Game, set and match

Most of us know someone with an encyclopaedic knowledge of sports results. However, in the modern age of big data is this person becoming redundant? Historic sports results are now catalogued and openly available; and large data models are used to predict match results based on specific player attributes. For instance, tennis players are analysed on factors such as service consistency, ability to hold break points, and unforced errors on each surface type. That’s obviously helpful for bookies, but the same systems are now reportedly being used by the players to gain advantage. Analysis of past performance for instance allows this year’s Wimbledon contestants to know when to play to the backhand, when to hit a topspin return, and when to lob a serve-and-volley attempt, with a different strategy for each opponent. Meanwhile, not to be outdone, the All England Lawn Tennis Club has some enviable data of its own. In a typical year, spectators drink 303,277 glasses of Pimms, whereas players eat 2,195kg of bananas and 166,055 portions of strawberries and cream are sold, each costing £2.50. This price, which has not changed since 2010, compares with an increase in the prize fund over the same period of 148%, making them the best served treat this weekend.