Getty Images
Part three of this series concluded with the role of combining various indicators and insights and using them to inform real-time decision making. The next step is live match analysis which, despite being the latest trend in performance management, continues to be severely underestimated.
Coaches continue to believe there are more important things to observe during a match than these insights and they still trust their eyes more than they trust the data. Most coaches believe that they should not – or are unable to – divert their attention away from the pitch.
This viewpoint was valid until very recently because live data was not as important as it is now – but things are changing.
Seven years ago non-contact injuries accounted for 20 per cent of all injuries sustained by players. Three years ago that had risen to 36 per cent, according to a Fifa study.
Anecdotally, this season it seems like every other injury is a non-contact one. Additionally, the dynamic of play is 35 per cent greater than what it was a decade ago, with higher speed bursts, harder decelerations and more sudden changes of direction. Throw in the fact that teams play more matches than ever, and you have the perfect conditions for player overload.
Now, most people focus on the output they can see from the player, whether it’s goals scored, shots, passes, speed or distance covered. Few analysts examine metrics relating to live physical condition, such as heart exertion and recovery, or monitor adaptability to non-rhythmic effort by constantly shifting tempo and dynamic change of player load.
Basically, the thinking is that if a player can still deliver adequate performance metrics then they are fine. But this analysis fails to take into consideration what this performance costs the player in terms of energy depletion.
Understanding more about how a player is conserving or replenishing energy during a match has benefits. This analysis can be as simple as tracking a decrease in heartbeat during certain moments, such as when a player touches a ball, or corners, free-kicks or pauses for an injury to another player – essentially tracking the heart rate reaction against exerted effort.
The value of real-time insights
In part two of this series, we discussed the example of a player who maintained a high heart rate despite a decrease in dynamics and work volume. The reasons for such underperformance are myriad; lack of preparation, actuate overload, mental stress, the quality of the opponent, or even arguing with the referee.
I had a conditioning coach tell me recently that he knew why six of his players were injured – the problem was that he discovered this through post-match GPS data analysis after the event. The players had been acutely overloaded by more than 30 per cent above their benchmark capabilities. He acknowledged that had he been able to get insights in real-time, these non-contact injuries could have been avoided.
Insights differ from basic real time parameters in that they deliver a cross reference that takes into account player condition, performance, and functional and medical data. This includes medical screenings, functional tests, acute and chronic load, leisure activities, recovery and stimulation levels.
Effective in-game management means players spend more time on the pitch and less time in the treatment room
The avalanche of raw data is now so vast that even the most experienced coaches and analysts struggle to process it all – especially in real time.
This combination of offline and online data insights is what delivers these real time snapshots and is where the issue of artificial intelligence (AI) and machine learning (ML) algorithms with regards to player condition and performance comes up most.
However AI and ML are yet another set of buzzwords used everywhere without a clear understanding of how such technologies should be deployed.
The constant evolution of sport
At Barin Sports, we pool data from the four main streams mentioned earlier and feed it into an algorithm which seeks correlation between the various parameters to deliver insights such as readiness to perform and non-contact injury likelihood.
Previously collected data gives us a snapshot of the state for the player as they prepare to enter the match and the algorithm is provided with live data as the game progresses, using predictive analytics to inform coaches’ decision making and in-game management.
The best part is most of these techniques can, and should, be part of pre-match preparation. This means a large analytics team is not required and coaches do not have to divert significant attention away from the match to take advantage of these insights.
Automated alerts ensure even small coaching teams can benefit.
There are many well-established expectations about what coaches should be doing during matches but professional sport is constantly evolving. Adaptability and agility through live insights are an essential part of the future.
To read part one of this series, click here. Part two is available here and part three is here.
