We paid* @Banouby to follow the nerds to the Sports Analytics Conference (*we didn't pay him)
I’ve been knocking around the sports data industry for a while now.
So long in fact that my first ever meeting was around trying to think of different phrases to go into the archaic automated text football commentary. It was the only time my ability to think of 6 different ways to describe the taking of a corner kick was useful to me, and I felt somewhat vindicated for wasting my formative years staring at Teletext and watching Gazzetta Football Italia.
It’s with that background, and subsequent years spent working for companies such as Opta, Sportradar and now Sportlogiq that I watched on with a mixture of bewilderment, awe and fascination at the recent Sports Analytics Conference held by the Seattle Sounders.
Here are a few things I think are bubbling away in the world of sports data:
1. We’re ready for the next evolution of football data
Given that I work for a company that specialises in computer vision data collection, this is straight from the you-would-say-that department but it’s increasingly clear that the event data that was once cutting edge is now just the starting point, especially for the bigger clubs.
What was once the next big thing is now simply what gets clubs to the start line.
The next edge is adding a new layer on top of that event data - new metrics collected by machines that benefit from added context around every event. For every pass, you now need to know what didn’t happen as well as what did. What options were there for the player, and did they choose the right one?
From this, we’ll very soon we’ll be able to quantify and analyse things like decision making, bravery and risk taking in real time. Exciting stuff.
2. There are edges, and there are edges.
As I mentioned, access to the event data is the starting point, and this is now a commodity with most top level clubs. There is still a big variation in how even that data is used and analysed though. Some of the work shown in Seattle by the likes of Arsenal, Liverpool, Toronto FC and the Seattle Sounders themselves showcase just what is possible with this standard data set.
It’s the clubs that have rinsed the possibilities of the event data that will gain the most from the contextual tracking data that is coming down the track.
Those who will benefit most from the next wave of available data will be the ones who have invested in getting the most out of the last wave.
3. Talent is the battleground
I’ve long thought that teams spend so much time, money and effort on scouting on pitch talent, but nowhere near as much scouting the off pitch support teams.
If you’re talking about extracting the maximum possible value in a club environment, there is certainly something to be said for investing heavily in the processes that enable you to attract, develop and retain the best backroom staff, instead of spaffing an extra 20 million quid on a reserve right back.
In analytics this is certainly true. The best talent will help you to sweat the data assets you have in the club already, across performance, physical and well-being data - potentially saving or making the club millions somewhere down the line.
4. Sometimes obvious = good
The Seattle conference had a lot of deep analytical stuff that I pretended to understand. But it also threw up the occasional reminder that complexity for complexity’s sake does no-one any favours. For example, the analytics talk from Liverpool FC showed a small element of the work they’d done on identifying Mo Salah before bringing him in. And this amounted to not much more than ‘he’s scored loads of goals and got loads of assists for years, in decent leagues’.
So while analytics and data led scouting is a crucial part of a modern club’s tool-kit, don’t overlook the fact that if someone looks really good, they probably are really good.
It’s interesting to me that these type of conferences are lot less about future gazing than ones in say, marketing. It’s very much around the here and now, or the very near future. I think a lot of this stems from the fact that the best work is still being done in secret, and there isn’t much of a culture of information sharing between clubs, for obvious reasons.
So it’s difficult to go to an event like this one, as informative and enjoyable as it is, and get a clear road map of the next five to ten years. How is performance data going to interface with health data, sleep data, or data from wearable tech? What are the edges to be gained from understanding how to connect all the different data sources within a club environment?
Some people probably know. But if they do, they’re not saying it in front of an audience of competitors.