Bill Belichick Is Great

I respect Bill Belichick more today than I ever have.

Last night he made a decision in the final minutes that led his team the New England Patriots to defeat. It will likely go down as one of the most criticized decisions any coach has ever made. With his team leading by six points and just over two minutes left in the game, he elected to go for it on fourth down on his own side of the field. His offense failed to get the first down, and the Indianapolis Colts promptly drove for a touchdown.

He has been excoriated for the choice he made. Everyone seems to agree it was a terrible blunder.

Here is why I respect Belichick so much. The data suggest that he actually probably did the right thing if his objective was to win the game. Economist David Romer studied years worth of data and found that, contrary to conventional wisdom, teams seem to punt way too much. Going for a first down on fourth and short yardage in your end zone is likely to increase the chance your team wins (albeit slightly). But Belichick had to know that if it failed, he would be subjected to endless criticism.

If his team had gotten the first down and the Patriots won, he would have gotten far less credit than he got blame for failing. This introduces what economists call a “principal-agent problem.” Even though going for it increases his team’s chance of winning, a coach who cares about his reputation will want to do the wrong thing. He will punt, just because he doesn’t want to be the goat. (I’ve seen the same thing in my research on penalty kicks in soccer; it looks like kicking it right down the middle is the best strategy, but it is so embarrassing when it fails that players don’t do it often enough.) What Belichick proved by going for it last night is that 1) he understands the data, and 2) he cares more about winning than anything else.

So hats off to Bill Belichick. This decision may have hurt his chances for the Football Hall of Fame, but it guarantees his induction into the Freakonomics Hall of Fame.


Truthseeker1

You also must ake into account how many yards were needed for a first down; how good the Colts' defense was; how good the Pats' rushing was etc., etc.

With less than 2 minutes left in the Chargers-Eagles game yesterday, the Chargers were up by 5 points and were quite deep into the Eagles territory. I silently urged the Chargers to go for it, but Norv Turner, who will never be considered a great coach had his team kick a (successful)field goal. They were up byu 8 points with a little over a minute to go. Before the kick, 3 things could have gone wrong--the kick blocked and maybe run back for some yards if not a touchdown; the kickoff run back for a touchdown or at least a lot of yards; and/or the Eagles scorea touchdown and the 2 point conversion to tie. The first did not happen; the second went to about the Eagle's 35 yard line, and McNabb completed 2 quick passes that brought the Eagles inside the Chargers' terriotory. A long last minute pass was intercepted in the endzone. Chargers won by 8, but in this situation might have been safer trying the 4th down play.

And by the way--Bill Belichick is not great. HE IS EVIL!!

Read more...

Kevin

Agree that the decision was probably correct. (Actually the data I've seen suggest it was a close call - either going for it or punting gave the Patriots about equal chances to win - about 75%).

However, I would hesitate to put him in any hall of fame. The man cheats to win.

Zach

What Belichick proved by going for it last night is that 1) he understands the data, and 2) he cares more about winning than anything else.

I fully agree with number 2 - that's why he doesn't care about cheating.

But, did Belichick proved that he understands the data by going for it? You would have to assume that he read Romer for that. Even though Belichick has a degree in Econ, and he is known for having being well-read in the situation, it does not prove that he understood the data.

If it proves anything, it proves that he didn't trust his defense to stop a rolling Peyton Manning in this particular scenario. It also shows a bit of arrogance (coupled with ignorance) - assuming that his team would never be in the situation of 4th and 2 and needing a timeout so he could challenge the last play before the two minute warning.

On the other hand, given that he likely didn't expect to have to challenge the last play before the warning (and that was a reasonable assumption - what is probability that you will have to challenge any given play?), the last timeout was probably used wisely.

Read more...

Joe Ranft

If you take this along with the decision by Rex Ryan to let the Jaguars score, and then the decision by the Jaguars not to score, we could be seeing the beginning of a major shift in the use of probability in football decision making, much like what you have seen in baseball over the past 10 years or so.

It seems to me Belichick looked at three simple outcomes:

First down = 100% likelihood Patriots win
Punt= 30% likelihood Patriots win
Failed 4th Down = 20% likelihood Patriots win

Kyle Robinson

I'm not familiar with Romer's study, but I would hope that it takes into account where on the field the ball is when the decision occurs. Going for it on 4th and short from the 50 yard line carries less risk than going for it on 4th and short from your own 20 yard line.

Ben D

It was not clearly the right decision, as you suggest. How much data is there on the '09 Patriots vs the '09 Colts during the final minutes of close games?

Nor was it clearly the wrong decision, as the football analysts suggest in hindsight.

It was a judgment call. There was no way to know the various probabilities a priori.

Hunt

Is #17 Eli Peyton's brother?

Leo

OK, but what signal did this send to his defense? Belichick outsmarted himself and no one else.

T

Has Romer done any studies on classless behavior after games? Belichick currently holds an unprecedented record in that category.

Dan C.

I'm a big fgreakanomics fan, but this is one of those classic statistical sports truths that don't hold up in reality ... just taking 4th down data and applying it here isn't enough ... you also need to factor in where they were on the field, at what the point in the game and how much time is left, who the opponent is (Peyton Manning) ...

the other thing is, you don't know *why* Bellichik made the decision ... was it in fact that "that 1) he understands the data, and 2) he cares more about winning than anything else." or just hubris ...

This reminds me of when statisticians say that baseball players don't 'get hot', and basketball shooters don't get 'in the zone', that it's just normal probablility running it's course ... anyone who has ever been 'dialed in' knows that isn't true - you can feel it, yu know you are going to make the bucket, you are performing at a higher level. It may be true that over the course of a year, a player will have better and worse performing periods, and they will average out ... but it doesn't mean that the player isn't playing better or worse in those periods ...

Read more...

Aaron K.

If going for it on 4th and 2 was the right call, then throwing it on 3rd and 2 was the wrong call.

JCB

I don't think this takes into consideration that the Patriots did not need a first down to win, they needed to prevent the Colts from scoring to win. In that case compare odds of 1) Pats making it on 4th down, 2) Colts scoring from Pats 28 yard line and 3) Colts scoring from own 20. I think only then can you tell the right decision.

an indy fan

I agree. Great decision!

murphstahoe

While you are analyzing yesterday's NFL action, analyze Maurice Jones-Drew taking a knee at the one setting up a FG to win the game on the last play, vs. scoring a sure touchdown and giving the ball back to the Jets. If Scobee is a sure thing from 25 yds, it's the right play. And if your kicker isn't a sure thing from there, you have bigger problems.

The best part of it was the move was made by a PLAYER - not a coach, in the heat of action. Which probably actually comes down to good coaching. The coach of the day yesterday was Del Rio.

Matt

Don't be silly! That type of aggregate data, averaged over a broad range of situations, is only useful as a general suggestion to give more consideration to the option of going for it on 4th down. In any given situation, you have so much more information (e.g. opponent's ability) available than that which is taken into account by the statistical analyses.

Brian

I don't agree with the call, even though I do agree that coaches punt too much, and punt way too much on fourth-and-short. Still, in a game in which ultra-conservative, fraidy-cat tactics seem to be predominant (are you reading this, Jim Tressel? Norv Turner? Dick Jauron? Andy Reid?), what a breath of fresh air this was!

Pat Palmer

Everyone's missing the real story here: the accumulated guilt of winning 3 Super Bowls by CHEATING! finally becomes too much to handle for Belichick. He snaps, and in a fit of madness makes the dumbest NFL coaching decision since Tom Landry let Vince Lombardi talk him into wearing Mrs. Lombardi's fur coat on the sidelines of the Ice Bowl.

WW

Jake, your point actually reinforces why it was such a great call on Belichick's part - when you're up against a great quarterback and offense, the difference in the percentage of the time that they'll be able to score an 80 yard td vs a 30 yard td isn't that big.

If you let
X = % Pats make the first down = % Pats win the game
Y = % Colts making a 30 yard td
Z = % Colts make an 80 yard td

If we assume the Pats make the first down 70% of the time (conservative), then their odds of winning the game are: 0.7 + 0.3(1-Y) = 1 - 0.3 Y

If the Pats punt, their odds of winning are 1-Z.

If we set the two equal, we get 1 - 0.3Y = 1 - Z...
0.3Y = Z

Even if we make Y = 1, as all the critics are saying, that means it still would make sense to go for the first down if the Colts have a greater than 30% change of winning. You don't give Manning a 30% chance of getting down the field with 2 minutes and time-outs to burn?

Read more...

Joseph

Let's not forget that a sample of 4th-down decisions is neither random nor uncorrelated. As someone else pointed out, Belichick did the "right thing" if and only if: he followed this strategy with other 4th down decisions, and variables such as opposing offensive strength, differing scores, etc. are taken into account.

As it is, it *looks* like a bad decision, and can only be called a "good" decision if he made *other* "good" decisions with 4th-and-short non-punts.

Felix

There is a paper on penalty kicks in soccer by Bar-Eli, Michael, Azar, Ofer H., Ritov, Ilana, Keidar-Levin, Yael and Schein, Galit
from Ben-Gurion University of the Negev and The Hebrew
University of Jerusalem in 2005 that concluded what you just said.