Belichick Decision was Correct Convention Wisdom is Sometimes Wrong

When the New England Patriots went up to the line last night on 4th and 2 from inside their 30 yard line, I thought for sure that they were trying to draw the Colts offside and could not believe when they snapped the ball. I told my son and wife that it was the worst coaching decision in history — in any sport!

Most of the sports analysts I have heard last night and this morning seem to agree, although when I look at it today from an analytical perspective, perhaps it wasn’t such a bad decision after all. In poker, which I play quite a bit, we often apply math to various scenarios. The math is only one factor that must be taken into consideration when making a poker decision, but it helps establish a base decision making criteria.

Many of you want to quickly dismiss an analytical approach to making this decision, but in truth, most football decisions come down to playing the percentages. The same football analysts who are dismissing the decision to go go for it on 4th and 2 are using math to justify their dismissal of the move. Their arguments usually go something like… can’t he trust his defense to stop Manning and the Colts? They would have had 70 yards to move down the field and Patriots had been stopping them most of the time. Most stands for a % in these analysts’ minds. Is it 40%, 50%, 60%. It has to be something, right? They are also making the mistake of applying math to some of the decision and not all of it — sure NE can stopping the Colts some significant % of the time, but how often will they make it on 4th and 2 and even if they miss, how often will the Colts score from 30 yards out? These are all factors that must be looked at. If not, why not?

Let’s apply some percentages to the various alternatives last night. These are just best-guesses, so feel free to use your own.
% of times that NE makes the 4th and 2: 75%
% of times that the Colts drive 70 yards and make a touchdown in 2 minutes with 2 timeouts (assumes a net 40 yard punt): 60%
% of times the Colts drive 30 yards and make a touchdown in 2 minutes: 70%

So given the above,
if NE goes for it they win .75 + .30*.25 = 82.5% (.75 they make it + 25% of the time they don’t make it but win anyway, because the Colts don’t score 30% of the time from 30 yards)
If they punt, they win only 40% of the time, since we are assuming that the Colts score 60% of the time.

In the above scenario they win twice as much as they lose if they go for it. Now you might bring down the 75% of the time that NE makes it on 4th and 2 and bring down the % of time that the Colts score from 70 yards, but I still think that NE wins much more than they lose given any reasonable odds. Here are some different %’s — harder for NE to make it and harder for the Colts to score from 70 yards, and easier for the Colts to score from 30 yards.
% of times that NE makes the 4th and 2: 60%
% of times that the Colts drive 70 yards and make a touchdown in 2 minutes with 2 timeouts: 40%
% of times the Colts drive 30 yards and make a touchdown in 2 minutes: 90%

If NE goes for it they win .60 + .40*.10 = 64% If NE punts they they win 60% of the time

So even here it is slightly better for NE to go for it, and most of us will agree that NE will make 2 yards more than 60% of the time and the Colts will drive 70 yards in 2 minutes for the win more than 40% of the time AND the Colts don’t score from 30 years all the time (if you disagree with this last point, keep on reading).

CONVENTIONAL WISDOM
I was listening the Colin Cowherd radio show on ESPN this morning and he was as dead set against the decision as I initially was. He dismissed any mathematical analysis related to 4th down efficiency and argued that the situation dictates the decision. Sure it does, but what is the situation and what are ALL the factors related to the situation? Colin, you can’t have it both ways. If you want to argue the situation, let’s discuss all the factors related to the situation, not just some of the factors.

So for example, Cowherd stated that teams would always punt from their 1 yard line if it was 4th and 1 and would often go for it from their opponent’s 1 yard line if it were 4th and 1. Why is that? Because if you are on your own 1 during the course of the game and miss, you are almost guaranteeing your opponent will score 7 points if you miss. But the bigger factors in my opinion, are: 1) even if you make it you are still on the 2 yard line and have to make more first downs; and 2) during the course of a game, your opponent will often settle for a field goal and will not go all out to score a touchdown. So in the normal scenario from your 1 yard line, you will hold your opponent to 3 points a large % of the time when you punt vs. guarantee them 7 points if you miss on 4th down.

Let’s take a look at this same scenario at the end of the game. Of course we could never have 4th and 1 from the 1, but what if we did? What if the Patriots were facing 4th and 1 from their 1 last night? If they make 1 yard they win the game. If they punt they likely give Manning the ball between 30-40 yards out and 2 minutes to cover the yardage. Most of you were conceding that it was virtually a sure thing for Manning to cover 30-40 yards in 2 minutes right? Well, if that is true, wouldn’t it be better to give Brady the chance of winning the game by getting 1 yard? How could it not be? Even if the %’s of success are as low as 30-40% and I’m sure they are MUCH, MUCH higher, isn’t that better than giving the ball to Manning when you know he is going to drive 30-40 yards in 2 minutes? Or wait a second? Is it not a certainty that he will move 30-40 yards?

Hey, guys, you can’t have it both ways. Sometimes, the conventional wisdom is just wrong.