Forecasting National Medal Totals at the Summer Olympic Games Reconsidered

Abstract

Objective: This article aims at explaining national medal totals at the 1992–2016 Summer Olympic Games (n = 1,289 observations) and forecasting them in 2016 (based on 1992–2012 data) and 2020 with a set of variables similar to previous studies, as well as a regional (subcontinents) variable not tested previously in the literature in English. Method: Econometric testing not only resorts to a Tobit model as usual but also to a Hurdle model. Results: Most variables have a significant impact on national team medal totals; it appears to be negative for most regions other than North America except Western Europe and Oceania (not significant). Then, two models (Tobit and Hurdle) are implemented to forecast national medal totals at the 2016 and 2020 Summer Olympics. Conclusion: Both models are complementary for the 2016 forecast. The 2020 forecast is consistent with Olympic Medals Predictions, although some striking differences are found.