Abstract
The Tour de France is the world’s biggest cycling event. The race attracts up to 25 million TV viewers per stage worldwide. In this article, we forecast TV audiences for individual stages of the Tour de France for five European countries where cycling is popular: Belgium, Denmark, France, The Netherlands and Spain. The predictions follow from on a multivariate ordinary least squares regression model that explains historical viewing habits for the Tour de France as a function of attributes of the individual stages, and contextual information such as TV channel and day. Although the accuracy of the forecasts changes from year to year and can be very different between TV markets, in most cases our predictions clearly outperform forecasts based on naive models. Our findings illustrate that a large part of the variation in TV viewership is determined by how the race route is designed by the race organizer, independent of actual race developments.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
