STATS is a sports data, technology, statistics and content company which collects live data on sports teams and generates insights through machine learning algorithms. It creates data through its innovative SportVU system, utilizing cameras to track players’ and balls’ speed, distance and positioning in sports such as football (or soccer), NFL, NBA, WNBA and MLB. For example in the NBA, six cameras are positioned above each court and photos are taken at 25 frames per second to provide live 3D positioning data (STATS, 2018).
STATS sells its information along with detailed analysis to teams to improve the teams’ performances. To stay ahead of the curve in sports data analytics, STATS needs to develop its machine learning capabilities to constantly look at the data captured to identify market leading insights for teams to utilize.
In the near term, the focus for sport analytics companies has been to build algorithms to predict likely behavior based on previous data. STATS has been developing cutting-edge technology, ‘ghosting’, which allows teams to predict scenarios based on where team members ‘could have been’. In effect, ‘ghosting’ allows teams to draw insights from games without needing to review hundreds of hours of game footage (Woodie, 2017). This has many practical applications such as identifying what is likely to disorganize opposition formations and how teams should have moved to minimize the opposition scoring (Le et al., 2017). Conversely, it allows teams to identify formations which seem to work better than others with the players that they have and they can adapt accordingly.
In the medium term, STATS has recently bolstered its Sales and Customer Success teams (STATS, 2018) and is looking to further options beyond its current sporting markets. As a leading provider of sports data analytics, STATS intends to build its customer base as technology improves in step. Given its large investments in technology, it is critical for STATS to identify potential customers and describe the current and future usefulness of its insights. An increased focus in this space allows STATS to further invest in R&D for potential new technologies.
As databases continue to grow, STATS has opportunities to utilize time-scale of its data to inform indicators of players’ future success. By having access to NCAA Basketball as well as the NBA, STATS has data available to assist NBA GMs identify players who may have high-potential to become good players at their team (Douglass, 2013). Successful NBA players would be identified and their NCAA Basketball careers could be mapped to current NCAA Basketball players to determine a likelihood of success. Teams spend a significant amount of money on scouting recruits and machine learning could significantly improve the efficiency and success rate of the recruitment team. This savings could be monetized by STATS through its potential offering.
Furthermore, STATS should try to build technology which predicts the ability for players on the same team to play with one another. So much information is captured when players play together and tendencies can be identified to determine what players or play styles best match each other. This application can be extended to the way teams make trades on players, unlocking many more win-win opportunities where player potential is unlocked due to the chemistry of players involved. Again there is a large potential to monetize this due to the large amount of time and money spent on brokering trades.
Another opportunity for STATS is to utilize their engines to identify the best training programs for players in between games to increase performance and decrease injuries (Woodie, 2015). By including analytics on training programs as well, STATS can provide an even more holistic perspective on what drives athletes to be able to perform at a high-level. Teams can set even more detailed training schedules and players can improve their play at a quicker rate. Generally improving the team’s rate of success is likely to increase other revenue streams like ticket sales and merchandise which can be very lucrative to a sports organization.
It is clear that there are many options for predictive analytics companies like STATS once they have unlocked so much data capturing potential with their innovative technology. STATS face an interesting problem of determining what set of insights they should target to unlock the most amount of revenue. What capabilities should STATS aim to develop? What other sporting markets should STATS aim to tap?
Word count (729 words)
Douglass, B. (2013). How SportVU Came to be the NBA Analytics Game Changer. [online] SportTechie. Available at: https://www.sporttechie.com/sportvu-pulls-nba-analytics-forward/ [Accessed 13 Nov. 2018].
Le, H., Carr, P., Yue, Y. and Lucey, P. (2017). Data-Driven Ghosting using Deep Imitation Learning. MIT Sloan Sports Analytics Conference. [online] Available at: https://s3-us-west-1.amazonaws.com/disneyresearch/wp-content/uploads/20170228130457/Data-Driven-Ghosting-using-Deep-Imitation-Learning-Paper1.pdf [Accessed 13 Nov. 2018].
STATS. (2018). Basketball Player Tracking for Pro Teams | SportVU | STATS. [online] Available at: https://www.stats.com/sportvu-basketball/ [Accessed 13 Nov. 2018].
STATS. (2018). STATS Strengthens Sales and Customer Success Teams by Hiring Three New Directors – STATS. [online] Available at: https://www.stats.com/press-releases/stats-strengthens-sales-and-customer-success-teams-by-hiring-three-new-directors/ [Accessed 13 Nov. 2018].
Woodie, A. (2015). Big Data’s Next Big Thing: Sports Training and Personalized Medicine. [online] Datanami. Available at: https://www.datanami.com/2015/07/29/big-datas-next-big-thing-sports-training-and-personalized-medicine/ [Accessed 13 Nov. 2018].
Woodie, A. (2017). Deep Learning Is About to Revolutionize Sports Analytics. Here’s How. [online] Datanami. Available at: https://www.datanami.com/2017/05/26/deep-learning-revolutionize-sports-analytics-heres/ [Accessed 13 Nov. 2018].