Tennis Myth Buster

Chunsang Ngai

Alejandro Casamajor Torrens

Data Analytics Bootcamp

INTRODUCTION

Association of Tennis

Professionals (ATP)

Focus on Grand Slam

DATA

Source         Kaggle

Period           2008-2018

4167 rows and 41columns after data cleaning

WELL-KNOWN FACTS

Higher ranking        how more likely to win?

Playing at serve         easier

Grass court          more aces

Taller player          more aces

1

2

3

4

Text

How likely are they to win?

1

Rank difference > 10

Easier when serving?

2

Differences Between Surfaces

Aces & DF per match

3

Roland Garros

Wimbledon

US Open & Australian Open

Is height a differential factor?

correlation coefficient = 0.58

4

MAIN ANALYSIS

3 Hypothesis

Harder to play against a Left-Handed

The longer the match, the more likely players are to make a DF and less likely to make an ace

Top players play better when facing a break point (under pressure)

1

2

3

80.1

72.6

Harder vs Left-Handed?

Top 10 players

YES

1

87% Right-handed

13% Left-handed

267 total players

Long match       more Double Faults?

NO

2

correlation coefficient = -0.03

Do top players perform better when facing a Break Point?

Most of them NO

3

Top 10 players

d = 2.49%

d

Top 41 to 50 players

d = 3.33%

LEARNINGS & IMPROVEMENTS

Github

Pandas

Statistics libraries (SciPy, plotly...)

Machine Learning

In progress

Thank you!