ON the exhibition floor of the Mobile World Congress in Barcelona in February, one of the stand-out displays was a large TV screen on which the tactics of the German football team were being analysed.
Enterprise systems company SAP was demonstrating how an application called Match Insights could gather data before and during a match and use it to influence the coach’s tactical decisions while the game was on.
Most saw the demo as a marketing exercise. But when Germany won the World Cup, systematically outplaying opponents with superior tactics, the data game became very real.
According to SAP, the journey started last year when national team general manager Oliver Bierhoff found that players were happiest communicating with each other on digital platforms. He commissioned SAP to develop an application that could facilitate the exchange of information, including data about opponents. SAP Match Insights was then developed in collaboration with the German team.
“This data can be converted to simulations and graphs that can be viewed on a tablet or smartphone, enabling trainers, coaches and players to identify and assess key situations in a match,” said Manoj Bhoola, a director at SAP Africa.
“SAP Match Insights synchronised data from scouts with the video footage from the pitch to make it easy for coaches to identify key moments in the game.”
The impact on the outcome of the World Cup is not as easy to quantify, but it’s given “big data” its biggest showcase yet.
“Big data is an incredible resource for coaches and players to contextualise information and draw well-informed conclusions to optimise training and tactics,” said Simon Carpenter, chief customer officer at SAP Africa. “It’s high time to make this type of information accessible to sports journalism and the fans as well.”
German soccer may have discovered big data, but it’s a path well worn by large enterprises.
“We have been doing it all along,” said Desan Naidoo, managing director for Southern Africa of global analytics company SAS. “But some of the aspects have changed. If you look at the volume and variety of structured and unstructured data, ranging from social networks to text and video, that has definitely changed. Ninety percent of all data ever created have been created in the past two years.
“This is unbelievable in itself. But now the requirement from clients to have access to this data has moved from running data through models for 18 to 24 hours to wanting access in minutes or seconds.”
And it’s not enough merely to analyse the data that are formally collected in organisational systems. “We’ve had to tap into social media data. We’ve had to restructure the way we do analytics to cope with the volumes. We’ve had to look at hardware changes and infrastructure such as in-memory analysis.”
The latter refers to loading relevant data into live memory so that it can be processed on the fly, providing usable information in seconds. A typical example is a customer going to a bank for a home loan: it can now run a risk profile and give an answer while the individual is waiting.
“In the past, if you based that risk profile on all the data sources the bank has, it would have taken hours,” said Naidoo.
“Having access in-memory means you can click a button and run a risk profile accessing all that data instantaneously. On top of that, analytics today can predict how that customer will behave, rather than being merely reactive, as in the past. That’s what big data means today.”
• Goldstuck is founder of World Wide Worx and editor-in-chief of Gadget.co.za. Follow him on Twitter @art2gee
• This article was first published in Sunday Times: Business Times