Tuesday, November 22, 2011

Decoding Garrett - Week 10 - Data - At Washington

The Cowboys won a game Sunday that is tough to fully explain from a statistical standpoint. From a production perspective, they accumulated their 2nd worst yardage day of the year with 354 yards. Only the blowout loss in Philadelphia saw fewer yards (266), while only the loss to Detroit saw more snaps (75). 73 offensive plays and barely 350 yards makes for a very poor 4.84 yards per snap. That would be easily the least efficient day of the season for the offense crew.

And as we saw late, it wasn't like the defense carried the team to victory, either. The defense has the dubious distinction of conceding 166 passing yards to Rex Grossman from the start of the 4th Quarter on. That is certainly not going to make the rest of the NFL take notice, either.

So, why did the Cowboys win on Sunday? And why are they playing winning football overall in the last 3 weeks during this streak?


Look no further than what is referred to around the NFL as the "Money Down". 3rd Down conversions can either sink a perfectly healthy ship if they are going wrong, or they can pull a team out of a bad spot if they are going right. There is almost nothing an offense does that has more impact than moving the chains when you simply must.

The sample sizes are not large enough in the last 3 games to say definitively that the Cowboys have figured out one of their biggest issues in October and November, but just one check of the 3-4 start versus the 3-0 run they are currently on takes you to this particular metric:

Weeks3rd Down Conv-AttRate - Rank
1-728-8234% - 22nd
8-1022-4351% - 2nd

HTML Tables

Just look at those results. The NFL average for 3rd Down conversions of any length is 38%. In 7 games, the Cowboys were sitting just above Washington's offense and just below Denver's as they ranked 22 out of 32 in 3rd Down conversions. A quick look at the bottom teams in the league and their talent levels, and you will quickly notice that it is made up of teams that just do not have the horses to compete. Dallas has no business down in that group. Meanwhile, the best 3rd Down conversion teams in the NFL have QBs named Rodgers, Brees, Brady, Roethlisberger, and Rivers.

Since that trip home from Philadelphia, the Cowboys - albeit in a 3 game sample so it is too early to consider this problem fixed - the Cowboys now are 2nd only to the Packers in moving the chains. And wasn't that the story on Sunday? Yesterday, we discussed at rather great length that the only reason Dallas won was because Tony Romo's 3rd Down play was off the charts late in the game. The Touchdown to Robinson, the TD to Witten, and of course the 3rd and 15 conversion to Bryant in Overtime all testify to this fact.

And how has Romo been in those two stretches of time on 3rd Down? Well, his QB rating in September and October on 3rd Down is under 80 at 79.6. That ranked him below Colt McCoy, Josh Freeman, and Michael Vick for 3rd Down marksman, and miles below the elite QBs. But, since the calendar turned to November, Tony Romo on 3rd Down is 2nd to only one passer in football - Aaron Rodgers. Romo's passer rating during that stretch sits at a near perfect 146.5. Off the charts.

Again, we don't want to over-explain a 3-game sample against 3 poor teams, but if this is a sign of a QB getting fully healthy and an offense finally coming together with a runner, a line that is playing better, and receivers that are starting to figure it out, then the sky could be the limit during the final third of the season.


It is certainly difficult to imagine back in training camp that the fan base would be depressed with the absence of a player they had never heard of, but knowing that Tony Fiammetta was not available in Washington and likely won't be on Thanksgiving, either, is starting to give everyone pause.

Fiammetta was not in the plans for the Cowboys at fullback, for they felt they were squared away with two other candidates at training camp, Chris Gronkowski and Shaun Chapas, a 7th round pick this past April out of Georgia. But, in camp, neither player was able to justify a roster spot, so around Labor Day, the Cowboys scanned the waiver wire and claimed Fiammetta from Carolina right before the start of the season and cut the other two loose.

Since then, we merely have to look at the following graphic to see how this team runs the ball in "under center" situations with their new FB and without him available:

W/ Fiammetta1057447.08
W/O Fiammetta1233723.02

HTML Tables

One thing is certain about the Cowboys in 2011 - that their running game changes from week to week. Many have seen these Fiammetta numbers and have properly pointed out that DeMarco Murray's run as the lead runner has coincided with the games Fiammetta has played, and that is mostly true. However, Fiammetta's performance in Week 3 against Washington opened up the lanes for Felix Jones with great success. Also, the game this week at Washington was a game with Murray, but without a fullback, and once again the Cowboys proved they have no back-up plan to Fiammetta where they needed 32 carries to get 89 yards. John Phillips is many things for this team, but lead blocker is not one of them.

Murray is getting 6 yards a carry and Jones is getting 4. Another interesting split is that the team averages 6.2 yards per carry running the ball at home and 3.3 yards per carry on the road. At Cowboys Stadium, the Dallas running game has not had a day under 100 yards. On the road, they have no games over 100 yards.

I would very much enjoy a candid discussion (as opposed to a guarded Jason Garrett in a press conference setting where he tries to say very little) with the Cowboys offensive coaching staff as to their theories on what is working and what is not in running the football, but it sure seems that Fiammetta is the one constant that is required for this team to really run the ball well.

Data from Week 10 at Washington

1st Down Run-Pass16-15
2nd Down Avg Distance to Go7.33
2nd Down Run-Pass12-12
3rd Down Avg Distance to Go7.70
3rd Down Run-Pass3-14
3rd Down Conversions8-17, 47%

HTML Tables

Drive Starters - 1st play of each drive can often reveal the intent of a coach to establish his game plan.

The Cowboys passed the ball to start nearly every drive. Deceiving because they seldom were downfield passes, but a series of high-percentage passes seemed like the plan of the day. This number factors in 2 kneel downs, also. So in reality, the Cowboys started 10 of 12 drives with a pass play.

Wk 1-At New York Jets: 13 Drives - 5 Run/8 Pass

Wk 2-At San Francisco: 10 Drives - 4 Run/6 Pass

Wk 3-Washington: 11 Drives - 5 Run/6 Pass

Wk 4-Detroit: 14 Drives - 7 Run/7 Pass

Wk 5-At New England: 11 Drives - 4 Run/7 Pass

Wk 6-St Louis: 11 Drives - 8 Run/3 Pass

Wk 7-At Philadelphia - 9 Drives - 2 Run/7 Pass

Wk 8-Seattle - 11 Drives - 4 Run/7 Pass

Wk 9-Buffalo - 10 Drives - 7 Run/3 Pass

Wk 10-At Washington - 14 Drives - 4 Run/10 Pass

Total: 100 Drives - 50 Run/64 Pass 44% Run

Here is the breakdown by groupings:

Before you study the data below, I would recommend that if the numbers for the groupings are unfamiliar, that you spend some time reading a more expanded definition of the Personnel Groupings here.

Here is what we learned yet again. The Cowboys try to cover up the absence of a fullback by putting 3 TE's on the field "13" personnel and lining them up as if they were in "22" - with a TE lined up as FB. The theory is that Phillips is too tall (Fiammetta is 6'0, 242 and Phillips is 6'6, 262) for proper leverage as a lead blocker at the point of attack. If "low man wins" in blocking drills, perhaps there is something to this theory. Regardless, there is a stark contrast in productivity between the FB lead plays when Fiammetta is there versus a tight end. You can see that in the "13" run plays below (17 carries for just 36 yards).

Totals by Personnel Groups:
PackagePlays RunYardsRunPass

Table Tutorial

Totals by Personnel Groups on 3rd/4th Down:

Table Tutorial

Shotgun snaps are fine on 3rd Down and in the 2 minute drill. But, we track this stat from week to week to make sure the Cowboys aren't getting too lazy in using it. They are not efficient enough to run it as their base, and with a 15%/85% run/pass split, there is no way the defense respects your running game.

This game saw the Cowboys seldom in shotgun other than 3rd Down. And as we said, they were doing quite well there.


Wk 1 - NYJ: 24/66

Wk 2 - SF: 32/66

Wk 3 - Wash: 27/62

Wk 4 - Det: 29/75

Wk 5 - NE 31/67

Wk 6 - StL 10/60

Wk 7 - Phi 39/49

Wk 8 - Sea 19/59

Wk 9 - Buf 15/61

Wk 10-Was 24/73

Total - 250/637 39.2%


Overall, we would rate the Washington win as part luck (Gano missed a kick to win) and part winning the turnovers and converting key 3rd Downs. Romo saved the day, but you don't want to play like that every week against inferior competition if you want to win the division.

And please email me at Sturm1310@aol.com if you have questions that I may clarify in an upcoming email blog.

No comments: