Brian Burke

And, you know, they put all the stats online now. So this is like 2006. I said, hey, you know what? We can just download the data. And by the end of lunch, we can answer this question forever. And that was the genesis of, you know, football analytics for me. When I began doing this, I hadn't read Moneyball. I didn't know that existed. It was an advantage because...

And, you know, they put all the stats online now. So this is like 2006. I said, hey, you know what? We can just download the data. And by the end of lunch, we can answer this question forever. And that was the genesis of, you know, football analytics for me. When I began doing this, I hadn't read Moneyball. I didn't know that existed. It was an advantage because...

And, you know, they put all the stats online now. So this is like 2006. I said, hey, you know what? We can just download the data. And by the end of lunch, we can answer this question forever. And that was the genesis of, you know, football analytics for me. When I began doing this, I hadn't read Moneyball. I didn't know that existed. It was an advantage because...

The baseball people tried to put it onto football for a long time. The kind of tools and the kind of analysis just doesn't work on football. I came from this military background, and I'm like, this is war. This is zero-sum, two-player game theory. And that paradigm took hold.

The baseball people tried to put it onto football for a long time. The kind of tools and the kind of analysis just doesn't work on football. I came from this military background, and I'm like, this is war. This is zero-sum, two-player game theory. And that paradigm took hold.

The baseball people tried to put it onto football for a long time. The kind of tools and the kind of analysis just doesn't work on football. I came from this military background, and I'm like, this is war. This is zero-sum, two-player game theory. And that paradigm took hold.

There's this optimization element to it. In the same way in the military, you have a mix of strategies. It's not like always do this or always do that. You have to be unpredictable in a way that keeps your enemy or your opponent on his heels. There's a famous thinker in military aviation named John Boyd who invented this thing called the OODA loop, if you've ever heard of that. And

There's this optimization element to it. In the same way in the military, you have a mix of strategies. It's not like always do this or always do that. You have to be unpredictable in a way that keeps your enemy or your opponent on his heels. There's a famous thinker in military aviation named John Boyd who invented this thing called the OODA loop, if you've ever heard of that. And

There's this optimization element to it. In the same way in the military, you have a mix of strategies. It's not like always do this or always do that. You have to be unpredictable in a way that keeps your enemy or your opponent on his heels. There's a famous thinker in military aviation named John Boyd who invented this thing called the OODA loop, if you've ever heard of that. And

Keeping the enemy confused and disoriented and in a state of ambiguity is one of the goals in American fighting theory. Football works the same way.

Keeping the enemy confused and disoriented and in a state of ambiguity is one of the goals in American fighting theory. Football works the same way.

Keeping the enemy confused and disoriented and in a state of ambiguity is one of the goals in American fighting theory. Football works the same way.

I was able to build something called expected points and expected points added. It's a point expectancy model based on down distance and yard line. Once I built that model, the very first thing I did was just aggregate by play type and And it was very, very clear at that moment that passing was far superior to running. Teams are running far too often.

I was able to build something called expected points and expected points added. It's a point expectancy model based on down distance and yard line. Once I built that model, the very first thing I did was just aggregate by play type and And it was very, very clear at that moment that passing was far superior to running. Teams are running far too often.

I was able to build something called expected points and expected points added. It's a point expectancy model based on down distance and yard line. Once I built that model, the very first thing I did was just aggregate by play type and And it was very, very clear at that moment that passing was far superior to running. Teams are running far too often.

And the way you know that is because if they're doing each in the optimum mix, the payoffs would equalize. There would be what people commonly refer to in game theory as the Nash equilibrium. As long as you have an intelligent opponent, you can assume that that equilibrium is going to be the optimum mix. And they were far out of whack. From that moment on, we knew that you need to pass more.

And the way you know that is because if they're doing each in the optimum mix, the payoffs would equalize. There would be what people commonly refer to in game theory as the Nash equilibrium. As long as you have an intelligent opponent, you can assume that that equilibrium is going to be the optimum mix. And they were far out of whack. From that moment on, we knew that you need to pass more.

And the way you know that is because if they're doing each in the optimum mix, the payoffs would equalize. There would be what people commonly refer to in game theory as the Nash equilibrium. As long as you have an intelligent opponent, you can assume that that equilibrium is going to be the optimum mix. And they were far out of whack. From that moment on, we knew that you need to pass more.

What year was this? 2008 is when I first did this, but it took years to permeate the football world. It was a slow process.

What year was this? 2008 is when I first did this, but it took years to permeate the football world. It was a slow process.