Posts
The first was the one you only cared about if you play fantasy football. Maurice Jones-Drew broke through the defense and took a knee at the 1 when the Jets were trying to allow him to score. By doing so, he allowed the Jaguars to run the majority of the remaining time off of the clock and kick the game winning field goal, as opposed to being forced to give the Jets the ball back with a chance to score. Mild argument ensued, but most people were more amused than upset. The Jaguars still won, after all, and 20 yard field goals are about as automatic as it gets.
The other has been discussed and dissected to the point of exhaustion. The Patriots, up 6 with just over two minutes to play, were facing 4th down and 2 to go. Instead of punting, the Patriots went for it and completed a pass for 1½ yards. The Colts scored a touchdown with little trouble, and the game was over. Every sports publication known to man had a take on the incident. It was Coach Belichick’s hubris. The Patriots didn’t trust their defense. Peyton Manning could have scored that touchdown no matter how far he had to go. In response to the largely incredulous mainstream media, the stat heads countered with the probability of victory for each decision. The verdict can be seen here, but the results are fairly simple: going for the first down gave the Patriots a 79% chance to win, and punting gave them a 70% chance to win. The counter-arguments from the players turned analysts were predictable. “Where did they play football?” “You can’t go for it, what if they don’t get the first down?” “I don’t care what the numbers say, I played football, and you have to trust your defense.” Or, my personal favorite, “It didn’t work, so it obviously wasn’t right.”
The topic of whether or not it was the correct decision has been beaten to death elsewhere, so I’ll mercifully spare you another diatribe about expected value or believing in your players. The situation was, by any measure, essentially a coin flip. The math supported the chosen decision, but not by an overwhelming margin. Yet, despite the comparable expected value (sorry, had to use it once) of each decision, I’m very confident that choosing to punt, regardless of the outcome of the game, would have brought coach Belichick a great deal less scrutiny. Football insiders, as well as fans, have by and large been resistant to changes in the statistical understanding of their game. Why, though, has this been the case? Why has the numbers revolution in sports eluded football?
Because football is a game we love for its subjectivity. You cannot measure how hard a tackle is, but you get awfully excited when a safety blows up a receiver going over the middle. Runs are enjoyable for their length, but just as much so for their style and power. The distance is measured more by phylum than specifics: a few yards is a decent run, 5-10 good, more than 15 long, more than 25 really long. Pass plays are short, medium and long. We know the exact yardage, but it only matters for first downs. Who cares if he got 4 or 6 if he only needed 3? Great football plays aren’t 3 pointers or home runs; they’re spectacular catches for first downs, tackles to prevent third down conversions, blocked field goals.
And yet, as much as the greatness of the plays, and the players, eludes measure, the circumstances are always measurable. There is always a down and distance, every play goes for a specific number of yards, the time left in the game always visible. Thus, the game becomes the ultimate expression of what all sports are: a mixture of the objective and subjective. We understand the game as a clash of titans, their collective actions immeasurable by any numbers that attempt to describe them, couched within the incredibly measurable and objective structure of a football game.
And somewhere along the line, the average fan never developed the ability to correlate the quantifiable with the human. We know the Patriots only needed two yards, but it seemed risky. We know the Jaguars could run the rest of the time off the clock, but damn it, you score the touchdown when you can. The fact that we have always measured the plays, the fact that we know every percentage and distance for every situation for as long as football has been played, and that we can extrapolate from these numbers the most correct course of action for nearly any situation does nothing to dissuade us from the notion that football is a game of people, not numbers.
Did this realization dissuade me from trying to advance the cause of the statistical revolution? Of course not. It just meant that I had to find an answer. What could make people embrace the increased knowledge of how to play a football game correctly? What could help people realize that the statistics aren’t something evil, designed to ruin their wonderfully subjective battlefield? In an instant, it occurred to me. Only in football could such a bizarre solution be true.
The players are, in fact, playing the hard hitting, scrappy, intangible dogfight we love so much. Nothing about this has changed, and nothing should. It’s the coaches that need to get their act together and embrace change.
How can this be the case? Because football players don’t make tactical decisions. They run the plays, and attempt to execute them correctly, with varying degrees of success. Their decisions are limited, as would be expected when 11 people are attempting to act as a coherent unit. This is distinctly unlike, say, basketball, where every player has the opportunity to shoot, pass, or attack their defender whenever they touch the ball. They’re certainly coached, and acting within a system, but ultimately they have to realize what decisions are good and bad.
Lets take an example of statistics changing the strategy of each sport. In basketball, it has recently been established that the corner (baseline) three-point shot is close to the most efficient shot in basketball. A good coach will tell his good shooters to take these shots, and his other players to pass to those players when they are in position to take the corner three. Ultimately, though, if the players do not set themselves up for these shots, the blame lies with them.
Conversely, football metrics focus on decisions made by the coach. If the down and distance is 4th and 2 yards to go, the players can’t make an inefficient choice; they will either punt as directed, or run the offensive play called for them. It is not their job to choose run or pass, punt or go for it, field goal or not, or any of the other decisions that drastically affect the outcome of games. It is their job to do what the coach tells them to do as well as possible.
Which brings us back to Belichick and going for it on that fateful 4th down, and the only thing I really want out of a football coach: common sense. Yes, it’s scary to go for it on 4th down, but it was 2 yards. The Patriots get two yards on accident. I’m sure the punt looked attractive, but Belichick had the confidence to trust the percentages and give his team the best chance to win. The outcome wasn’t the probable one; sometimes it won’t be. But he did what was right, and the players did the rest. If he punted and lost we might not have judged him as harshly, but that would be wrong. That would be the bias of the misinformed, confusing our love of the human element with outright ignorance. We wouldn’t like the game as much if coaches punted on third down, or never passed. I hope that someday we feel the same way about coaches who punt with impunity and ignore history. Coaches can change while the players stay the same. The game we love will still be there. We can still have the magic of intangibles if we embrace the math.
This is quite a liberating realization. We are free to enjoy the subjectivity of the game because we have compartmentalized it away from the objectively correct decisions behind the game. We can appreciate the nerds who pored over mountains of data involved in determining the win probability of every decision in every circumstance, then watch 22 grown men beat the hell out of each other to see what actually happens. We can watch our players fight like hell and love them, then berate our coach for stupidly punting. We are now free to enjoy the most beautiful reality of sport, and in many ways life itself: the idea that trial and error, experience, and objective fact can take us far, but our results will ultimately depend upon things like intuition, heart, and a myriad of other intangible qualities that will supplement our knowledge and give us our best chance of success. Our coaches' reason and our players' passion can coexist, distinctly separate and yet intertwined. We can respect the nerds and root for the jocks.
Or maybe I just need to yell at Belichick and ask the poindexters where they played football. This isn’t rocket science, after all.
Image via The New York Times
You can email Chris with questions or comments at TheSportsKiosK@gmail.com
Migrating comment over from Reader...
ReplyDeleteFootball more than any other sport eludes statistical definition. Of course, there is definitely a lot you can decide on stats, nobody denies this. But the influence that momentum yields in football is incomparable (I can't prove this obviously, just going off anecdotal experience).
Baseball is a series of individual battles, with almost each encounter being separate from the one immediately prior, as far as personal interaction goes. Sure there are hot and cold streaks, but after a few years, you get a pretty good idea of what player you have.
Basketball just figured out the stats phenom...and really, it's only the Rockets who have figured it out collectively. Someone shared an article not too long ago that it was cheaper and more efficient to hire a bunch of 3 point shooters than down low bangers. And teams are doing just that.
But football...at least the NFL...more than any other sport do thing change rapidly. The Cardinals almost won the Super Bowl last year. THE CARDINALS. Adrian Peterson was Purple Jesus JUST ONE YEAR AGO, now he's a massive liability. RBs come and go like the seasons.
I'm not saying hardcore statistical analysis shouldn't be used. It should be, and already is. But coaches aren't pure generals. They aren't moving pieces. They know the players, they know how they feel, and where they are going. Belichick is a bit of a savant as far as strategy goes, and so was Bill Walsh. But both coaches would tell you right away how much feel goes into their decisions in a game. Bill Walsh invented scripting plays for Christ's sake, and he would never argue the value of momentum and emotion in a game.