Living La Vida Woody
I chart the Cowboys games, looking for play calling motifs, organizational quirks, situational matchups, etc. Back in week two, when the Cowboys lost their shootout home opener to the Giants 33-31, I made a small notation in the margins that Jason Garrett was calling a lot of "heavy" sets, most notably the 22, a two running back, two tight end, one receiver formation.
In week three, I noted that Garrett was going heavy in a similar way, using the 13 set, with a single back, three tight ends and a lone receiver. What I didn't do was break down the game tapes by packages, noting which personnel groups made the field, and with what frequency, so my notations lacked context. Was the reliance on heavy sets an oddity, something Garrett did for just those two games? If he did, he was foolish to abandon them, since Dallas gained over 200 rushing yards in each contest.
Fortunately, my buddy Bob Sturm, a host at 1310 The Ticket in Dallas, shares the tape-analyzing sickness. During the season, Bob produces weekly charts, which he calls the "Decoding Garrett" series, which breaks down how many times the Cowboys use each of their various personnel groupings in a game. You can see one of them here.
The collected charts give us a comprehensive look at Garrett's philosophy. How does he like to attack opponents? Is he a mad bomber in the Mouse Davis or Patriots-era Josh McDaniels mode? Is he a modern day Al Davis acolyte, working to blend a power running attack with a deep-drop, down-the-field passing game? Is he a balance freak, who mixes and matches runs and short and intermediate throws, in the style of his predecessor Sean Payton? Is he none of the above? All of the above?
In part three of my series on the running attack, I catalog Sturm's individual breakdowns into one big chart, which demonstrates rather clearly that Garrett took Wade Phillips' request to run the ball more very seriously, and that he did so by using a running style more familiar to Woody Hayes than to Norv Turner, Ernie Zampese, Tom Landry or any of the other people to carry a play chart on the Dallas sideline.
The 30/40/30 Break
The Cowboys used eight different personnel sets week-to-week last year, and spinkled in a handful of plays from exotic sets like the wild cat or a 5 WR set or a no WR set in some cases. The team used thirteen different looks all told. How then do you classify how Garrett attacked opposing defenses?
I break the formations into three types, which are classified by the number of receivers they employ. I classify "standard" sets as any which use two receivers, and either one back and two TEs (the 12 set) or two backs and a lone tight end (the 21).
I define "spread" sets as any which use three or more wide receivers, with the intention of overwhelming opposing secondaries. On the other side of the strategy spectrum are what I term "heavy" sets, meaning they have just one or no wide receivers. These are the 22 set, with two backs and two TEs or the 13 set, which puts three TEs on the field to either block or receive, around a single back.
A breakdown of the raw play totals produces the following results, from 993 scrimmage plays*
| Package Type | Number of Plays Run from Type | % of Total Offensive Plays |
| Spread | 337 | 34% |
| Standard | 385 | 39% |
| Heavy | 271 | 27% |
*Sturms total of 993 plays is 7 plays more than the official number of 986. I suspect this is due to charting a few plays which were called back by penalty. Since those seven plays make up less than one percent of the total, and therefore won't skew the tables in any meaningful way, I'm going to work with Bob's figures.
The raw numbers show a fairly restrained offense, from a personnel grouping perspective. Dallas goes to spread sets roughly one third of the time. Of equal interest is the surprisingly high percentage of heavy formations Garrett deploys. How many of you expected to see Dallas going with only one wideout almost 30% of the time?
More odd is that this appears to be Garrett's preferred methodology, using base sets 40% of the time and then balancing the 60% of his remaining calls between heavy and spread formations. Don't believe me? Let's look at some of Dallas' games, and compare the spread-versus-heavy splits:
- Giants I: 20 spread plays; 19 heavy plays
- Panthers: 19 spread; 19 heavy
- Chiefs: 14 spread; 13 heavy
- Falcons: 12 spread; 22 heavy
- Seahawks: 20 spread; 20 heavy
- Eagles I: 18 spread; 17 heavy
- Redskins I: 18 spread; 20 heavy
- Raiders: 13 spread; 19 heavy
- Chargers: 18 spread; 18 heavy
- Saints: 20 spread; 23 heavy
- Eagles II: 22 spread; 20 heavy
In nearly every game, Garrett called roughly the same number of heavy formations as spread sets. Three games on the schedule skewed heavily towards the pass. In Denver, the running game was thwarted and Dallas abandoned the run for the last 20 minutes of the game, trying to catch up. The same thing happened in the Meadowlands, where defensive and special teams breakdowns put Dallas in a deep hole. Romo threw 55 times that day in a vain attempt to bring Dallas back.
The most lop-sided passing game, the one and only '09 Cowboys game plan which can be termed a full spread attack, was used against the Packers. Garrett called some plays from spead sets which got receivers open early, but Romo missed his throws. Garrett kept doubling down on the pass and by day's end produced this line:
Spread - 40 plays; Standard - 12 plays; Heavy - 6 plays
That line also produced Dallas' ugliest offensive output of the season, a 17-7 loss which was nearly a shutout. There are several commonalities between the three spread-happy games. First, and most important, all were losses. Next, they were far less effective than the more balanced game plans outlined above: Dallas scored an average of just under 14 points per game in those three contests. It scored just over 24 points per game in the other thirteen.
If we adjust for those three games where the score got out of hand, or where Garrett lost his sense of equilibrium, we get the following breakdown for the other 13 games, an 11-2 stretch which saw Dallas lose a game by two points and the second by three points after failing to score from a 1st-and-goal situation on the Chargers one yard line.
- Spread plays -- 241
- Standard plays -- 308
- Heavy plays -- 236
That's almost a symmetrical 30-40-30 distribution of plays. (It breaks down 31-39-30 if you want to be real precise.) If you look at the overall pass-to-run ratio in those 13 games, where the defense kept the score tight and Garrett could stick with his preferences, you see Dallas threw 53% of the time and ran 47% of the time. Those are far more balanced numbers than '07 when Garrett called runs 44% of the time and especially better than '08, when the run-to-pass ratio was 42% to 58%.
Phillips asked for more runs and Garrett delivered. He did so by frequently removing a receiver and adding an extra blocker, and telegraphing his intent. Garrett wasn't cute about his heavy sets: he called runs 77% of the time when he used the most common 22 package and 66% of the time when he inserted the 13. The Cowboys ran a neat 75% overall from heavy formations. Defensive coordinators knew what was coming -- but it didn't matter.
Average yards per rush:
- Combined standard and spread sets: 4.4 yards/attempt.
- Heavy sets: 5.4 yards/attempt.
The Limits of Heavy Running
5.4 yards per rush will rank your offense near the top of the league, if not at the top. Given the success of heavy running, why didn't Garrett do it more? A little reflection should produce the answers. The Cowboys line in its '09 incarnation, was mistake-prone. Flozell Adams took his copious share of penalties. So did Leonard Davis, who ranked just behind Adams in most penalties taken in the decade. Tight end Jason Witten, despite his Pro Bowl pedigree, draws an inordinate amount of flags and took almost as many as Adams in '09. Recall how his back-to-back flags sabotaged the opening drive of the wild card game.
Garrett calls an aggressive game plan because calling six or seven straight runs will put him, somewhere along the line, in a 1st-and-20 or a 1st-and-15, which pushes the heavy sets on the field and brings the receivers on.
Dallas also lacks the across-the-line muscle to run all facets of a power game. Dallas runs draws, counters and tosses very well, but has struggled for a while now on simple isolations, dives and off-tackle runs. Think of the two 4th-and-short failures at Washington. Recall the inability to convert against San Diego. The Cowboys successfully used power I attacks in places like New Orleans, but they could not do it consistently.
In some ways, the heavy sets are an admission of Dallas' o-line foibles; Garrett is, in effect, doing with eight blockers what Turner and Zampese's offenses could do with seven. Until the Cowboys get better guard play, the game plans Garrett unveiled last season are likely to continue, because the Cowboys ran effectively from their heavy formations most of the time. I doubt that Phillips, with his increased attention to defense and special teams, will ask Garrett to run any less in 2010.
If you looking for a shorthand synopsis of this coming year's running game, it's probably this: "we're heavy. What are you gonna do about it?"
Thursday: What can we conclude from Dallas' "heavy" run tendencies?