Finding the perfect swing

In the sports world, there are few things that are considered as difficult as a good golf swing. The player experiences the perfect swing maybe only a few times in their life. However, researchers from Borås and Skövde are at present scientifically devising a method for understanding how a golfer can zero in on the perfect swing.

Picture from testing of golf swings

“No one has previously looked at data like we are about to.“ ”We measure how a group of golfers technically works.“ ”We can then, for example, find similarities in groups with a comparable skill level,” says Rikard König, researcher at the University of Borås.

The basis of the research is the management of a large amount of data. 500 golfers make 10 strokes. The strokes are recorded both with cameras and a Doppler radar. The radar “looks at” a total of 26 different variables for each stroke, from the moment when the club hits the ball until the ball is out on the course. The originality of this research is however not the actual collection of all the data, but the way in which it is transformed into pure information, through what is called “data mining”.

“We have three intermediate goals or levels with the research,” says Rikard König.
“First we find new or improved algorithms for data analysis that describes a particular group of golfers' results; this is the basic research. We then want to find a method for combining video results with data from the Doppler radar. Finally, the goal is to find a method for the application.”

He is collaborating with two students working on their degree projects and a golf pro in order to develop a practical and applicable method. The research is being carried out jointly for the University of Borås and the University of Skövde. Region Västra Götaland is responsible for the funding, with the aim of creating forms of cooperation between the universities. The idea is for the universities to jointly develop research expertise.

“The golf swing research is both scientifically relevant and applicable.”  The development of algorithms for data analysis is very current. Similar research on algorithms to describe the lowest common denominator in large data collections (known as Big Data Analytics) is being conducted in many fields, such as the automobile and pharmaceutical industries.

Despite the work of the research group from Borås and Skövde having only been underway since January, there is already hope for good results.

“In a simple preliminary study, we have seen that it is possible to describe a good swing. Now it's time to gather data from our 500 golfers and begin processing the material,” says Rikard König.

Text: Gunnar Fägersten Novik
Photo: Ulf Nilsson