Motion Graphs Sports Analysis
I am a huge fan of TED.com. TED started in 1984 as a conference where the founders would invite some of the worlds most interesting people to the stage to deliver a presentation no longer than 18 minutes. It’s main premise is to discuss Technology Education & Design. The conference costs a small fortune to attend – and even if you have the money you still have to apply for access!
The good news is that all the talks are made available free online. The fact that the talks are short (18 mins) makes it an excellent format to dive in listen to a talk and then continue on with your day. Some may be related to your field but often even when they are not they can be uplifting and motivational.
My favorites;
- If you are working in education or teaching young kids anything I really advise you to check out Ken Robinson says schools kill creativity.
- For an example on how to make stats interesting, exciting and just to see an unbelievable presentation Hans Rosling shows the best stats you’ve ever seen.
- And my final recommendation is to have a look at David McCandless: The beauty of data visualization. Although the data used in this example is not applicable to sport I think the speaker throws up some very interesting questions about how we actually present data. Maybe some ideas you can adapt to your analysis work?
So what has all this got to do with sports?
Well, when I was watching the Hans Rosling talk I thought the software used was pretty innovative (never seen anything move in Excel before) so I thought I would see if it could be applied to sports analysis. I picked just one example of possessions of each player in 5 minute intervals. But you could easily look at any aspect of your key performance indicators and you could certianly look at combining multiple games.
Below is a quick video I have made explaining the process of creating a Motion Chart and some of its key features.
I have also embedded the Motion Graph for you to have a play with.
This type of analysis won’t apply to all data, but I think it’s worth examining the power that improving visualization might have on your stats. I will certainly be testing this out on some real data in the near future.
Any thoughts?
