Paragraph of the day, red zone playcalling edition

Red Zone Play-Calling

On a first down Red Zone play, teams are more likely to score if it’s a run than a pass if they are at the 8 yard line or closer. Anything between the 9 and the 20 favors a pass on first down. That doesn’t mean that 100% pass is the optimal strategy, just that the play calling should favor the pass (or run inside the 9). For goal to go situations after first down, second down is the ultimate OC’s choice. From anywhere 10 and in on second and goal running and passing have nearly identical touchdown percents. On third and goal, the run still holds up strongly. A called run is more likely to score a TD on anything from the 6 and in than a pass, which owns 7 and up. Again, not saying the strategy should be 100%, but there is real value to favoring the run inside the 7.

There is more data driven situational analysis here.

First down means everything

According to an exhaustive study of NFL play data conducted by Yale professor Cade Massey, what happens on first and 10 in an NFL game is a powerful indicator of who will win.

According to Dr. Massey, an assistant professor at the Yale School of Management, the ability of an NFL team to meet certain benchmarks on each down is one of the best predictors of whether a drive will be successful. When it comes to first down, he says, the magic number is four. That’s the number of yards Mr. Massey says teams need to gain on first and 10. Those that do, he says, are more likely to be successful in making a first down and keeping the drive alive.

. . . The four teams in Sunday’s playoffs have different approaches to first down—and rates of success. The Pittsburgh Steelers, who play the New York Jets Sunday for the AFC title, run the ball 55% of the time on first and 10 in the first three quarters of a game. (The fourth quarter isn’t included in the calculation because play-calling can be largely dictated by whether a team is ahead or behind). These runs by the Steelers rarely catch the opposing defense by surprise. . .  [T]he Steelers are daring the opposing team to try to beat them in a head-on collision.

. . . The Steelers have only managed to attain that magic mark of four yards or more on first downs 48% of the time—a number that puts them in the bottom half of the NFL. So far, the Steelers have been able to compensate for their lack of success on first downs by connecting on long passes on other downs. They’ve also excelled in the reverse role: Pittsburgh’s defense is the best in the league on first downs, holding opposing teams to less than four yards about 59% of the time.

. . . According to the numbers, the Jets are similar to the Steelers in that their first-down defense is better than their first-down offense. . . .

The Green Bay Packers rank No 2. in the NFL this season in recording successful plays on first and 10. The team also likes to pass on that down and distance 54% of the time—more than any other team in the playoffs and all but four in the NFL.

When Green Bay played the Atlanta Falcons in the divisional round of the playoffs, the Packers managed to run successfully on first and 10 early in the game with running back James Starks. But later in the game they began lining up in running formations—sometimes with two tight ends—and instead running play action passing plays. Those plays faked out the Falcons and led to big gains.

So who is likely to prevail? Green Bay is by far the best first-and-10 team left in the playoffs. But their opponents, the Chicago Bears, are also one of the best defenses at stopping offenses in that situation. . . .

Read the rest here. This is not a shocking result, but it’s possible to draw the wrong conclusion. I think the wrong answer is to pick plays that have extremely low variance at the expense of expected gain — i.e. the plunge into the line that, while it rarely loses yards, doesn’t average much, with the thought that you just want to avoid negative plays and want to get close to that four yard gain. As the chart below indicates, your probability of getting a first down in three plays depends far more on your expected gain than it does the variance.

chart

And, of course, game theory is relevant because you might significantly improve both your expected gain and variance on first down with simple consistent gainers, like runs off tackle, quick passes, and so on, by taking high variance chances, like play-action or some other kind of play that can significantly keep the defense off balance.

All that said, I think the upshot of Dr. Massey’s analysis is that most first down playcalling is not good, and too often puts the offense in a bad spot. If you show me a team that is good on first down, I’ll show you a good offense. Indeed, the best offenses look at it like they are playing under Canadian rules: if you only have two downs to get a first down, you approach the problem quite a bit differently. Third down shifts the burden away from the defense to the offense; better to avoid as many third downs as possible.

Smart Notes 1/18/2010

1 Is it possible for a defense to be “good against the run” or “good against the pass,” or is it merely good, mediocre, or bad? Chase Stuart, in two excellent posts heavy on the game theory (available here and here), shows that, at the very minimum, it’s difficult to say anything meaningful about a defense other than to comment on its general effectiveness; the two phases are too inextricably intertwined. For fans and commentators I think this is correct, though from a gameplanning perspective it remains possible to identify which defenders are most dangerous and what is most difficult to accomplish, not to mention whether the defense is tilting to the pass or run — i.e. extra defensive backs or guys in coverage, or extra run defenders.

2. Survivor bias on the gridiron. From the Freakonomics blog.

3. Tim Tebow’s loping release. During the broadcast of Florida’s bowl game, Brian Billick showed exactly what is wrong with Tebow’s release: It’s long, he brings the ball down too low (this motion generates no additional power or accuracy), and it exposes the ball both to a fumble and to a defender who might break on the ball. See it here (h/t Doc Sat):

The word I had gotten was that Scott Loeffler, Florida’s quarterback coach, had made significant progress with Tim on this but that come gametime, well, a player’s gotta play how he knows how. And Tebow had earned the right to play his way. Yet it is troubling to the lack of progress, and it will hurt him in the draft. But what if it was worse, than a lack of progress — what if Tebow actually regressed on this point? Check out this video which charts Tebow’s release over time, and you be the judge.

4. “Football Island”:
(more…)

Spurrier wants balance: Is he asking the right questions? Are his critics?

Steve Spurrier watched the game film of his offense’s horrible performance against NC State and concluded: we weren’t aggressive enough. And people are ridiculing him for it.

Steve Spurrier has watched the entire N.C. State game twice and part of it a third time.

The South Carolina coach reached two conclusions: The Gamecocks were too conservative offensively in their 7-3 win in Raleigh, and such an approach is not going to cut it this weekend at Georgia.

“We had a pretty conservative game plan. I didn’t realize how really conservative it was until I watched the game twice now – almost three times,” Spurrier said Sunday. “We wanted to give the running game a chance, so we did do that. But we obviously need to try for some big plays along the way a little bit more probably.”

USC’s run-pass ratio in the opener was nearly 2-to-1, with the Gamecocks running 42 times and attempting 22 passes (although some of those rushes were scrambles by or sacks of quarterback Stephen Garcia).

Still, the attack looked much too plain for a coach credited with introducing the SEC to an intricate downfield passing attack in the 1990s.

And while Spurrier is not ready to scrap the Gamecocks’ revamped rushing scheme after one game, he made it clear he wants to see a more balanced attack against Georgia.

“We certainly can’t bring that game plan to beat Georgia on offense. I don’t think we can,” Spurrier said. “But we don’t want to send Stephen back there and get sacked and run around all night either. We’ve got to get us a balance between runs and passes that we can hit and look like a good offense.”

The buzz has been that Spurrier must be nuts — hey, he’s already given up on the run game. But look at the numbers. I’ve previously talked about a notion of “balance” that only looks at the number of runs or passes or the total yards with rushing and passing as being misleading, and that a far better metric is comparing the expected — or, in lieu of that, average — yards per attempt of each, though, since passes are riskier than runs, passes should still average more (have a premium). The reason is because the defense will respond to your playcalling; it’s a game theory thing.

So let’s look at the numbers. Overall, the Gamecocks averaged a measly 2.57 yards per rush, and an okay 6.7 yards per attempt, though with an interception. There can be problems at looking at the raw numbers, particularly on third down where the result is binary: convert or fail to convert. So let’s look at first down, where clearly the optimal strategy is to maximize your expected gain.

The sample is small, but on first down South Carolina ran the ball 16 times and averaged a mere 3.06 yards per carry. They threw it nine times for 78 yards (and no INTs), resulting in a very healthy 8.67 yards per attempt. I can safely say that Spurrier should have called more first down passes. The OBC’s instincts are right. His playcalling was too conservative, at least on first down, which is the most important down in football because there are more first downs than any other down.

This post is about game planning and play-selection

Suppose first that you wish to cross a river that is spanned by three bridges. (Assume that swimming, wading or boating across are impossible.) The first bridge is known to be safe and free of obstacles; if you try to cross there, you will succeed. The second bridge lies beneath a cliff from which large rocks sometimes fall. The third is inhabited by deadly cobras. Now suppose you wish to rank-order the three bridges with respect to their preferability as crossing-points. Your task here is quite straightforward. The first bridge is obviously best, since it is safest. To rank-order the other two bridges, you require information about their relative levels of danger. If you can study the frequency of rock-falls and the movements of the cobras for awhile, you might be able to calculate that the probability of your being crushed by a rock at the second bridge is 10% and of being struck by a cobra at the third bridge is 20%. Your reasoning here is strictly parametric because neither the rocks nor the cobras are trying to influence your actions, by, for example, concealing their typical patterns of behaviour because they know you are studying them. It is quite obvious what you should do here: cross at the safe bridge. . . .

[Now s]uppose that you are a fugitive of some sort, and waiting on the other side of the river with a gun is your pursuer. She will catch and shoot you, let us suppose, only if she waits at the bridge you try to cross; otherwise, you will escape. As you reason through your choice of bridge, it occurs to you that she is over there trying to anticipate your reasoning. It will seem that, surely, choosing the safe bridge straight away would be a mistake, since that is just where she will expect you, and your chances of death rise to certainty. So perhaps you should risk the rocks, since these odds are much better. But wait … if you can reach this conclusion, your pursuer, who is just as rational and well-informed as you are, can anticipate that you will reach it, and will be waiting for you if you evade the rocks. So perhaps you must take your chances with the cobras; that is what she must least expect. But, then, no … if she expects that you will expect that she will least expect this, then she will most expect it. This dilemma, you realize with dread, is general: you must do what your pursuer least expects; but whatever you most expect her to least expect is automatically what she will most expect. You appear to be trapped in indecision. All that might console you a bit here is that, on the other side of the river, your pursuer is trapped in exactly the same quandary, unable to decide which bridge to wait at because as soon as she imagines committing to one, she will notice that if she can find a best reason to pick a bridge, you can anticipate that same reason and then avoid her.

The above passage is from here. Can you explain in what way this informs play-calling and gameplanning? Here’s an (incomplete) hint.