Erasmus university rotterdam



Download 1.62 Mb.
Page1/20
Date20.10.2016
Size1.62 Mb.
#5462
  1   2   3   4   5   6   7   8   9   ...   20

The favorite longshot bias

in ATP and WTA tennis

What is the best strategy: playing favorites or risk-love?


ERASMUS UNIVERSITY ROTTERDAM

Erasmus School of Economics

October 2014

Author: Jacky Quint

Student ID: 363710

Study Program: Master Economics and Business: Accounting & Finance

Thesis Supervisor: Prof. Dr. E.A. de Groot

Abstract



The favorite longshot bias is a phenomenon where bettors overvalue longshots and undervalue favorites. This results in the fact that a betting on favorites in sport matches give a higher return than betting on the underdogs, also called the longshots. Most literature is focused on the bias in horse racing, this is also the sport where the bias is found for the first time.

In this paper, the favorite longshot bias in the tennis sport is investigated. This paper uses a large dataset, which contains 37.220 odds from ATP and WTA tournament in the years 2009 to 2013.

The results do not show an existence of a favorite longshot bias. In addition, there is no strong evidence for a favorite longshot bias in WTA or ATP tournaments separately. Investigating the bias per years or surface does show a bias in ATP tournaments in 2011 and ATP tournaments on hard-court (outdoor), but the overall conclusion is an absent of a favorite longshot bias.

Preface

This thesis is the result of a research assignment in order to finalize the Master degree in Business & Economics (Accounting & Finance).


I greatly appreciate the help of my supervisor Prof. Dr. E.A. de Groot. I would like to thank him for all the valuable comments and feedback, but especially for his patience when I was struggling for months with choosing a subject. I also would like to express my gratitude to all others who have contributed, directly or indirectly, to this thesis.
Jacky Quint

Rotterdam, October 2014


Table of Contents


Abstract 2

Preface 3

1.Introduction 5

2.Literature review 7

2.1The early evidence 7

2.2Favorite longshot bias in other sports 9

2.3Favorite longshot bias in tennis 10

2.4Contradicting literature 11

3.Research question & hypotheses 14

4.Methodology 16

4.1Data 16

4.2Statistical research 17

5.Results 17

5.1Results ATP and WTA tournament (2009-2013) 19

5.2Results gender 21

5.3Results over the years 25

5.3.1 Results over the years - ATP and WTA tournaments 25

5.3.2 Results over the years – ATP and WTA tournaments separately 29

5.4Results surface 36

5.4.1 Results surface - ATP and WTA tournaments 36

5.4.2 Results surface - ATP and WTA tournaments separately 43

6.Conclusion 48

7.Discussion 50

7.1 Contribution 50

7.2 Limitations & further research 50

References 52

Appendix 54

A.Overview of ATP tournaments 54

B.Overview of WTA tournaments 57

C.Historical winners Australian Open 1950-2014 59

D.Historical winners Roland Garros 1950-2014 60

E.Historical winners Wimbledon 1950-2014 61

F.Historical winners US Open 1950-2014 63




  1. Introduction

The introduction of this master thesis will introduce the research topic and current state of knowledge. The research question and the research methodology of the study will be briefly presented. The main empirical findings and the contribution of the study will discussed shortly. Finally, the structure of the thesis is provided.


Griffith (1949) mentions the favorite-longshot bias for the first time. The favorite longshot bias is a phenomenon where bettors overvalue longshots and undervalue favorites. This results in the fact that a betting on favorites in sport matches give a higher return than betting on the underdogs, also called the longshots. In the early literature, there is a lot of investigation on horse racing (Ali 1977, Arsch & Malkiel 1982 and McGlothlin 1956). Later, there was more investigation according to the bias in other sports, like football (Andrikogiannopoulou & Papakonstantinou 2011, Cain, et al. 2000 and Vlastakis et al, 2008) and golf (Shmanske, 2005) and tennis (Forrest & Mchale, 2007). In section 2, the literature overview, many articles that support this favorite longshot bias are discussed.
The current literature (Forrest & Mchale, 2007 and Cain et al, 2003) suggest that there will be a favorite longshot bias in tennis matches. There is also a lot of contradicting literature that found a reverse favorite longshot bias, but this is mainly in other kind of sports. As will be conclude from the literature review in section 2, there is a lot of research done according to the favorite longshot bias in different sports, but not much according to tennis. In addition, does this study have a much larger database and more specific results than other studies (Forrest & Mchale, 2007 and Cain et al, 2003).
This study aims to answer the following research question:

Is there a favorite longshot bias in ATP and WTA tennis?’


To answer this question, data is collected from www.oddsportal.com. The total database contains 18.610 tennis matches, which automatically will result in 37.220 odds. This odds will be distributed in categories and the mean return per category will be calculated. After this calculation the results will be analyzed by a t-test.

The results do not show an existence of a favorite longshot bias. In addition, there is no strong evidence for a favorite longshot bias in WTA or ATP tournaments separately. Investigating the bias per years or surface does show a bias in ATP tournaments in 2011 and ATP tournaments on hard-court (outdoor), but the overall conclusion is an absent of a favorite longshot bias. This is not in line with the current literature of favorite longshot bias in the tennis sport (Forrest & Mchale, 2007 and Cain et al, 2003).


This thesis is structured as follows. In the next section the literature will be reviewed. To have a good overview, there are tables with summaries of the relevant literature. Thereafter the research question and the hypothesis development are discussed. In section 4 the data and the empirical research are explained. Then section 5 contains the results of this research. After analyzing the results, in section 6 and 7 these results will be conclude and discussed.



  1. Download 1.62 Mb.

    Share with your friends:
  1   2   3   4   5   6   7   8   9   ...   20




The database is protected by copyright ©ininet.org 2024
send message

    Main page