Parimutuel system of the Japanese race track: Voting and ranking process

Abstract

We investigated a time series of betting rates in a pari-mutuel betting system of Japanese horse races.  Principal component analyses and cluster analyses suggest that the level of expectation for the favorite is the most important factor when classifying races. The dataset of the odds determined before the races used in this study is exceedingly large and detailed compared to previous studies. We also conducted detailed research on whether bets placed on the favorites tend to be concentrated just before the race.