by Rebecca Merrett

How sensor data analysis keeps cricketers in top shape

Mar 20, 20153 mins
Data Mining

Professional athletes are turning to sensor data analysis to bridge gaps in their performance and improve their game. This is what biomechanics expert Dr Edouard Rene Ferdinands from the University of Sydney is working on with cricketers.

“If you can enhance the sensory data, the feedback loop… then you are optimising the learning process. If you optimise the learning process, you have a way of optimising performance,” he said at an SAP event today held at the Melbourne Cricket Ground.

One of the biggest focus areas for Ferdinands is improving the performance of fast bowlers, as they play a crucial role in the game of cricket. The aim is to increase speed by using motion sensors to study a player’s movement in detail and see where coaching intervention can be made.

Instead of waiting for a high performing bowlers like Shane Warne and others to pop up, taking a data and scientific approach can help create that talent, he said.

“Do we have to wait another 20 years for the next Shane Warne? If you can find a way of actually training this, to achieve this in a systematic way, then you have a much better chance of holding the top position in world sport for a longer period of time without those drops in performances.

“There’s one Shane Warne, but imagine if you had two [in one team]. That’s the aim, to start to produce more than one Shane Warne,” he said.

The process to doing this is to first feed all the sensor data into one central repository where it can be processed, sliced and diced, he said.

Then top performing, history making bowlers are examined to form a model that others are assessed against to identify the gaps. Coaching intervention programs are then formed to fine tune a bowler’s movements in great level of detail.

Results from the model also need to be evaluated to see whether it was effective in helping improve performance or not and whether it needs to be tweaked.

As humans come in all shapes and sizes and genetic makeups, Ferdinands said taking into account individual characteristics when building a model is an important part of ensuring its success.

The physical condition of a cricketer’s body also affects performance. Ferdinands uses 3D motion maps of the kinematics of the entire body when doing analysis to see where are the areas that endure most pressure.

For example, if parts of the body and the spine don’t carefully align when a fast bowler puts much force and energy into bowling, there’s great risk of injury.

“So it’s not about the coach saying, ‘I’ve done the job, look he is bowling better.’ But has the change [in technique] led to a better aligned lumbar spine?”

Ferdinands also uses 3D motion maps to look at the legality of bowling actions, rules around specific movements to ensure cricketers aren’t cheating in the game. The analysis of this helps coaches find ways to address this more effectively in their training programs.

Besides motion sensors, Ferdinands also uses sensors that detects a person’s anxiety levels.

“An athlete that is anxious does not move as well,” he said.

Playing cricket or any sport professionally can be a high pressure, stressful job, Ferdinands pointed out. So data analysis around anxiety could help identify players who are prone to this and build programs around dealing with intense moments in a game.