How the science of analytics can raise football’s game

This post was contributed by Nick Eisen, Business Engagement Reporter, Birkbeck School of Business, Economics and Informatics

Football

Studying the data could revolutionise football’s development, according to a conference hosted by Birkbeck’s Sport Business Centre in February.

In collaboration with the Sport Business Centre, sports analytics provider OptaPro’s 2016 Analytics Forum attracted international attendance when the event was held on Wednesday 10 February at the University of London’s Senate House.

The Birkbeck Sport Business Centre had also hosted the previous two editions of the annual conference. The collaboration with Opta is part of the Centre’s mission to provide a forum for informed debate on emerging areas of best practice in sport management, of which the field of player performance analytics is one.

Winning formula: finding present indicators of future performance

Presentations showed how analytics could help to build winning teams and described research quantifying correlations between specific elements of play and performance. Recognising such correlations could help coaches, managers and others focus on elements indicating correlations with longer-term success. Some initial findings also cast doubt on pieces of received wisdom.

Attendees networking at the event

Attendees networking at the Analytics Forum in Senate House

Focusing on the Premier League, analyst Joel Salamon found like-for-like comparisons to be valuable indicators of future performance. For example, he found correlations between chances a player created early in a career and chances created later on, as with shots: like predicted like. Conversely, Salamon’s study found no significant correlation between how young a player was when first recognised as a top-level regular and subsequent success.

One of blogger Sam Jackson’s findings from his study of goalkeepers, and how they deal with crosses, suggested that the game attributes too much importance to how tall a goalkeeper is.

Speakers acknowledged that their findings were preliminary, required larger studies involving more players and teams, and could be refined to offer more sophisticated indications of future success or failure. But, the speakers argued, the studies already suggested analytics was a valuable addition to other skills in the game.

Translating research into practice, numbers into words

Dean Oliver presents at the OptaPro 2016 Analytics Forum

Dean Oliver presents at the OptaPro 2016 Analytics Forum

One question from the floor was: how could the mathematics of this research translate into a language that would win over players, coaches and others? Guest speaker Dean Oliver, renowned statistician of the US National Basketball Association (NBA), addressed this.

Citing his NBA work, Mr Oliver suggested that analytics experts adapt to the sport rather than try to teach the sport about analytics, and that analytics were an addition to existing approaches, not a replacement for them. Introducing analytics was likely to be a gradual process that would have to overcome resistance with persuasion, fostering cooperation and managing expectations by avoiding overconfident claims. Two of Mr Oliver’s key phrases were: “Relate everything to wins”, and: “Understand the existing process and where you can help”.

This Forum offered an exciting view of the contribution analytics could make to football.

With thanks to OptaPro (view tweets from conference)

Hosted by Birkbeck Sport Business Centre (on twitter and LinkedIn)

Find out more

The next School of Business, Economics and Informatics Open Evening, will be held on 14 April, 17:00-19:00

Share

Leave a Reply

Your email address will not be published. Required fields are marked *