Seattle Sounders score with data visualization fed by wearables

SUMMARY:

Seattle Sounders FC makes extensive use of Tableau based data visualization, fed by wearable devices and GPS data aimed at keeping players at peak fitness. It’s working.

[sws_grey_box box_size=”690″]SUMMARY – Seattle Sounders FC makes extensive use of data visualization, fed by wearable devices and GPS data aimed at keeping players at peak fitness. It’s working.[/sws_grey_box]Seattle soundings standings in 2014

Sports science is becoming increasingly sophisticated. At Seattle Sounders FC, David Tenney the team’s fitness coach explains how its use of wearable technology, GPS data and data from triangulated video shoots are used to build fitness profiles that are visualized in Tableau and then used to optimize training and fitness plans for the star players.

But before getting into the meat of this he warns:

Team coaches don’t usually come into the sport with statistics degrees and when you look at how the the most successful coaches operate, they tend to overtrain some players and undertrain others. So as someone who uses data to predict outcomes, you have to show something that makes sense to them. Otherwise, don’t do it – you’ll waste your time.

What does Tenney use on the device side?  No consumer grade crap for them.  Instead, Tenney has amassed a veritable geeks paradise of gadgets, devices and technology including:

  • Catapult GPS
  • Polar Team 2
  • OmegaWave
  • Match Analysis
  • Fatigue Science

Of these, Tenney finds that the GPS systems used for measuring the amount of effort a player is putting in are almost useless inside the stadium but that the accelerometer technology works well. To compensate for the GPS issues, they use a set of cameras set up in an XY configuration so they can triangulate player activity from the resultant video footage. How does Tenney use the data?

He offered an interesting subtext to the plot that describes how the culture of sports players has to be managed:

The problem with pro athletes is that as they get more accolades, they don’t always do the things needed to stay in good shape. Using the metrics allowed us to get player buyin. So for example we could correlate things like poor sleep patterns with a likelihood of injury based upon a player’s fitness recovery. We can better predict who will break down and when and that’s a big deal when you’re trying to ensure that players are fresh.  But then when you’re forecasting say 23% will fail, the coach turns and says – well I’ve still got 77% likelihood of everything being OK.

Sleep patternsIn this case, Tenney credits the buy in of head coach Sigi Sdhmid who was an accountant in a past life before turning to soccer coaching. Schmid has a background in numbers that made it a lot easier for us to persuade the team as a whole to both trust the data and use the devices.

Why is sleep so important and why should it be measured? According to Tenney, across all sports, it has become known that the very best performers regularly get 9-10 hours of sleep per night.

In the image to the left, taken from Tenney’s presentation, you can see for instance a grayed out area where a player has not been getting enough sleep on a couple of days and the impact that has in the successive four to five days in terms of peak fitness.

The algorithm used by Seattle Sounders plots the impact of poor sleep habits and provides direct feedback to the players. They can easily see and understand how their bodies are responding and adjust behaviors accordingly.

What about other aspects of measurement? Sports fitness is an evolving science that is increasingly dependent upon data. Tenney quotes the example of a period last year where the Seattle Seahawks used technology to figure out how to best optimise player fitness while going through different time zones.

Are you impressed?  I was and especially listening to Ravi Raminen, the person responsible for building the algorithms and providing the initial visualizations that help Seattle Sounders make the best team selection decisions. As a programmer, he was used to being a rock star but in this field of endeavor, he had to learn a supporting role.

The real assets, the real stars of the show are the players. You’re there to support them and to gain their trust. We have to do that all the time because many of them have little obvious incentive to be compliant in wearing heart monitors or sleep measuring devices.

The results speak for themselves. While the season still has some weeks to go, Seattle Sounders are at the top of their league, two points clear of their nearest competitors. Where they end up is still up for grabs and both Tenney and Raminen are quick to acknowledge that building predictive models and individual fitness plans for players remains a moving target. Even so, both remain confident that the quality of data they’re hefting is good enough for them to be confident about outcomes.

Disclosure: Tableau Software funded most of my travel and expense for attending Data 2014

    Comments are closed.

    1. ladylaff says:

      Interesting. So does this mean data is replacing dope in elite sports?

    2. says:

      ladylaff Funny but as you know, dope testing is routine these days. More to the point, Americans are fanatical about sports stats and so this fits in very well with a way of viewing things with which they’re already familiar. The results are certainly worth assessing.

    3. ladylaff says:

      dahowlett ladylaff the story about the Oakland As use of data to assemble the optimal team was quite interesting. Admittedly, I learned about it through ‘Moneyball’ but based on a true story about data analytics. It’s interesting to see how the Seattle Sounders are using data to examine things related to individual performance, like the consequences of sleep deprivation.

    4. says:

      ladylaff dahowlett I got to listen to the author of Moneyball yesterday – really interesting stuff. More on that in an upcoming story. 😉