw to him. His yards per route run is his lowest of the past 5 years.
Worse yet, there's a clear downward trend. His YPRR in 2011 shows he was starting to lose a step two years ago. Had the Dallas Cowboys not targeted him more with deeper routes in 2011 than in any other year besides 2008, his yards per catch and target would have significantly dropped in 2011 and I might have noticed this last year as opposed to now.
All of this data says to me that there is nothing extraordinary about the way the Dallas Cowboys used Jason Witten in 2012 as opposed to his career. There's nothing that I see in here that suggests the drop in his yards per catch and yards per target was caused by any change in his game plan for his career.
In fact, while there *is* a change in the game plan for 2012 from 2011, that change masks a drop in efficiency that started two years ago. This supports the argument that the drop in efficiency was caused by some other factor than how he was used, and that factor is probably age.
Now, there were other arguments against my overall thesis. If I can, I will address some other ones in another post when I get to it. One suggestion was that he was really good in clutch situations. A quick glance, without plugging all of these numbers into a spreadsheet suggests that in 2012 he was no better than in the past, and perhaps worse. If someone else wants to go through all of this data, feel free :)
Another suggestion was that everyone in the Cowboys dropped because the OLine was so bad and the running game was wretched. Again, a quick glimpse without really breaking down all of these stats for every one of the Dallas receivers suggests to me that this is also not true. For example, Dez's YAC/rec is trending up. Miles's is trending down.
In any case, this post is probably long enough and gives you all something to chew on for a bit.