Great analysis! My sense is that PA can work in the red zone, but needs to be run on 1st down. After that, it's a waste.
Pull up the NCAA official stats and Michigan’s red zone efficiency looks great, ranking third with scores on 93% of trips. Brendan Gibbons had a lot to do with that as Michigan connected on 14 field goals in 46 trips. But as tends to happen in these situations, the truth is much more complicated the NCAA would have you believe.
After the concept of fumble luck, 3 <> 7 may be the second statistical pillar of MGoBlog. The NCAA’s stat does not believe what we believe. Their rankings are based on a simple equation:
[Times scoring in the Red Zone]
[Trips to the Red Zone]
For the NCAA 3=7. An equally simple measure that has been strangely ignored is Points Per Trip (PPT). By that measure (and taking out meaningless second half trips and games against FCS teams), Michigan drops to 44th at 5.2 PPT in 36 qualifying trips.
Red zone Efficiency is a very easy stat to overreact to. The sample size is small and a couple of fluky plays can swing the ranking either way. When you expand the study beyond the end result of the trip and look at the 110 individual plays that comprised Michigan’s 2012 red zone offense, there is at least a little more sturdy basis for evaluation, although the smaller the sample set, the more likely there is a large piece of luck involved in any outputs, whether positive or negative.
To evaluate each play I looked at the touchdown percentage for drives at each possible possible down, distance and yardline from inside the 20. Every play either makes the offense more or less likely to score a touchdown on the drive. A first and goal from the 1 yard line results in a touchdown on the drive 91.4% of the time, therefore a touchdown is worth 8.6%. Second and goal from the 1 results in a touchdown 87.3% of the time so getting stopped on first down is worth –4.1%. Each play is evaluated based on its impact to Michigan’s chances of scoring a touchdown on the drive. Even though the odds of a field goal dropped slightly as you move back within the 20, for this study I just wanted to focus on the effect on potential touchdowns.
Michigan ran 43 first down plays on their qualifying red zone trips last season and put themselves in a situation more likely to result in a touchdown on 47% of them. Even though their plays were slightly more likely to be negative than positive, the positive plays had a higher magnitude, resulting in a net positive of about 52%, or half of a touchdown.
Second down was where the problems started. Michigan ran 39 qualifying second down plays in the red zone and only 14 of them bettered their chances of reaching the end zone. Michigan finished at –221% on second down, a loss of over two touchdowns due to poor second down performance.
Michigan actually held up well on second down rushes, improving their odds on 12 of 23 second down rushes. The problems were centered around second down passing. After the Robinson to Gardner touchdown on the first 2nd down red zone pass of the season, Michigan went 0-9 with 2 sacks on the next 11 pass plays. Michigan quarterbacks locked into Devin Funchess and Jeremy Gallon in these ill-fated situations as the were targeted on 7 of the 9 incompletions. The incredibly surprising play action was not the only issue, only 2 plays were noted as play action in the UFR’s and another 3 were listed as waggle or rollout, but one of those was the initial touchdown.
Where Michigan struggled on second down they excelled on third down. Michigan got a first down or touchdown on 16 of 28 third down plays and reversed their second down loss with a +324% change in their touchdown odds on third and fourth down. Michigan’s binary down success was largely driven via the pass but the situation greatly changed when Devin Gardner came on for Denard Robinson. Denard was 1-5 with a sack on third down while Devin Gardner went 5-5 (all for first downs or touchdowns) with a sack. Where the second down plays were focused on two different players, Gardner third down passes were to 4 different players on the five completions.
Gardner’s third down prowess continued on the ground with a +122% rating on five third down red zone carries. The lack of confidence in the traditional running game around the goal line was evident as only 4 of 13 red zone carries on third and fourth downs were taken by running backs. Toussaint and Vincent Smith both went 1/2 on their attempts.
So Devin Gardner was pretty good in the red zone. Over all plays he was +432%, or over 4 touchdowns added over the course of the season. In fact, Gardner’s success was probably unsustainably good. I don’t have touchdown’s added for all players, but if you look at pure points added in the red zone, Gardner’s five game red zone average was the second best season ever to Tim Tebow’s 2007 Heisman season. Gardner is really good in the red zone but it is going to be very tough to sustain this level for a full season, only one player ever has.
But what about the other Wolverines?
The only other Michigan player to finish with a positive number was Justice Hayes, by a hair. Hayes’ singular red zone carry against South Carolina netted him a 2% increase. Among the other running backs, Thomas Rawls was –12%, Vincent Smith was –66% (although he was actually the most valuable receiver) and Fitzgerald Toussaint was –117%. All three were making positive plays less than 50% of the time.
Denard finished with a slightly negative red zone contribution for the season, with –39% but on a team low 39% positive ratio. As mimicked by his career, Denard showcased a lot of valuable game changing plays in the red zone, but struggled with consistency. In the end, his 2012 red zone negatives outweighed his positives.
On the receiving side, targets of Vincent Smith, Jeremy Gallon, Drew Dileo and Devin Funchess all finished on the positive side while Roy Roundtree was the sole receiving target to end with a negative rating with pair of 3rd and Goal targets from the 7 falling incomplete.
As noted above, red zone efficiency is fickle stat and can easily swing. With that said, based on small sample size splits, here are some pros and cons heading into the season.
The two biggest things that seem like more than just fluky outcomes of limited play counts are the success of Devin Gardner in the red zone in both running and passing and the failures passing the ball on 2nd down. Some of this is due to the incredibly surprising play action, 5 of the 12 UFR’d plays where listed as PA, rollout or waggle, but the other six plays weren’t any better.
At this point I have no clue how to keep my expectations for Devin Gardner on earth. There are lots of sample size issues with only five games under his belt but those were five pretty spectacular five games from him and he was at his best in the highest leverage situations. I don’t think he can do it for a whole season and hopefully the defense and running game mean he doesn’t have to, but man, that guy made a lot of plays last year.
Great analysis! My sense is that PA can work in the red zone, but needs to be run on 1st down. After that, it's a waste.
Your mention of "Points Per Trip" (PPT) and scrubbing out meaningless FCS numbers got me thinking.
Is there such a thing as a "Points Per Trip When It Mattered" metric. That is, a metric to measure red zone success when Michigan is either behind, tied, or has an insufficient lead to assure a win.
I guess that's a further refinement of the "scrub out FCS numbers." It would also scrub out B1G games where Michigan has the game in hand.
There's a certain opportunity cost to not scoring when the game is on the line.
In this study and most of them I do, I include all plays for the first half of the game but only plays in the second half if the drive starts or ends within two touchdowns. Depending on defintions could tweak the When It Mattered sentiment but this should be about what you were looking for.
You make a great point about the usual way of calculating red zone effeciency being inadequate. My feeling is that if Michigan can get production out of the RB position this year, look out. Gardner will be one of the most dangerous and effective weapons in college football.
I agree 100%
I am also more bullish on Gardner than most and I think he will be a heisman finalist... so I think his arm will be the difference and the running backs will have room because of that...
I had Queens of the Stone Age in my head the whole time while reading this
The funny thing about everyone saying that Gardner's stats being "unsustainable" is that we're talking about a season in which he wasn't even playing QB half the year and had to switch midseason. Under normal circumstances, you'd expect him to improve, if anything.
The 2nd down struggles remind me of the 2nd half flailings in the Ohio game, in which Borges apparently ran some plays in order to set up other plays which he never got the chance to run because the Buckeys were stymying M. The problems with 2nd down plays in the red zone may be related to attempts to display tendencies that, perhaps, cause defenses to make errors on subsequent downs.
In terms of Devin Gardner's crazy '+ value' in the red zone, how much can be attributed to his own badassness and how much can be attributed to the fact that opposing coaches did not yet know how to game plan for Devin&Michigan due to the Denard-to-Devin switch?
Sorry for the run on sentence. Last year, Devin had two advantages in his starts. 1) He was a new starter and, thus, old team tendencies would be broken and 2) Opposing teams still had to spend time and practice preparing for the possibilty of Denard playing QB.
As much as I want to, I'm not sure I can embrace all the crazy Devin hype yet. He had several factors working in his benefit during those starts. Even in the bowl game prep, no one outside of THE TEAM knew how much Denard could and would do, much to the benefit of Devin as quarterback.
...about pure red zone %, as you have, and cannot stand not accounting for FG v. TD. After all, when a defense is defending the red zone, allowing a FG is considered success.
That said, I don't think PPT is good, either. Football is a field position game, and is as much about consistently getting the ball into the opponent's red zone more often than you allow them into yours.
For example, Arizona only scored 4.82 PPT, significantly less than Michigan. However, Arizona made 78 (!) trips to opponents' red zones, and thus averaged 38 PPG. Michigan made a measley 46 red zone trips in 2012, so even if we scored 8 points on every single trip, that would've yielded 368 red zone points compared to the Wildcats' 376.
On the flip side, Wisconsin yielded 5.375 PPT and was dead last in FBS in defensive red zone efficiency with opponents scoring 94% of the time. However, they only allowed 32 trips to the red zone, which equates to 172 points allowed. Bucky gave up 19 PPG last season.
Perhaps red zone total point differential, or even simply the difference in the number of trips, would be more indicative of red zone success or failure. For example, Alabama made 62 red zone trips while allowing a paltry 29--a difference of 33 trips. Even if the Tide scored 3 on every single one of their possessions and allowed a TD on every one of their opponents, that's only a deficit of 17 points for the entire season.
Over 14 games, Alabama scored 352 points in the red zone while allowing just 120; a difference of 232 points total and 16.57 red zone PPG! (As an aside, Alabama allowed just 153 points all season, meaning they gave up on 33 points from outside their red zone, or 2.36 PPG. Talk about limiting big plays.)
I agree that red zone only tells part of the story. It's a stat designed to balance out total yards. Moving the ball up and down the field and scoring in the red zone can be two very different things, but I don't think it's a deceiving stat. I think most people take it for what it is. People don't look at rushing stats and say that they mean everything and neglect passing stats. Like most stats, red zone efficiency is a measurement of a team's success in a specific facet of the game, I hope I didn't give the impression that this was any more than that.