Creating the Case:
I was asked by a very talented and enthusiastic young group of people. At that point, the decision problem was trivial, how could I possibly say no? I genuinely enjoy supporting student-driven initiatives, especially when they are both intellectually challenging and socially relevant. From a utility-maximization perspective, it was a dominant strategy.
Much more complicated than expected, an excellent reminder that ex ante expectations are often biased. The case had to be interesting, relevant, and solvable using a wide range of methods, given the heterogeneous backgrounds of the participants. The hardest part was not the idea itself, but making everything crystal clear and preparing a clean, well-structured dataset. Since teams only have a very limited time, there is absolutely no room for ambiguity. Confusion is not an advanced econometric technique.
What impressed our Case maker?:
The creativity was remarkable, and no two solutions looked alike. I was particularly impressed by how teams managed, in just two days, to produce something that in some cases resembled a nearly publishable paper. Thanks to intense collaboration, full concentration, and probably a heroic amount of caffeine.
Advice for the player:
First: less is more. Teams sometimes opt for extremely complex models, which increases the risk of overfitting and consumes valuable time. Time that is then missing for interpretation and clear exposition. Solving the case is important, but explaining the solution clearly is just as crucial.
Second: look at your data. Skipping proper exploratory analysis and summary statistics can lead to results that simply do not make sense. So get to know your data before you start torturing them. They tend to behave better when treated with respect.
The Econometric Game Impact:
It is a great opportunity to show some of the very best students, many of whom will eventually contribute to making the world a better place, that econometrics is not just an abstract collection of formulas living happily on a blackboard. Instead, it demonstrates that well-chosen models, combined with good data, can address concrete societal problems. In short: econometrics actually does things, and does them quite well when used responsibly.
I hope it will attract even more attention from industry, with cases and datasets motivated directly by real industrial needs and concrete questions. The ideal setup is one where the problem originates in practice, and the academic community translates it into a rigorous econometric framework. That way, everyone wins; and econometrics continues to prove that it is both theoretically elegant and practically indispensable.