Frank Frigo

You've got a lot of active decisions you have to think probabilistically.

I would say that the uncertainty component is more realistic of what people encounter in the business world and in social decisions.

Chess is a great game, but chess, if you make...

The best move, it's the difference often between a 100% winning path and a 100% losing path.

Whereas in backgammon, it's more probabilistic.

And I think that is more like the decisions we encounter in real life.

We built proprietary models for a variety of applications in sports, in commodity trading, in health care, in education.

The idea was to take a game-based approach to solving difficult problems.

There's a lot of data out there.

but there's not a lot of good ways to take that data and steer it towards a particular objective.

So in commodity trading, it might be ROI.

In education, it might be how to reduce freshman attrition rates.

In healthcare, it was how to reduce hospital readmissions for cardiac patients.

And then, of course, in sports, one of the big, big breakthroughs that we applied was we took a lot of ideas from the modeling of games

and applied it to professional football.

Originally, when I was just watching the game as a fan and I was a pretty seasoned backgammon player, I would look at what appeared to be the inherent risk aversion bias of head football coaches.

They would get to fourth down decisions

and they would sort of pucker up and relinquish possession.

You're saying they're very risk-averse.

At least they used to be before we came onto the scene.