6 min read
This edition of Analysts Assemble takes us into the exciting world of professional sports analytics.
Analytics consultant Samir Abid tells us about his journey from engineering to analytics and how his experience of these worlds has put him in pole position to help professional sports teams with their insights.
We also hear how Samir is using the knowledge he has built up over the years to help others work better with their own data.
Tell us a bit about yourself, how did you get into the data space and what does your data journey look like so far?
My name is Samir Abid and I run a sports analytics consultancy called Pace Insights. We work primarily with British based Olympic and Professional sports teams to help them make their data useful (harder than it sounds!)
My background is actually motorsports and automotive engineering. Specifically “vehicle dynamics” which involves the design and optimisation of suspension and steering systems.
Computer Aided Design (CAD) and Computer Aided Engineering (CAE) systems were fairly new when I first started in the industry. The established guys were not so interested but I was. I’d built and upgraded a couple of PC’s whilst at Uni (as you did that then!) and always found them easy to use. I was always helping my Mum how to use her computer and also others at Uni with theirs.
Maybe I was also a bit lazy in that if I set up the computer right, it could do the work for me – less hassle and more accuracy. It didn’t bother other people so much but I always got frustrated having to repeat the same thing twice. Still does actually!
A car suspension system is surprisingly complicated. It has non-linear inter-dependencies and means you can quickly get lost trying to design and develop them. Some clever software guys had built these “multi-body-dynamics” tools which used maths to enable us to simulate the whole system.
These were physical models using lots of simultaneous equations. They were pretty crude and slow to use but I got into it and enjoyed running the models at work. To speed things up I then learnt about using stats models to create response surfaces using Monte Carlo and other similar types of analysis. We’d create a model off a model (eek!) but they proved really useful – I was even able to publish them into Excel with some slider bars so the other Engineers in the team could use them.
That was my first experience of developing “data products” for others to use and the lessons (nee mistakes!) I made then stick with me through to today. I just kept at it until they had something useful. It was really rewarding.
I was actually lucky enough to work on the new Aston Martin DB9. We did this from a clean sheet of paper (quite rare these days). Despite being a high-end brand, they didn’t actually have a lot of cash. In fact, where we’d normally have 3 prototype phases, they could only afford 2.
This meant the team were relying even more than ever on the simulation results from my models to specify and design everything from the suspension arms, the brakes, the steering rack, even where to position the engine. When it came out, Top Gear loved the car (and the handling!) so much they created a whole new section on their “cool wall” to put the car in! Good times.
A good friend of mine from Formula 1 had then moved to UK Sport. He asked me to help with a few projects in about 2010. It became clear that sports were getting more technical and starting to use data more on their programmes but didn’t really have the skills or experience. I’ve been helping them get more value from their data, reports and processes ever since.
What’s a typical day look like for you in your current data role? Which tools and languages do you use? Big team/small team/lone wolf? Remote/office based/co-working space?
As a consultancy the nature of the work varies wildly. We’re based out of an office and workshop in Leamington Spa. This is a great central location for most of our customer base. Over the years the team has grown and shrunk, peaking at 8. These days I tend to front the work and call on a select group of proven freelancers when needed.
It is probably not that cool but the majority of our work has been developing “apps” in Excel over the years. This was purposefully done as all our customers have it, it makes a great prototyping environment, can be programmed (with VBA), requires little/no “environmental” maintenance and can easily handle most of the data types and data volumes our customers have.
I have a thing for making reports and charts look good. This is possible in Excel (one customer actually didn’t realise they were using Excel on one training session!) but clearly Excel does have significant limitations.
I’ve even been able to try some “live” dashboard development on-the-fly in front of a room full of grumpy coaches! That clearly breaks every 101 rule about how to avoid looking like you’ve no idea what you’re doing but in fact the risk came off much better than I expected. The rabbit was pulled from the hat. Cue much debate about the consequences of what they were seeing rather than trying to comprehend what they were seeing (success!).
I highly recommend trying it sometime.
You’ve built up a good following through your own blog and newsletter. How important do you think it is for data professionals, at all stages of their career, to share publicly what they are doing and learning?
The blogging is relatively new thing. Writing, unlike maths and models, doesn’t come easy. I’ve set up the side project Your Data Driven where I am hoping to share some of my experiences over the years in the hope of helping others.
One thing I am finding is that it really makes you think about assumptions and jargon. I really want the content to be valuable and “actionable” rather than hand-wavey-bs and I think that is why I find it so difficult!
For anyone to do that well will certainly help their career because, whilst the headline is data, it is really all about people. Being able to communicate your ideas clearly is so important and often not done that well. Hopefully I will get better by doing more of it.
Where do you see your own data career going next? Building on your technical skills or moving into a more management-based role?
I really enjoy the consultancy and I work with a fascinating industry. Adopting value-based pricing has really helped square away some of my own issues about doing consulting (i.e. the obsession with buying / selling time, which has no correlation to outcome success and penalises people like me who like to work really fast). The roller coaster of feast or famine does get a bit much after a while.
Of all the things I’ve done over the years, helping others to do their work better / easier has given me the most pleasure. In the future I’d like to work out more ways I can help them away from implementation of bespoke apps.
Recently I’ve done more strategic consulting for a couple of customers and thoroughly enjoyed that. I’d also like to see where I can take the blog – if people are interested in what I’ve got to share then that could lead to creating some small products which would be a good challenge.
I’ve been successfully building them on a bespoke basis so I’d like to see if I can make something to successfully solve an issue everyone is facing.
If you had a list of “best-kept-secrets” (websites, books, coaches) that have helped you, which would you recommend?
I like the work of Edward Tufte. There are a few good “Data science in Excel” books which I think are worth a read – they build ML models from scratch using sheet formulas alone!
Other than that, the best people who ever helped me are the people I’ve worked with – in terms of their feedback, ideas and often crazy ways of using my work (like printing out an Excel dashboard I designed onto a 4 metre wall in the boardroom. True!)
What is the number one piece of advice you give to aspiring data scientists?
Don’t assume you know the problem you’re being asked to solve. I’ve a blog article coming out on this soon but I’ve distilled this down into the “3x knows.” These are “I don’t”, “I think I” and “I do”.
Most people start at “I think I know” and it just ends in disaster more times than not. Start with “I don’t” and get the person you’re doing the work for (your “customer” i.e. the person who will judge the value of the output) to tell you in laymen’s terms what they are expecting.
Repeat this back to them in your own words. Get them to tell you again. Then get them to email you a brief. Then check again!
It is a pain but it is better to have this exchange at the beginning rather than at the end …
Where can readers find you online?
Please sign up to the newsletter at www.yourdatadriven.com if you want to hear me share some thoughts each week (there are other benefits of course!)
If you’re a pro sports team, contact me at www.paceinsights.com.
Also I’m on LinkedIn – Samir Abid.