I apologize for failing to post an article yesterday, as I’ve been entrenched in the Cowboys’ 2010 playaction pass numbers. I love studying Jason Garrett’s use of playaction, particularly because of the statistical “anomalies” that arise from season to season (although the consistency of these numbers actually makes them anything but anomalies). The idea that such seemingly unique numbers can develop on such a consistent basis is absolutely fascinating to me.
I’ve already posted a couple of analyses on Garrett’s 2010 playaction use, including one after Week 16 which compared this season’s playaction statistics to those from 2009. I’ve republished those results below with the Cowboys’ Week 17 game in Philly added into the totals...
- Of the 109 playaction passes, 14 were thrown 20+ yards downfield (12.8 percent).
2009 Comparison: 4.8 percent
Analysis: Garrett certainly made an effort to get the ball downfield following playaction looks, but this was one of the only areas in which he improved.
- Dallas ran screen passes on 53 of their 528 non-playaction passes (10.0 percent). That screen rate nearly doubled on playaction passes to 19.3 percent.
2009 Comparison: 22.9 percent screen rate following playaction; 7.1 percent otherwise.
Analysis: We see a bit of an improvement here, but that’s probably due to the higher overall screen rate. The Cowboys did average a solid 7.76 yards-per-pass on playaction screens, due in large part to Felix Jones’ average of 15.0 yards-per-catch on those sort of plays.
- Of the 100 playaction passes attempted, just 43 were to the right side of the field. **NOTE: There were only 100 playaction passes attempted due to six sacks and three scrambles, i.e. 109 total playaction passes called.
2009 Comparison: I say “just” 43 because 63.9 percent of 2009 playaction passes went to the right side.
Analysis: 2009 seems like an aberration.
- The Cowboys still ran just FOUR playaction passes with 1-4 yards-to-go. That is only 2.96 percent of the 135 overall plays in that range.
2009 Comparison: 4/132 (3.03 percent)
Analysis: Incredible. These are the kind of numbers that get me excited (I’m a strange individual). Seriously though, the EXACT same number of playaction passes with 1-4 yards-to-go on nearly the exact same number of opportunities.
The idea that Garrett doesn’t utilize playaction in “obvious” running situations is mind-boggling to me. These numbers must change in 2011.
- 62 of the 109 total playaction passes were with exactly 10 yards-to-go. That’s a rate of 56.9 percent.
2009 Comparison: 59.3 percent
Analysis: Wow. The similarity of those percentages alone is nothing short of amazing, but the fact that Garrett utilizes playaction so much in such a specific situation is just as incredible. I’m not necessarily against this tactic, as the majority of these passes came during first and 10 situations when most defenses, mistakenly, are playing to stop the run. Still, the rate should be a bit lower if for no other reason than an increase in short-yardage playaction looks.
- The Cowboys again ran more playaction passes with 20+ yards-to-go (six) than with 1-4 yards-to-go (four).
2009 Comparison: Five playaction passes with 20+ yards-to-go; four with 1-4 yards-to-go
Analysis: For the life of me, I cannot figure out why Garrett calls playaction passes in such obvious passing situations. It isn’t as if the Cowboys have been successful on them, averaging just 4.5 yards-per-pass. Again, the consistency here is astounding to me.
- Only 24 of the 109 total playaction passes came with less than 10 yards-to-go. That’s just 22.0 percent.
2009 Comparison: 19.8 percent
Analysis: I feel like I’m stating the obvious in claiming that someone who has watched football for only a week would realize that, perhaps, more than one-fifth of a team’s playaction passes should come with less than 10 yards-to-go.
Spread vs. Tight
The other playaction study I published this offseason broke down the Cowboys’ playaction passes from spread and tight formations. I noticed that, contrary to my prediction, the ‘Boys were far more successful on playaction looks from spread formations, averaging over four more yards-per-attempt in 2010.
I’m still not entirely sure why we see these numbers. It’s possible that a small sample size is at work, although the large discrepancy in passing efficiency seems to make the 53 play sample size a bit more valuable.
My best guess is that the situations in which Garrett calls playaction passes (i.e. very few “obvious” running situations) is the largest contributor here. If the Cowboys ran more short-yardage playaction passes, I presume the efficiency of playaction looks from tight formations would increase due to defensive expectations. Short-yardage + tight formation = expectation of run.
Overall Playaction Efficiency
You can see below that the Cowboys simply aren’t getting the job done on playaction passes. The 6.29 yards-per-play is atrocious, particularly when you consider the situations in which playaction passes are generally run: ones with high upside. With Garrett calling so many playaction passes with 10 yards-to-go (56.9 percent), we know the Cowboys are generally in “normal” down-and-distances–not short-yardage, and not too many second or third and longs.
The sack rate on playaction passes is down from 8.7 percent last year, but the sack rate in general decreased in 2010. You can also see quite a nice completion rate on playaction passes, but looks can be deceiving...
Screen Passes Following Playaction
As I mentioned above, Garrett loves to dial up screen passes following playaction looks, doing so 19.3 percent of the time in 2010. That’s about double the screen rate on non-playaction passes (and certainly a major reason for the high completion rate).
One of the reasons Garrett utilizes a playaction look before many of his screens is because, often times, he isn’t running “traditional” screen passes to the running back. Instead, Garrett likes to suck the defense in toward the running back by showing playaction, then throw a quick screen or bubble screen to a receiver. Actually, 71.4 percent of playaction screen passes went to a player other than a running back. That rate dropped to just 40.0 percent on non-playaction screens.
It was relieving to see Garrett take some shots downfield following playaction looks this season, but I’d still like to see more than 12.8 percent of playaction passes travel 20+ yards. The Cowboys could probably maximize their playaction effectiveness by stretching the field on closer to 25 percent of playaction passes. At worst, the increased rate of deep pass attempts would open things up underneath.
It’s also obvious the ‘Boys desperately need more playaction passes in running situations. A less than three percent playaction rate in short-yardage situations (1-4 yards-to-go) is a joke, as is the 22 percent of playaction passes with less than 10-yards-to-go, and the 56.9 percent rate of playaction looks with exactly 10 yards-to-go. These numbers have remained uncannily stable from 2009, proving we’re witnessing something inherent to Garrett’s play-calling rather than an aberration.
Those of you who know me know I like the Garrett hire and I think he’ll improve considerably as a head coach. He’s certainly shown the ability to adapt in other areas of his coaching, but he’s late to the table on this one. Garrett is young, confident and aggressive, but if he doesn’t show the willingness to aggressively change his playaction calls, it will be difficult to reverse the team’s fortunes in 2011.