Olympic Games, marathons and a European soccer championship: this year’s summer has been a dream for data scientists interested in sports. Joost Kok, director of the Data Science research programme and board member of the Leiden Centre of Data Science, is a prominent researcher in this field. During his talk, he elaborated on several of the data science and sports projects at Leiden University.
Data analytics can be applied to sports in many ways, as Kok pointed out: discovering the ideal training plans for speed skating, detecting match fixing, or analyzing soccer games to find out what situations lead to changes in ball possession. The researchers in these projects always share a similar goal: discovering patterns in large amounts of data. These patterns help us uncover hidden knowledge, which can help athletes and sports teams to perform even better.
Of course, the same methods can be applied to more than just competition sports. For instance, in collaboration with the Faculty of Social Sciences, Kok and his colleagues analyzed how children interact on playgrounds. Another interesting research project is on active and healthy living for elderly people, using wearables for data collection.
Why have the possibilities of data science developed so quickly in the past years? Three main factors contribute to this, as Kok explains. Firstly, more and more data is becoming available. Secondly, new algorithms are being developed, which makes data analysis more effective. And last but not least, the computational infrastructure is constantly improving: better and stronger machines, faster connections.
Data Science research programme
The purpose of connecting data science and sports is twofold: it yields new insights on sports and at the same time it serves as a source of inspiration for new data science research. In a similar way, the recently launched Data Science research programme will be connecting data scientists with researchers from other faculties at Leiden University. This, says Kok, will be a great source of inspiration for data science research.