In this special guest feature, Raf Keustermans, CEO of Sportlight Technology, discusses how professional sports teams around the world measure everything — individual player performance, team performance, how much an athlete is training, player speed and endurance during a game, injuries sustained, goals scored, assists made, passes blocked and much more. Strong data collection is important, but what can be even more crucial is consolidating all of that information to make it usable. Sportlight is a fast-growing sports tech startup building the next-generation athlete tracking technology platform using LiDAR and AI. Raf is an experienced entrepreneur, he was previously co-founder and CEO of mobile games startup Plumbee which was acquired by Sony’s GSN Games. Prior to that Raf worked as an executive for companies in gaming, betting and advertising, including Electronic Arts, Kindred Group and BBDO.
In the increasingly sprawling universe of sports information, nearly everything that currently can be measured is being measured. Franchises and organizations invest small fortunes collecting data in the pursuit of even the slightest edge on the playing field. They employ teams of professional wonks and deep thinkers to expand their reach of quantifiable data, storing those results in small cityscapes of servers.
With so much to observe, and so many ways to interpret the data, different types of statisticians are turned loose on the numbers, armed with a variety of models and algorithms that spit out all sorts of projections. All that numbers-gathering churns out data that can, and very much does, influence team decision-making. But It can be difficult to harness. Collecting and storing data is a vastly different enterprise than making sense of it.
So how do sports teams and clubs consolidate all this wonderful data? The ability to point a device and take a measure of almost countless aspects of performance doesn’t necessarily come with the knowledge of which, exactly, are the most valuable aspects worth measuring. And even if you’ve figured out that much, you may not know quite what to do with the information once you have it. How are sports organizations to make practical use of all this data?
In the Premier League, clubs use modern technologies to precisely record measurables from speed to change-of-direction quickness. That phenomenon isn’t especially new. But what they have added more recently are modern data sensibilities, analyzing those results and meshing them with other performance measurables to arrive at statistics such as strike rate and expected goals – metrics that reveal more about the effectiveness of process and potential, removing some of the randomness that accompanies outcomes.
Just the same Major League Baseball clubs house troves of advanced analytics, the latter of which are proprietary to each team. After Oakland Athletics general manager Billy Beane ushered sabermetrics into baseball, herds of Ivy League statisticians descended upon baseball’s front offices all with different algorithms that net predictive analytics. And they each have individualized methods for storing and interpreting that data. Just the same, basketball tapped into analytics, which has unveiled critical insights about player usage — and over-usage.
What these components amount to is a language – a way to understand whether, for instance, a side is optimizing its assets, whether a club manager is making the right moves and whether a player may be overachieving or perhaps has yet to access untapped value. Applied data adds context to what may appear to be chaos.
Just the same, data can create chaos if not used appropriately. Too many individual analytics could ultimately be contradictory.
That’s why it’s important to constantly re-evaluate the data you’re consolidating. Yes, the end goal, of course, is to house your data in a succinct way that helps player personnel departments make decisions. But consolidating lesser-accurate data points can derail any player personnel department.
Where teams across sports — from the Premier League to the NBA — fail in their data collection is in the way in which they measure these analytics. It’s critically important to find data collection technology that accurately assesses the performance metric you are analyzing.
Creating that language begins with identifying the most important data, or most valuable data sets and interactions, then consolidating that information into a program or app – a model that helps an organization make decisions. The idea: store and sort this data (even if you aren’t yet able to put it to use) in a repository that allows for its academic organization.
Consolidated data, even at its best, is not a panacea for sports franchises. Metrics are another tool in the kit, but they are open to interpretation and ultimately, more prescriptive than predictive. Still, sports organizations play the percentages every day. Data that arguably makes its decision-makers just a bit smarter – or better informed – than the competition may be the small advantage that turns a loss into a win or transforms a contender into a champion.
Sign up for the free insideAI News newsletter.
Join us on Twitter: @InsideBigData1 – https://twitter.com/InsideBigData1
