Scoring bias caused by services in table tennis: a statistical analysis

Abstract

In table tennis, servers have the potential advantage of scoring bias when serving. However, the length of the scoring bias, i.e. the shot number where any bias is eliminated, has not been clarified. This study aimed to clarify the level and length of scoring bias occurring at services in table tennis. We analysed 45 men’s singles matches (3,665 rallies) and 29 women’s singles matches (2,352 rallies) from the 2012 London Olympic games and 49 men’s singles matches (4,404 rallies) and 31 women’s singles matches (2,320 rallies) from the 2016 Rio Olympic games. The statistical analysis revealed that services generate a low scoring phase at the second shot and slightly high scoring phase at the third shot. Moreover, the level of the scoring bias became trivial after the third shot, although a small scoring bias could remain. Players should therefore be cautious of a scoring bias until the third shot. In the gender comparison, the scoring bias observed in matches of male players was larger than that of female players up to the third shot. This result indicates that male players are more likely to take advantage of service than female players. In the winner/loser comparison, it was found that losers use the service to create scoring bias as effectively as winners do. Losers’ inferior skills and tactics in the shots after services were the major factor in the difference in the occurrence of missed shots between winners and losers. Finally, we found that the performance of each shot number should be analysed separately up to the third shot, as the remaining effect of the service is remarkably different among shot numbers. The results of this study contribute important suggestions to the conventional methods of performance analysis that empirically separate a rally into three phases.