BACK TO PROJECTS

the motivation

Fantasy Premier League (FPL) is an online game that follows the real-life Premier League Football season. Given a virtual budget of £100m, each ‘manager’ has to assemble a team of players, who score FPL points based on their performance in real life games. Famous players who scored highly in previous seasons are more expensive. Therefore, good FPL managers are those able to spot cheap, unknown players with the potential to score big.

By collecting extensive amounts of data pertaining to both FPL manager decisions as well as the matches themselves (e.g. line-level detail for over 10,000 shots taken across over 400 games), CB91 was able to create a series of interactive dashboards to help FPL managers explore the performances of players and teams in the Premier League.

the product

Use the button in the bottom-right of each dashboard for a full screen view, and start exploring! Note - these dashboards work best on desktop/tablet browsers.

The GitHub repo for this project and summary of other findings can be found below the dashboards.

Watch the video below for an example use case...

1) PLAYER DASHBOARD

2) SEASON SUMMARY

3) TEAM DASHBOARD

the GitHub repo

View the readme on the project's GitHub repo for a full breakdown of the project.

This details the webscraping and data munging steps taken, as well as detail on further insights generated by network analysis and unsupervised learning, performed on the dataset.

the techniques

Analysis conducted using Python libraries and APIs:

• SQLite3 and Pandas for data mining and munging
• Splinter and Requests for API calls and webscraping
• Network X for graph theory analysis
• Tableau, Matplotlib, and Seaborn for data visualisation

the findings

This presentation was created to be presented to a non-technical audience in under 10 minutes

Note - data correct as at the end of March 2020.

the blogs

"Of course, the original marriage of data and sport is ‘Moneyball’, the Michael Lewis book that tells the story of the Oakland Athletics’ 2001 baseball season, for which their manager Billy Beane assembled a highly successful team of cheap ‘misfit’ players with appealing underlying statistics. And while statistics can help professional teams to claim a place in the annals of sporting history, enthusiastic amateurs are starting to benefit as well..."