The news that English Premier League soccer team Liverpool is utilizing deep mind artificial intelligence to explore in-game tactics is a watershed moment for sport. AI can now impact many other sports to deepen our understanding of team performances, beyond what is possible through ordinary human cognition.
Published in the Journal of Artificial Intelligence Research, the initiative opened up possibilities for using GPS data, computer algorithms and sensor data, among other things, to see patterns to aid the coaching process.
The study suggests that models on data can be used to predict how teams and players tend to react to certain situations. For example, in soccer, if a long pass is played down the right wing, certain players tend to run in a particular direction, compared to others. This level of minute detail may be used by coaches to calculate new gameplans and adjust training regimens to take advantage of this new information.
Sporting leagues around the world will be taking notice. China has invested in domestic leagues in recent years, and a homegrown soccer culture is emerging. Clubs such as Guangzhou Evergrande and Shanghai SIPG command legions of avid fans who may also be a source of interest for this new technology. An investment arms race among soccer clubs has been on-going for many decades already, and uptake of AI may be the latest battleground behind the scenes of the world's best soccer teams.
This phenomenon is not just apparent in soccer. In the United States, baseball is also undergoing an AI revolution, where countless pitches and catches are analyzed and broken down to ensure that the next season is a successful one.
However, technophobes can rest assured in the knowledge that AI is unlikely to be replacing managers anytime soon. The nuances of the game are things that only a lifetime of passion and social intuition can achieve. However, this will mean that in the coming decades, such AI systems may well be used in tandem with managerial talents to create a seamless system. Automated AI feedback for key game moments may be relayed back to players based on the analysis of thousands of similar historical scenarios. Locker-room conversations may never be the same again in the presence of deep mind machine learning.
It is well known that athletic performance relies heavily upon successful strength and conditioning expertise, too.
Weight training, for example, is universal across all professional sports in the modern day, with even darts and F1 drivers having vigorous exercise programs carefully calculated to maximize performance.
AI can make use of data collected through conditioning sessions and provide feedback in real time to maximize results. The advent of wearable technologies in recent years may further combine with this trend to create bespoke training programs and save time for coaches. A careful eye is needed especially when it comes to detecting levels of activity nearing chronic injury and overexertion, and the more eyes the better, whether they be human or not.
The AI revolution also reaches into how we enjoy and consume these sports. Using clever programming, AI can ensure that cameras pick out highlights of a game and automatically put them into an edit to distribute for television and mobile online platforms.
Natural learning platforms which are able to process and recognize raw data of interest to an audience, and then transform them into meaningful forms of media, are vastly improving the capacity of the media to cover sport. This means that more high-quality games are reaching passionate fans. In the coming decades, AI will not only improve the calibre of the athletic feats seen, but also the quantity of them too.