Evaluating Strange Forecasts: The Curious Case of Football Match Scorelines
35 Pages Posted:
Date Written: February 23, 2019
This study analyses point forecasts for a common set of events. These forecasts were made for distinct competitions and originally judged differently. The event outcomes were low-probability but had more predictable sub-outcomes upon which they were also judged. Hence, the forecasts were multi-dimensional, complicating any evaluation. The events were football matches in the English Premier League. The forecasts were of exact scoreline outcomes. We compare these with implied probability forecasts using bookmaker odds and a crowd of tipsters, as well as point and probabilistic forecasts generated from a statistical model suited to predicting football match scorelines. By evaluating these sources and types of forecast using various methods, we conclude that forecasts of this type are strange. We argue that regression encompassing is the most appropriate way to compare point and probabilistic forecasts, and find that both types of forecasts for football match scorelines generally add information to one another.
Keywords: Forecasting, statistical modelling, regression models, prediction markets
JEL Classification: C53, L83, G14, G17
Suggested Citation: Suggested Citation