An economic analysis of best-of-five ATP matches from 2004-2014

Sunday, October 11, 2015: 9:40 AM
Darius Conger, Ph.D. , Economics, Ithaca College, Spencer, NY
Lilyana Mladenova , Sport Management and Media, Ithaca College, Ithaca, NY
This paper compares the  success determinants of Association of Tennis Professional (ATP) players over the decade 2004-2014  in best-of-five matches occurring during  Grand Slam—Australian Open, Roland Garros, Wimbledon and U. S. Open—tournaments, the most prestigious and lucrative tournaments on the professional tennis circuit. While talent differentials, physical superiority and past successes are among the factors distinguishing the two match-type outcomes (3-set matches vs. 5-set matches), the psychological aspect of competition may also be a substantial contributor to success in lengthy tennis matches. Thus, the objective of this study is to extend previous findings by accounting for unanticipated match occurrences and the players' ability to react and respond to the situations, thereby providing empirical verification for the possibility of talent differentials being confounded with psychological abilities.  If confirmed, this study will re-assess the confounded contributions of talent and physical characteristics to the probability of winning matches by isolating the psychological aspects within those characteristics.  Additionally, the inclusion of all ladder matches gave us an opportunity to make some inferences about both upsets and blowouts in light of strong playing fields and substantial prize money differentials between tournament ladder rounds.

To accomplish the above, we utilize match information from the (ATP) web site, http://www.atpworldtour.com/Scores/Archive-Event-Calendar.aspx, and the commercial product Tennis Navigator, http://www.tennisnavigator.com/ , from which we compile measures of player talent, physical characteristics, career success overall and against specific opponents and develop measures of psychological pressures occurring within matches.  Statistically, we employ an appropriate variant of regression models which allows for certain player characteristics which vary over our data period to be appropriately accounted for in 5-set matches, both over time and between tournaments.

JEL Code:  Z21

Alternative JEL Code:  Z20