smallchou.com/blog

Big anticipation among NFL fans for the Colts-Patriots game tonight. Best current rivalry in football. The game’s two biggest stars. Amazing plays all night. And an unreal finish. 35-34 Indianapolis.

I’m a big football fan and I think I’m very knowledgeable about it, but I’m certainly not qualified to question any of the coach’s decisions from a football standpoint. But from a pure decision-making point of view, I find it humorous how fans and analysts alike are ripping Belichick’s decision to go for it on 4th-and-2 from the Patriots 30. Just check out all of the “Belichick’s so DUMB” comments on ESPN.

It was an aggressive decision, but was it unreasonable? Here are the pertinent facts from before the play:

1. Patriots were up 6 points.
2. Down was 4th-and-2 on their own 28-yard line.
3. Patriots had moved the ball pretty much at will for most of the game.
4. Peyton Manning had just led his team on two LONG touchdown drives that each took less than 2:08. The last one was 1:37. He hadn’t used a timeout on either of those drives.
5. It was the play immediately before the two-minute warning, so no booth challenges.
6. Patriots had no timeouts (had used the last two before the play), so no challenges for their benefit, period.

There is a “conventional” decision, which is definitively to punt in this situation. It’s the “safe” decision (for some reason). Almost every coach in the world is going to punt (apparently save for one). Should they? I would argue that simply saying, “You’re supposed to punt” is dumb.

As any good poker player will tell you, what matters is making the best decision with limited information. Football has the additional component of execution (whereas ‘execution’ in poker usually just means the random sequencing of the cards), but that doesn’t mean the right decision shouldn’t be made. How’s the decision supposed to be made?

1. What is the probability that you will convert that 4th-and-2? If you convert, let’s assume you win.
2. If you don’t convert, what is the probability that you can stop the Colts from scoring a touchdown?
3. What is the probability that you will stop the Colts from scoring a touchdown if you punt? Let’s assume that if the Colts do that, then the Patriots lose.

If you aren’t considering those questions, even on a gut or impulse level, then you’re not doing your job as a coach. Let’s place some estimates.

1. Let’s say the Patriots think they convert that 4th-and-2 60% of the time.
2. Let’s assume the Patriots think they have a 10% chance of stopping the Colts if they don’t convert.
3. Then, consider that the Colts are moving the ball *really* well. It’s not outside the realm of possibility that they have, say, a 40% chance of taking the punt and scoring a TD to win the game. I don’t think that’s totally crazy.

If you take those numbers into account, then it’s clear what you should do: go for it. 64% of the time, you’ll win by going for it. 60% of the time you’ll win by punting.

Now I’m not saying those are the numbers. Maybe a 40% chance of Peyton scoring is crazy high. Is it? I don’t think it is. And hey, maybe a 60% chance of converting a 4th-and-2 is crazy low. After all, over the last three years on 3rd and <3 or 4th and <3, Tom Brady is 23 of 29 (79.3%) for 221 yards.

At the end of the day, I think Belichick actually made a pretty reasonable call. I think it was the right call, even. If I was coaching the team and only worried about winning the game (i.e. not worried about losing my job), I would have made the same call. And if he was playing poker and lost the hand, he would be just considered unlucky.

Now, if you want to get on Belichick and Brady for wasting two timeouts before the play, meaning they could not challenge the spot, then be my guest. But, given Belichick’s usual decision-making, I think this can just be chalked up to results not reflecting the decision. After all, as all good poker players know, good decisions are not correlated to short-term results.

And by the way, I come out of that game (especially after also watching his press conference) liking Belichick a lot more. He clearly thought they had a better than 80% chance of converting, so he went for it. That’s belief in the system and his players.