STATS attended the South by Southwest (SXSW) SXsports® conference in March, where CTO Darryl Lewis presented alongside Joe Procopio, Chief Product Officer at Automated Insights, a STATS-owned technology company, on the development of automated sports analysis.
Below are some of the key points made during the session, as Darryl and Joe spoke about some of the major innovations in automated analysis and the likely future direction of the industry.
Watch the video…
Automated Content is on the Rise
Automated Insights’ Wordsmith platform is at the forefront of the automated content revolution. Using big data to automatically write human-sounding, professional narrative, the platform is able to publish in real-time and at scale. Able to produce over 2,000 per second, Wordsmith generated over one billion articles in 2015, each with their own client-specific lexicon and tone.
Automated content isn’t about replacing writers, it’s about helping media companies, brands and broadcasters handle the evolution of sports data and make it work for them in terms of expanding the scope of their coverage and meeting the needs of the connected consumer.
Real-Time Data is the Future
STATS captures data from 100,000 sporting events each year, with coverage taking in over 70 leagues across 211 countries. This amount of data capture sees over two petabytes of data transmitted through STATS’ systems each year, although currently the scale of the operation is costly and data cannot be processed in real-time due to the amount of manual data entry involved.
STATS is constantly striving to streamline data collection processes and seeking ways to deliver information in real-time to support in-game decision making for coaches and improve digital experiences for fans. To achieve this, STATS is investing in progressive technologies to engage teams and fans with instant, in-depth data feeds.
Sports Content For All
The proliferation of data at all levels of sport and the growth of automated content services means that sports coverage is no longer restricted to elite leagues and teams. Any team that keeps data from its games, whether they be professional, amateur, high school or little league, can be the focus of automated content such as match previews, recaps or statistical summaries.
Automated platforms are able to personalize content to certain players, giving people the opportunity to access their own data or read about how their kids played in a little league game. This sort of content is increasing the depth, personalization and power of sports data and can even act as a motivation for individual participation in sports.
There are vast amounts of performance data available to professional sports teams, but the challenge has always been finding the most effective way to apply it to the coaching process. In combination with advanced machine learning, the wealth of sports data possessed by companies like STATS will one day have the potential to give rise to real-time alert systems that could give coaches live recommendations based on technical and physical performance indicators.
The realization of these real-time decision support systems would be a major step forward in terms of the application of data, directly influencing the action as it happens. For example, coaches could access instant play recommendations or injury risk alerts that have a positive impact on their decision-making during games.