Tuesday, September 07, 2010

Decoding Garrett '10

As seen Here: http://cowboysblog.dallasnews.com/archives/2010/09/bob-sturm-decoding-jason-garre.html

Today, is where you begin to find out if I am what you are looking for. I believe I bring different topics and discussions to the forum than some, and I quickly concede that my interests are not for everyone. For instance, if "who Tony Romo is dating" or "what Terrell said about Romo on Wednesday" is incredibly interesting, then what I am about to discuss may make your eyes roll back in your head and induce a deep sleep.

Today and pretty much every Tuesday all season long, I will offer information for you under the heading of "Decoding Garrett" which is a summary title of my 3 year obsession to get inside Jason Garrett's head and attempt to figure out his game plan and objectives for the Dallas Cowboys offense.

Since 2008, on my personal blog We have broken down every Cowboys game by personnel package and this helps us to identify what Garrett is trying to accomplish.

Here is the game summary from the December Redskins Game so you can see what I am talking about. We break down the big plays (both of the good variety and the bad) to try to engage in our own level of film study and attempt to understand Jason Garrett on a level much deeper than fantasy football or just looking at the final score.

Why did certain strategies work? Why did other objectives just not make sense? Every game, I try to figure that out with data nd video. So, hopefully those of you who enjoy football from a "chess game" perspective will really get into this.

However, we must walk before we can run, so here are a few things that you will need to know as we enter week 1.

Yesterday, when I posted about Tashard Choice I referenced "S11" as their 3rd Down package, and there was one comment that I wanted to address this morning:

Robert42078: Bob, do me a favor either on air or in this forum. Can you go over the personel packages and their corresponding names.

Absolutely, Robert. And I very much believe that this is something that is essential to understand a game plan in the NFL (or any level of football). After every play, an offensive coordinator looks at his play sheet and decides what he wants to do on the next snap. He must consider down-and-distance and pick something that makes sense. Once he decides, he sends out the proper personnel package to execute the play. Meanwhile, across the field, the defensive coaches for the opponent observe what Garrett is sending out there and then he matches up with a proper personnel change of his own (Nickel, Dime, Base, etc).

By simply watching a game on television, you can easily identify the personnel that an offense is sending out there. And once you know that, you are prepared to keep a tally on a notebook to see for yourself what the Offensive Coordinator is attempting to do.

Remember, there are 2 important things that a defensive player is attempting to do pre-snap. 1) He is looking at personnel and making sure from their film study and their defensive game plan what his responsibility will be. He usually has time to absorb this piece of information during the huddle and get a proper play call from his own sideline. Then 2) He is looking at formation. This deploying of the 11 offensive players are a subset of the personnel. You can have "12" personnel on the field and it can mean many different things. If you are a middle linebacker and both tight ends are lined up off right tackle, you must make sure you identify enough of your mates to neutralize the "strong side" run formation on that side. If, instead, Witten is flexed out in the slot, then you are thinking pass, and that Bennett is serving to pass protect while we better get some help on Witten before he runs the verticals that he is so likely to run with effectiveness. The worst part about formation is that it can happen at the last possible second before the snap with pre-snap motion and adjustments part of every NFL offense.

With that in mind, we must go back to personnel groups. If we are going to attempt to process some of this information, we must first all agree on terminology.

Below, please find the Cowboys offense broken down by personnel groups.

For any of this to make sense, you have to know what all of the different packages mean. Basically, it is very simple. Every Offense in the world has 1 QB and 5 Offensive Linemen. Therefore, if 11 players are on the field, then that leaves 5 players who can join the QB in skill positions and the Offensive Coordinator has to choose how to deploy those 5. So, the groups are simple. "11" means 1 RB and 1 TE, so you add those 2 numbers together (1+1 = 2) and subtract that number from 5 to get how many WRs are on the field at the time. (11 will mean 3 WRs, of course).

Personnel PackageDescription
111 RB, 1 TE, 3 WR
121 RB, 2 TE, 2 WR
131 RB, 3 TE, 1 WR
212 RB, 1 TE, 2 WR
222 RB, 2 TE, 1 WR
232 RB, 3 TE
S11Shotgun, 1 RB, 1 TE, 3 WR
S12Shotgun, 1 RB, 2 TE, 2 WR

Table Tutorial

Do you get it? In all of the packages, the first number is the number of RBs, the second number is the total number of TEs. And "S" means Shotgun. So, when you watch the game on tv, you can easily identify the package before the snap to see what Garrett is doing.

And when you do all of this, you will quickly see patterns. You will see that the Cowboys are in Shotgun 94% of the time last season on 3rd Down and 4 yards or more. You will see that the Cowboys on 1st Down are 52% run in their 456 1st Down snaps last year. But, all a defense has to do is look at the position of the QB. If Jason Garrett shows shotgun on 1st Down, the Cowboys are passing 83% of the time; under center they run 65% of the time on 1st Down.

The reason we have to keep track of this is simple to me. If conventional wisdom is that the Cowboys are a good "shotgun" team with "11" personnel (last year it was generally Choice, Witten, Crayton, Williams, and Austin), then we have to ask the question why they were more effective passing the ball in "12" personnel in yards per play and passer rating. Romo threw the bulk of his interceptions when they went to the "S11" package and did not get the results to make it worth the while.

Anyway, Here are the 6 different Personnel Groups that Jason Garrett called more than 16 times in 2009. My logic is that we should attempt to try to keep this complicated topic as simple as possible. So, I am not saying the Wildcat wasn't important, but if it wasn't used more than 20 times all year, then we will not bore you with the Data. I have included a picture with each of them, so you know what you are looking for:

S11 Personnel (1 RB, 1 TE, 3 WR - In Shotgun) - The MOST Run Package in 2009
This is the default package for any 3rd Down that is not 3rd and short. This is also the default 2-min drill offense. This is the bread and butter of the offense for the entire Jason Garrett era.

Run/PassYardsAvg Per Play% of Snaps
293 Total Snaps20386.96-
44 Runs2856.4815%
249 Passes17537.0485%

HTML Tables

22 Personnel (2 RB, 2 TE, 1 WR) - The 2nd Most Run Package in 2009
In 2009, this is what the Cowboys loved to do on 1st and 2nd Down as they were attempting to demonstrate their physical running style. What is odd about the '09 Cowboys is that the stats say that they were a "power running team". And they were in "22". The problem is, when they needed short yardage late in the year against the Chargers or the Vikings, they knew they couldn't get it. A power running team has to be able to convert 3rd and 1. In Minnesota, the Cowboys didn't even try. 3rd and 1 became a passing situation.

Run/PassYardsAvg Per Play% of Snaps
181 Total Snaps10015.53-
140 Runs8125.8077%
41 Passes1894.6123%

HTML Tables

12 Personnel (1 RB, 2 TE, 2 WR) - The 3rd Most Run Package in 2009
The most versatile look available. If the defense tries to stack against the run, the Cowboys can audible into a pass with 4 weapons on the field (Austin, Williams, Bennett, and Witten). And, if they sit back in a pass posture, the Cowboys pound the rock with 7 Offensive Linemen. It is a very popular base look around the league, instead of the traditional "21" look.

Run/PassYardsAvg Per Play% of Snaps
167 Total Snaps11757.04-
75 Runs3214.2845%
92 Passes8549.2855%

HTML Tables

21 Personnel (2 RB, 1 TE, 2 WR) - The 4th Most Run Package in 2009
Since the beginning of time, this is how football looked. But, Jason Garrett knows that 21 means Deon Anderson is on the field. He is a fine blocker, but not a threat to do anything else. If he takes him off and puts on a 2nd TE or a 3rd WR, the offense is far more threatening.

Run/PassYardsAvg Per Play% of Snaps
104 Total Snaps6336.09-
70 Runs3014.3067%
34 Passes3329.7633%

HTML Tables

S12 Personnel (1 RB, 2 TE, 2 WR - In Shotgun) - The 5th Most Run Package in 2009
A variation of S11, where Martellus comes on for a 3rd WR. Slightly difficult match-up for the secondary, but it also allows TE pass protection help without Witten having to do it.

Run/PassYardsAvg Per Play% of Snaps
118 Total Snaps9337.91-
9 Runs343.788%
109 Passes8978.2392%

HTML Tables

13 Personnel (1 RB, 3 TE, 1 WR) - The 6th Most Run Package in 2009
A variation of the "22" that worked John Phillips in Deon Anderson's spot. They threw a TD to Phillips in the Eagles playoff game out of this if memory serves.

Run/PassYardsAvg Per Play% of Snaps
64 Total Snaps2604.06-
43 Runs1683.9167%
21 Passes924.3833%

HTML Tables

Below are the final overall numbers for the offense by package. It is interesting to look at 2008 vs 2009. So much talk about how the offense was going to be different and more "Romo-Friendly". As we noted above, the number of runs was pretty much exactly the same as it was in 2008. Losing Terrell Owens and trying to run more efficiently meant only 1 thing in totality in 2009 - more Martellus Bennett and John Phillips.

Look at the uptick in "22", and "S12" - 187 more snaps this season (nearly 12 more PER GAME) with 2 TE Sets. Most of that is Martellus Bennett. I have no idea why his production is so very low. It is alarmingly low (15 catches???) but the object of the game is not to make Bennett a great fantasy football player. It is to make the Cowboys offense efficient. And, it seems difficult to argue that these 2-TE sets are the bread and butter of the 2009 Dallas Cowboys. And that offense was very, very effective.

And for all of the 3 Tight End Sets, now you see why the John Phillips injury is thought of as a rather large deal on the coaching staff.

2009 Final Numbers

PackagePlays RunYardsPercentage

Table Tutorial

2008 Final Numbers

PackagePlays RunYardsPercentage

Table Tutorial

So, like I said, this isn't for everyone. But, I hope it is for you.


jamezyjamez said...

Looks like he prefers the 4-4-2, occasionally morphing into a 3-5-2 to achieve better possession in midfield. This guy is a genius.

Doctorjorts said...

Thanks, Bob. I really look forward to your analysis this season. I was wondering if you might consider keeping track of avg. down and distance for these packages, too? It looks like the running and passing yards/play go inverse of what the personnel would lead you to expect (high rushing/play out of S11, high passing/play out of 21) so I would think that's because they're going counter to what the defense expects/what the down and distance would dictate as conventional wisdom. I think context-based stats are much more interesting, and paint a much more complete picture. Otherwise, we're misled into thinking the Cowboys ought to run out of S11 until the defense shows they can stop it, which is a ridiculous suggestion.

Roger Light, artist said...

this is great stuff. i would love to see a breakdown of which packages and plays are leading to the biggest plays from our top play-makers.i.e. what specifically produced the biggest gains for witten and miles and jones and choice. this would be great to see, because the casual fan thinks too simplistically (bench Barber, stick Jones in there, look how much better his average per carry is!) I couldn't tell you the packages and plays the looked the worst for Jones, but there were definitely some poor fits for him last year to go along with some absolute money plays. It would be very awesome to this Cowyboys nerd to see what works, what doesn't for each player. We generally only get this analysis when we are talking about Roy sucking on pass routes to the middle of the field. (I would rather not even hear about him until we have some good news to report)