Bookie Beaters

This visualisation explores the predictability of each team in the English Premier League. Bookmakers' pre-match odds are compared with the final results of each match over the course of a season, to give a diverging bar graph describing which teams have been the most, and least, predictable.

The most heavily predicted (blue) and most surprising (red) match results are highlighted in a more saturated colour, to enable identification of outliers and trends, and to find the most surprising results overall.

Bookie Beaters visualisation - Click to view interactive version

Data for this visualisation is from Football-Data.co.uk.

Common Connections - Football Pass Relationship Explorer

Swept away by World Cup fever in summer 2018, we started to create a visualisation that would help fans, coaches and analysts explore how football (soccer) players connect with each other, during the 90 minutes of a football match.

Passing Relationship Explorer - click to view interactive version

Our approach was to depict each player as a node in a radial node-link diagram, with the thickness of the link corresponding to the total number of passes between each pair of players.

This created a tidier visual style than simply displaying the players 'in formation', and then connecting them. Passes are made from a multitude of positions anyway, so representing the path as a static position does not provide any absolute spatial information.

We retained the context of which player plays in which position by including a pitch map, that allows selection and comparison of players between teams.

Animating the rotation of the radial node link diagrams and the bar charts, helps the viewer follow the transitions, and adds a bit of life to the vis tool.

That's it! This dataset was just one match from a Women's Super League dataset provided by Statsbomb. The match was Manchester City v Chelsea, on 24th February 2018.


Visualisation by Gary Rendle and Aziz Niyazov.