Thursday, September 29, 2011

Denard's passing vs. offensive structure

Denard's passing numbers are down (obviously), after Al Borges brought a new "pro style" offense to Michigan. Recently, Borges took some responsibility for Denard's struggles (emphasis mine),
Well, it’s a work in progress with our offense. That’s the thing … because it’s different. Now part of that, too -- and I’m going to take the rap for that a little bit. I’ve got to get him some better throws. I’ve got to put him in position to complete some more balls so he can gain some confidence and gain some rhythm. Get in a little bit of a zone. He’s a capable passer, you know, but as a playcaller you have to consider everything we’re calling in terms of the passing game. This kid really threw the ball well in two-a-days and threw the ball well in spring. He did. All his numbers were better numbers than now. I think game situations are different. As he learns about how to do this, you’ll see progress. Because he does have a good arm, and he has an accurate arm when he’s comfortable. But part of that has to be my responsibility to get him in better situations to complete some throws.
A few months ago, I did a post about Denard's passing abilities on obvious passing downs that indicated not only that this regression would happen, but that he was probably a little overrated as a pure passer last year. The basis of that post was the definition of an obvious passing down:
The first step was to establish the criteria for OPD. The basic parameters:
  • 2nd down and 10+ yards to go
  • 3rd down and 7+ yards to go
  • 4th down and 7+ yards to go
  • All situations where over 20 yards were required for a first down were omitted
  • All first downs were omitted
Given Denard's success last year and the conclusion of that post (31/71 passing for 5.7 YPA, 0 TD, and 5 INTs), it seemed likely that Denard wasn't in a whole ton of those long distance situations last year. A quick sanity check bears this out: the goal of Rich Rodiguez's option offense was always to stay ahead of the chains and pick up first downs. The logical assumption, then is that the offense last year wasn't in many long down and distance situations, or at least Denard wasn't throwing the ball from long down and distance situations as much as he is this year.

With the help of MGoBlog's Upon Further Review series, I charted all of Denard's passes from 2010 (until the Ohio State and bowl games, neither of which received a UFR), as well as all of Denard's passes through the first three games of the 2011 season (Western Michigan, Notre Dame, and Eastern Michigan). My hypothesis was that when Denard threw the ball in 2010, he was probably throwing it in more surprising situations (e.g., 2nd and 5, 3rd and 3, etc). In other words, a defense was likely to be preparing for a run and was caught off guard. The results:

Down Comp/Att Average yards to go YPA
1st 54/76 10.39 10.16
2nd 54/87 8.21 10.87
3rd 39/69 7.54 6.7
4th 3/6 7.17 10.5
Down Comp/Att Average yards to go YPA
1st 13/24 10 13
2nd 13/21 8.14 9.86
3rd 6/15 6.53 5.93
4th 0/1 5 0

Those numbers are, well, really surprising.

There are some explanations here. Rodriguez was far more likely to run Denard in long down and distance situations. Borges seems to run Denard early in series to get to these short down and distance situations. When he does put the ball in air, there's usually less ground to cover for a first down. Another reason Denard threw in long-distance situations so much last year was because the offense took far more penalties than they have this season, necessitating longer passes.

My other inclination was that Michigan's offense had so many big plays in 2010 that the higher average yards to go were less impactful because of the team's ability to pick up chunk yardage. However, in the 11 games surveyed last season, Michigan averaged 2.45 plays of 15+ yards per game. In 2011, they are averaging 2.66 plays of 15+ yards per game. Recalibrate that for the strength of schedule (as 2011 gets more difficult, you would expect that number to drop), and you probably have a similar, if slightly lower number in 2011 than in 2010.

Unfortunately, these numbers sort of contradict Borges comments: it's not necessarily the structure of the offense that's hurting Denard's passing numbers. He appears to be throwing the ball in favorable (or at least more favorable than last year on average) situations. Given the indecision that we saw against SDSU and Denard's general accuracy problems, I find it difficult to imagine Borges finding more favorable situations for Denard to throw in. The flip side of that is that Denard's decision making is something that can be improved and probably will get better as the year progresses.


Post a Comment