Predictive analytics can help beat the opponent on the other side of the net

In sport, effective data analysis is reliant on context and a level of relevant information.

Putting data in context is essential to understanding its impact and importance. Greater context can provide more effective measurement.

Technology is already driving change in sport, providing an accurate way of understanding performances, both in training and competition.

The most common, hawkeye, has already made a significant difference to the way teams and individuals in sports such as tennis and football train and operate.

The most technologically advanced teams measure everything from physiological tests such as lactate, heart rate, breath by breath, velocity and power data to more biomechanics tests such as step, length, velocity, force and power.

This is all in addition to one of the most common data analysis tools – video.

Contextual information offers a much richer understanding of activities. The challenge comes when organisations look to combine information to understand its context.

>See also: AV technology can benefit the office environment during sporting events

To really understand who is a better athlete, power would be a much better determinant than speed. While environmental data is a variable, power is relatively constant.

To use a recent example, if we look at Rafa Nadal, who recently missed out on winning the Australian Open Final against Roger Federer, there are some interesting stats that illustrate the importance of context.

In his semi final against Dimitrov, Nadal’s first serve percentage was 73%, a whole 7% better than this year’s average.

If you were Federer preparing to face Nadal in the final, you would probably be made aware of this and it on its own might cause you to adjust your preparation or game plan.

However, this would in fact be a mistake. Rafa Nadal did not win nearly as many points on his first serve as he has done throughout the tournament or career.

The point is that all of this data creates a full picture of Nadal’s game. Understanding how different pieces of data fit together is therefore essential to putting information in context.

Once this principle has been grasped, the next challenge is actually combining data to produce relevant and useful analytics.

Developing predictive analytics is often a core aim of data scientists. In sport, predicting how an opponent might set up or react to condition changes will help data scientists advise teams or athletes on how best to play or race in order to win.

>See also: Why the future of sports is in the cloud

For sports teams, developing predictive models is actually relatively simple, the challenge comes when they want to feed in real-time data. Particularly in sport, data driven insights need to arrive soon after the event.

Understanding how an athlete or team performed a day after activity is less useful as it leaves little opportunity to trial an immediate change or improvement. Late reports become interesting history not active contributors to real time decision making.

The challenge comes from the manufacturers of monitoring equipment. Each produces information in its own proprietary formats, and often at different time frequencies.

As a result, combining, let alone understanding timelines is incredibly difficult, especially in a tool such as Excel.

British Athletics has combatted this challenge by working with us to build their own bespoke tool, enabling them to generate in-session analysis using combined data and multiple video.

As a result, to truly understand data in context, teams should be investing in their data management and analytics tools to enable easier, faster and more effective data extraction from equipment.

>See also: Place your bets: do odds improve as sports enter the predictive age?

Systems should then be able to report out visualisations, taking into account environmental data to provide true context.

Combining, collating and understanding data is a real challenge, that is not just unique to sport. In all industry sectors though, the principles are essentially the same. Context and timing are key. For organisations struggling to understand or manage data, engineering could provide a solution.

The profession tends to take a relatively strict and straightforward approach to innovation, following a research, do, review and repeat process until optimum outcome is achieved.

Ultimately data is a science, so following a scientific and engineered approach will ultimately pay the greatest dividends.

 
Sourced by Samir Abid, CEO of Pace Insights

Avatar photo

Nick Ismail

Nick Ismail is a former editor for Information Age (from 2018 to 2022) before moving on to become Global Head of Brand Journalism at HCLTech. He has a particular interest in smart technologies, AI and...

Related Topics

Data Analytics