Can Positive Thinking Fuel Economic Booms?
A study of data from 17 countries provides a nuanced answer.
Can Positive Thinking Fuel Economic Booms?Tobias J. Moskowitz: So how did I become interested in sports?
Robert B. Gramacy: Oh, well I love sports.
Matt Taddy: Students were so interested in it.
Tobias J. Moskowitz: I grew up in Indiana, therefore by law you have to love college basketball, which I do.
Richard H. Thaler: I took out my own subscription to Sports Illustrated when I was about 12.
Matt Taddy: I kept finding it as a tool for teaching things.
Robert B. Gramacy: I’ve always wanted to see if I could apply some of the techniques that I know in statistics to making sense of sports data.
Richard H. Thaler: If I can make people believe this is actually work, then I get paid, which is great.
Narrator: Professional sports is a global multibillion dollar industry. The Dallas Cowboys franchise alone is thought to be worth some $2.3 billion. With so much money, not to mention pride and fans’ energy on the line during every game of every season, there’s a lot of pressure to deliver wins.
But here’s news for owners, coaches, and players out there: some of your biggest fans are researchers who have metrics, models, and theories you definitely want to know about. They have some suggestions for how to help you win. Or, to help your competitors win. Unless you like losing, you might want to hear them out. How do you build a winning team? Can you do it without a superstar? Is one second-round draft pick better than two third-round picks? How can you measure individual player performance in a team sport such as hockey? Faculty at the University of Chicago Booth School of Business have been tackling these questions for years.
Tobias J. Moskowitz: People always say “there’s no I in team,” right? You need a team to win a championship. The individual can’t do it on his own. And the classic example is Michael Jordan with the Bulls, where for years he was toiling away in Chicago, scoring 63 points in that famous game against the Celtics, and then they ended up losing year after year, and it wasn’t until Scottie Pippen emerged and they had some other players like Horace Grant and Oakley and other people who came in that they were able to win.
So people always point to examples like that and say, “See? It’s the team that matters, not the individual.” Well, again, this is something that can be answered with data, so we put this to the test. It turns out that if you want to win a championship, you pretty much need at least one superstar, if not two. The last 23 out of 24 NBA [National Basketball Association] championships have included one of the following six players, which is Magic Johnson, Larry Bird, Shaquille O’Neal, I think it’s Hakeem Olajuwon, Lebron James, and Kobe Bryant, right? Six players, right? If you don’t have a top-five player, someone who’s year after year in the MVP voting category, you might make the playoffs, but you’re not gonna go much further than that.
I think the statement that “there’s no I in team” is uttered the most for basketball. Now, you might ask, why do people say this if it’s not true when you look at the numbers? A lot of it has to do with incentives. If you knew that all the mattered was getting the ball to Michael Jordan and you could just sit back a watch, your effort level’s gonna go down. Why try so hard if it’s clear that you don’t matter that much?
Narrator: So, you need stars on your team. But how do you recognize who is actually a star? Are the current metrics any good?
It turns out that some metrics are very imperfect. But researchers are inventing better measures.
Robert B. Gramacy: The question that we had in hockey analysis was: Could we work out how a player contributes to the success of his team? And the simplest way to do that is to think about scoring goals, because goals are what you need to win games. And you need to prevent goals also to win games. We tried to find the simplest model we could that jointly entertained all of the players’ on the ice contributions to goals being scored for or against teams. Plus/minus is probably the simplest metric for tabulating this information. It gives a +1 to every player who is on the ice for the goal-scoring team, and a -1 to everyone on the ice on the goal-scored-against team.
Matt Taddy: I could be the worst possible player, ever. I am not a very good hockey player at all. And if I’m on the ice next to someone who’s a fantastic player, and I just happened to always be on the ice next to them, my plus/minus is gonna look fantastic, because that guy’s a really good hockey player, and he’s gonna score even though I suck and I’m skating into the boards in the opposite direction and that sort of thing, right?
So the issue that we said is: Can we build a model that says, OK, what are the effects on whether this goal was for or against our team?” OK, conditional on not just me being on the ice, but me and everybody else who I’m skating with on the ice?
And then we also include things like a goal for is more meaningful when you are the penalty kill, for example. So we include the fact of what sort of special-teams info there was there. There are just other things, you know? Some teams have plenty of other effects that make them better or worse than other teams. The coach, things like that, so we include effects for team. So what’s left over at the end is the contribution. We call it the partial contribution of an individual player after removing or controlling for the effect of all his teammates, his line mates, things like that.
Narrator: Building a better metric may be new to hockey, but it’s been done in baseball. The sport has what’s called sabermetrics, applying statistical analysis to baseball records in order to evaluate and compare players. [Financial journalist] Michael Lewis wrote a book about this called Moneyball. It was later turned into a movie staring Brad Pitt. So could the same idea work on the ice?
Robert B. Gramacy: The take-home message there is that it’s possible for a team on a low budget to compete with teams with much higher budgets. And I think that that’s true in hockey too. We found a lot of examples of potential for forming teams on very low budgets that would be at least as good as the best teams. What I would like to do is to be able to simulate games, and that’ll help understand the odds that teams will win games vis-à-vis the odds that they will score goals or prevent goals from being scored.
Matt Taddy: We’re saying there’s only about two-thirds of players that we can really measure as being better or worse than average. However, that doesn’t argue for not using statistics and data in hockey. What it argues for is putting the knowledge that you have into building better models beyond the baseline that we’ve provided.
Narrator: But let’s say better metrics help you identify the best players. You still have a challenge. How do you get the best players to join your team at a good price? Once again, research offers suggestions. Look at the NFL draft, the annual event at which National Football League teams select their players.
Richard H. Thaler: We’ve discovered that there’s something that in the league they refer as “the chart.” And the chart is simply a table that lists the value of each pick. It turns out all the teams use this chart, so if I have the 10th pick and I want the first pick, the chart will tell me what combination of picks I have to give to get up to that first pick. All the draftniks now have the chart in front of them when they’re watching the draft, which has become not only a three-day televised event but a movie starring Kevin Costner.
And so what we were trying to do is say, well is that chart correct? What we found is the most valuable picks in the draft, most valuable meaning the quality of the player they get minus what they have to pay the player is maximized somewhere early in the second round. Now this is an enormous anomaly, because you can trade the first pick for five early second-round picks, and according to our analysis, each of those picks you get is worth more than the one you gave up. We’ve worked with three different teams over the years, and they sometimes listen to us, but oftentimes the owner will fall in love with some player and then all the numbers go out the window and the owner gets what he wants.
Narrator: Richard Thaler’s idea in context is this: yes, stars are important, and metrics help reveal the best athletes, but when it comes down to picking players, you still shouldn’t put all your hopes on one person. Or even a few.
Richard H. Thaler: If you ask me, “Which is the best quarterback prospect?” I would say, “I have no idea.” But if you’re sure you know, I’m willing to bet against you.
Narrator: Some teams may not be willing to apply the research—yet. But analytics is becoming more and more prevalent in professional sports, and movies such as Moneyball and Draft Day are making the public more aware of the data that produce wins.
(sports announcer voice) We may be just at the beginning of a sports analytics (normal voice) revolution.
A study of data from 17 countries provides a nuanced answer.
Can Positive Thinking Fuel Economic Booms?Should the government and insurance companies subsidize travel to healthcare centers?
Why Medical Tourism Could Be Good PolicyThe podcast examines economics’ influence on public policy in the US.
Capitalisn’t: Has ‘Thinking Like an Economist’ Distorted Our Politics?Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.