Sport Informatics and #UCSIA15

300

Without a formal qualification in Sports Science, I’ve had to work hard to understand the reasons behind the types of analytical work I’ve undertaken for clients. My degree was as far from sports science as you could get (business and engineering), but I’ve built a working knowledge of performance, athlete monitoring, and the types of statistical analysis professional sporting organisations need.

That experience does count for a lot. Then I heard about a free course, Sport Informatics and , that seemed ideally suited to my existing skills and experience. It’s free because it’s short, and it’s short because there’s a lot of self-guided learning involved. For a linear, process driven person, it’s an unusual, almost alien, way to learn.

And so I began the course, a new dedicated notepad at the ready, and my trusty four-colour Bic.

The course is designed around three streams; Gates A, B and C. A for “general admission”, new-comers wishing to be guided through the ideas and concepts. Gate B for volunteer access, those with experience who prefer self-directed learning; and Gate C coach or manager access for those with a specific interest and specialist knowledge.

I chose to be led through Gate A.

Sport Informatics and as a title seems right up my street. But oh boy, were the first weeks an eye opener. There has been some deep, deep water: I would never have thought I would sit through a 68 minute Stanford University lecture on Machine Learning. Thanks to Wikipedia I now understand what Occam’s Razor means (of all the things it could have been, the most simple explanation was best). And “R” is no longer just the eighteenth letter of the alphabet.

There have been plenty of softer subjects too, playing to my strengths; story telling with data, vs strategy, appropriate types of chart to visualise data.

I’m almost three quarters of the way through the four week course now. The Performance Monitoring theme has definitely helped my understanding. In many ways it’s reinforced the initial education I received from Nick Broad.

I genuinely had my eyes opened to the possibilities of data visualisation in sports by Nick. The two major projects we worked on together would have been transformational for the organisations he worked with. The course content is delivering the same message, albeit a few years after Nick evangelised to me.

The biggest takeaway for me so far is how similar elite sporting organisations are to the more “normal” businesses I work with. Everyone has data, everyone knows they need to do something with it, and rarely does anyone know what questions should be asked.

Big technology companies offer to organise data with a high-profile partnership, but analysts and coaches still struggle with spreadsheets. They capture every piece of data they can just in case they stumble over the silver bullet for success. A 25 megabyte spreadsheet is not unusual.

But they simply can’t see the information for the data.

No one needs this “big data”, they need questions answered. Those answers need to be in context, relevant, and actionable.

The Seattle Sounders, who I know also worked with Nick, state that “the real value is blending all the data together”. It is, almost.

Nick and I built huge data structures linking every GPS, HR, subjective, objective, and contextual data point together. It allowed questions to be asked across multiple contextual themes.

But the bottom line for me now, is not creating value by blending all the data together, it’s blending just the right data together to answer the specific contextual question. Big and more doesn’t mean better, it means less time to analyse and find answers.

For now, with one week of the course to go, I’m left intrigued by John Lythe’s “night time” role of helping sports scientists with their Excel skills (ExcelTricksforSports). I’m not certain that for a sports scientist acquiring additional technical skills is the answer, but at least it’s something I understand and can build upon with my own software tools and experience.

David Penny

300

Tags: , ,

One Response to “Sport Informatics and #UCSIA15”


Leave a Reply

Your email address will not be published. Required fields are marked *