Project: Is there evidence that referees are biased?

Project

1 of 2

CBM LEVEL: Beginner / Intermediate / Proficient / Advanced

IS THERE EVIDENCE THAT REFEREES ARE BIASED?

In this project, you will use data from the Premier League (2020–2021) to find out if there is any evidence that the referees were biased in favour of the home teams.

This guide helps you through the problem-solving process with helpful hints and questions to direct you.

There are 20 teams in the English Premier League that play each other at home and away grounds throughout the season, making a total of 380 matches. There is a popular belief amongst fans and players that playing at your home ground is an advantage. Is this true?

	Football referees have to make quick decisions.Could they be influenced by the home crowd?

The Wolfram Language code will only be enabled once you Make Your Own Copy of this Wolfram notebook.

You will need to sign in to the

Wolfram Cloud»

with a free Wolfram ID to do this.

➜

Evaluate this to download data from the 2020–2021 Premier League (source»).

data=Import["http://www.football-data.co.uk/mmz4281/2021/E0.csv"];

(The codes used in the first entry of the data are explained here».)

A version of this data with a reduced number of fields is available in a variable called premierLeagueData2021.

premierLeagueData2021

➜

Follow the

Computational Thinking Process

to determine if there is any evidence that football referees in the Premier League make decisions that are biased towards home or away teams.

Tips:
• Extract the relevant parts of the data (i.e. parts of the game that are ultimately decided by a referee).
• Perform a

hypothesis test

You may find the following code useful:

DistributionFitTest

Dataset1

Dataset2

,"PValueTable"

LocationEquivalenceTest

Dataset1

Dataset2

,"PValueTable"

VarianceEquivalenceTest

Dataset1

Dataset2

,"PValueTable"

Histogram

Dataset1

Dataset2



DEFINE QUESTIONS

What is your precise question to tackle?
Can you write a formal hypothesis test?
What is a suitable level of significance for such a test?
What assumptions are you making?

ABSTRACT TO COMPUTABLE FORM

What sort of data are you using?
How reliable is the source?
Are there any outliers in the data?
Which tools will you use and why?

COMPUTE ANSWERS

What statistics have you found?
How did you compute them?
How did you overcome any problems?

INTERPRET RESULTS

Should you accept or reject your null hypothesis?
What do your statistics mean?
Do your results seem correct?
How can you check that it is a sensible answer?
How much does the answer depend upon the assumptions you made?
What could you do to improve your answer?
Are there further questions you could ask to develop your analysis?

End of project.

Primers

END OF PROJECT

Resources to Explore

	Wolfram\|Alpha
	See cool things Wolfram\|Alpha knows about

	Wolfram Demonstrations Project
	See 12000+ interactives from users

	Wolfram\|Alpha Notebook Edition
	Explore further with linked free-form input

	CBM LinkedIn group
	Ask questions and discuss ideas