Mladen Sormaz: Data analytics in the multi-club model
Written by Training Ground Guru — April 11, 2023
MLADEN SORMAZ is Director of Football Analytics at multi-club ownership group 777 Partners.
Speaking at TGG's Big Data Webinar last December, he explained how analytics works within the multi-club model:
DELIVERING INSIGHTS ACROSS CLUBS
Mladen Sormaz: We have five main clubs we work with directly as an analytics department - Genoa, Standard Liege, Red Star Paris, Vasco de Gama and Melbourne Victory (* since the talk another club has been added to 777's portfolio - Hertha Berlin).
There is a central analytics department of three of us who cover all the technical work and each has contact points at each club that were developed at the earliest point possible. You need to have a good relationship with - at a minimum - the Sporting Director and a technical member of the coaching staff.
It doesn’t have to be the same across all clubs, but having a similar structure to deliver new insights is vital to making a success of a central analytics department.
I have found it’s easier to start lean and have the flattest hierarchy possible and to have people who can be club facing and talk to people as well as doing the technical side. I have my Director of Research and Development, Dan Nichol, and Senior Data Scientist Elliott Stapley. They are both experienced with football data and good at communicating with people.
Any person from the clubs can approach any of us directly; it doesn’t have to go through me as the head. We need to focus on their needs, which means the maximum amount of availability.
Face time is key. All three of us attend games at all the clubs where possible. It is important to show your face, because you are invested in what is going on and need to feel that and build personal relationships. You are always more personable if you have had any time on the ground with someone, versus being remote.
And it allows good periods of time with club practitioners and staff, so conversations can happen spontaneously and ideas can be generated in a less formal setting.
Compared to a single club, there is a massive workload across our portfolio and we need to balance and satisfy all expectations across them.
If we’re not careful, we could be seen as the outsiders, especially as we work at the holding group level, and we need to be wary of things like differences in preferences and process in views of football and data suppliers.
You can watch Mladen's whole presentation if you purchase the Big Data Webinar on-demand. You get a total of eight sessions:
- Rodrigo Picchioni: Fully automating and integrating data at Atletico Mineiro.
- Mladen Sormaz: Data analytics in the multi-club model.
- Tyler Heaps: Data-driven decision-making at AS Monaco.
- Emily Angwin: Making an impact with data across the women's pathway.
- Fabio Nevado: How La Liga clubs synchronise tracking & event data with video.
- Akhil Shah: Data engineering - How to get from good to great.
- Jonny Whitmore: Quantifying player decision-making with predictive data.
- Lucy Rowland: Differences between club and international analytics.
Further things to consider are that there could be cultural and language barriers, so you have to make sure there is always a successful mode of communication.
So how do you solve the technical part? Again, that is the easier bit:
- Standardised data suppliers where possible, making sure everyone is getting the same information. Some leagues have specific deals with tracking data suppliers or longer-term contracts with GPS vest suppliers and you can’t change that.
- Centralised databasing, which is something you would do at one club anyway, but it makes even more sense at the group level, so all clubs can leverage the same data infrastructure.
- Consistent insight delivery timings and report structures. Build around how all your different clubs work, figure out what their workloads are and how their match weeks work. Try and give them the most consistent insight reports as possible, so everyone is getting the same quality of information.
- Insight report walk-throughs are even more important when you’re part of a multi club. You’re remote with some visits, so having time on Zoom or Teams to walk through every single new bit of work is even more important.
- And finally, have an open process to receive feedback, iterate and improve on your work. The benefit of having clubs across the world is they all have different things they bring to the table that you can work with.
Another key, as I mentioned earlier, was the balance of consistent KPI reporting with some flexibility. This is the 80/20 rule I talked about at Leicester. We have a set pre-match report that every club gets based on needs they have expressed to us.
What we do on top of that is any ad hoc work we think each club might re-use, especially if it’s specific either to their game model or situation. You need consistency across the whole group, but you also need the flexibility, because football is varied and people’s approaches and goals vary.
On the non-technical side, the key things, as I outlined at the start, are good structure and links.
So these are my key takeaways for making analytics a success in a multi-club:
- Always remember you are part of a social and cultural endeavour here. Be prepared to build for more diverse situations, different game models and different recruitment challenges. Some of our clubs are salary capped, others aren’t and that needs to be built for and addressed.
- Build efficient and comprehensive databasing. Technical synergy is the easiest win in a multi club because it’s the clearest benefit to everyone.
- Remember to build strong links not only with your central ownership, but also with each individual club. The point of you is to serve them. That will help the work to be used effectively.
- Provide that solid consistent base of work, with the consistent reports with the same formatting, but with the flexibility to be able to pivot to a diverse set of circumstances. Have enough bandwidth and people on hand to be able to do ad hoc or specialist bits of work that one club might need at any one time.
- And finally, remember to automate. The workload is massive, so make sure anything that will be used repeatedly is automated at the earliest possible moment.