Sunday, August 19, 2012

Rocky Long's idea is intriguing to me

I stumbled across a mind-blowing story about San Diego State coach Rocky Long the other day and started a reactionary post but then totally forgot about it. Derp.

Anyway, an excerpt from said story:
Call it Rocky Long’s “Great Go-For-It Experiment.”

After reading articles about an idiosyncratic Arkansas high school coach who never punts, always onside kicks and has tremendous success doing it, Long is toying with the idea for his Aztecs of no punts or field goal attempts once they've driven inside an opponent’s 50-yard line.

Conceivably, San Diego State would go for the first down whether it needed a couple of inches or 10 yards.

And yes, Long — who apparently hasn’t yet tried it all in his 40 years of coaching — is serious about this.

“It makes sense,” he said, seeming almost giddy in talking about the possibilities. “Additional plays would allow you to score a lot more points,” he said. “It also puts a whole lot of pressure on the defense.”
Words do not do this revelation justice; pictures/videos are necessary.

Astonishment indeed.

I'm not sure whether I'm more amazed that somebody might actually do this or that nobody has done it before. I suppose the latter should be expected considering that most of the guys in charge grew up in an era when the forward pass was controversial and probably still think that way because the media requires them to; freakin' Bill Belichick (whose job security on a scale of 1 to 10 is about a 17) got totally destroyed a couple years ago for going for it on a fourth-and-2 in his own territory even when it totally made sense and the game itself meant little. But most media people are dumb and should be disregarded as such.

The reason the former is worth consideration even in the face of public OUTRAGE: the numbers. Some really smart people have run a lot of numbers to figure out some Moneyball-type stuff and have come to the determination that going for it on fourth down usually makes more sense than gaining 35-ish yards of field position, especially when field position dictates that going for it likely would result in an immediate increase in the odds of scoring points.

David Romer's piece -- which focused on NFL numbers -- is the most famous one and resulted in coaches calling him an idiot despite (a) the aforementioned numbers and (b) Romer himself saying that the numbers were basically just intended to be a reference, kinda like the odds every pro poker player knows by heart and presumably considers before deciding whether to put a million bucks on his/her hand of choice.
Romer said that the goal of his study was not to “replace coaching decisions with a computer” but to inform their understanding of the game. “You ought to know when you are making these decisions what the statistics say, on average,” he said.
I agree; Ron Zook probably doesn't.

Anyway, here's one (relatively) easy-to-read graph that was put together recently by The Mathlete over at MGoBlog based on college data:

This maximizing-win-likelihood graph is based on a hypothetically average offense, defense and special teams and pretty clearly indicates that punting from inside the 50 is almost always a bad idea unless it's a fourth-and-long from outside the 40; that's it. Field goals are a little more ambiguous but are typically the way to go from between about the 10-yard line and the 25.

BTW, regarding the requisite "winning by three at the end of a game" rabble: No kidding. Common sense should prevail in situations in which various other factors (such as time or extreme weather conditions) have to be considered. Again, these are statistics that are meant to provide odds/information and not necessarily make the decision in all circumstances.

As for this particular case study, San Diego State probably doesn't have a hypothetically average offense, defense and special teams. The offense and defense actually were pretty close to average last year and should be again this year, but the kicker and punter will be spectacularly named true freshman Seamus McMorrow, who is (a) a true freshman and (b) a true freshman. Expecting any consistency or guaranteed level of quality would be unwise. FYI, San Diego State had its first scrimmage Saturday, and McMorrow didn't attempt a field goal and averaged 36.2 yards a punt, which isn't good (36.2 would have been somewhere outside the top 120 nationally last year).

So in that regard, going for it a lot might just be preferable to sending out a punting game that's crappy or a kicking game that's a total mystery. It also might not be an improvement: San Diego State reportedly went 2 for 14 on fourth-down attempts over the first two days of scrimmage-type settings in fall camp. It probably goes without saying that 14 percent would be an awful number that would pretty quickly put an end to any desire to avoid kicking the ball since it would render the "average" probabilities irrelevant. I'm gonna bet they won't go 14 percent for the season on fourth downs, though; they were 103rd nationally last year at about 39 percent (9 for 23) and are likely to go up just due to reversion to the mean.

Regardless, this soooooooo needs to happen, both for the entertainment value and for a data point that includes the results of all those drives with a fourth down somewhere between the 50 and the goal line. And it sounds like it might actually happen since Long does what he wants with his hot body:
Whatever he decides, Long said it won’t be because he doesn’t want to take the heat.

“It’s has nothing to do with the armchair quarterbacks and how they view this, I promise you,” Long said. “I don’t do anything anymore that I don’t think is best for the football team. I never worry about the critics.

“When I first became a head coach, I think I did. I don’t anymore. I’m an old guy. I’ve been coaching a long time. I’m mature enough to know better.”
Rocky Long FTW. There's also this:
Long is no mathematician, and he said he can’t decipher the data, just the conclusions – which he’s not sure he trusts. He said he still would like an unbiased SDSU math professor to take a look at the data and offer an opinion.

But seriously: Rocky Long FTW. Do it, man.


Post a Comment

Powered by Blogger.