Community Spring Cleaning week is here! Join your fellow Maveryx in digging through your old posts and marking comments on them as solved. Learn more here!

Analytics

News, events, thought leadership and more.
LukeM
Moderator
Moderator

Predicting the Euros Webinar: Hackathon and Key Takeaways

 

On June 10th 2021 @wdavis and I hosted a webinar to discuss predictive analytics, particularly when applied to determining the outcome of football matches. This blog post will recap the key points we covered and provide links to resources and material to help further your exploration.

 

We are also running a hackathon with the workflows and data we presented. Please see below for details and downloads and get involved in the comments!

 

If you missed the presentation or want to re-watch then head over to https://pages.alteryx.com/euro-2020-on-demand.html to watch any time.

 

Key Takeaways

 

Analytics and Data Science doesn’t need to be complex

The nature of data science requires constant updating, iteration, and improvements. This means that its very straightforward to start out as you only need a basic model to begin with. Any speculation down the pub about who may win a match is a data science prediction – what’s stopping you translating that into an Alteryx workflow?

 

  1. Think of your business problem and make sure you understand it. Which of my customers should I target with advertising? What players are playing in the lower leagues that are future stars?
  2. Consider the “features” or data points that may provide intelligence and inform an answer? Have those customers bought similar products in the past? Are the players showing similar traits to current stars?
  3. Get hold of the data and drag it into Alteryx!

LukeM_0-1623328214480.png

 

 

Any analytics problem is just an analytics problem – footy or finance

All analytic projects require a similar approach and the same skillset. Learnings from the Moneyball film or the success of Dave Brailsford and his cycling team can be applied to your line of work. Additionally, why don’t you get in touch with other Alteryx users in your organisation from different departments and find out how they are solving problems?

 

There are always good takeaways to help you iterate on your projects. Some suggested reading:

 

 

Use the framework to your advantage

Leverage the simple steps of the Data Science Lifecycle as highlighted in our presentation. This is key to optimising your approach and simplifying the process.

 

Read this great article here which details all the stages and provides insight to assist you on your journey.

 

There is also fantastic self-paced training in the Academy which will walk you through the Data Science process. Check it out!

 

 

LukeM_1-1623328214531.png

 

 

Focus on the actionable insight. What is the desired outcome?

Very rarely is a project solely about a single data point outcome or producing a predicted value. Always keep in mind the reason and original purpose that drove you to try to solve the problem.

 

You may be trying to predict the likelihood of each of your customers to churn…but your end goal is actually to produce a report to each account manager of their key areas to focus. Or is it to feed information to marketing of where to target? Alternatively, are you looking for the drivers that persuade customers to stay?

 

We’re trying to predict football results. But really our focus was to educate and inform. So our model is simple and easy to explain and the reporting output interesting and engaging.

 

Ask why? Challenge the status quo

Billy Beane, Dave Brailsford, Elon Musk, Martin Luther King Jr, Tim Berners-Lee, Julius Caesar, Dean Stoecker…

 

Don’t be afraid to try something different, try to find a different answer or ask questions of what is commonly accepted as correct. The great thing about Alteryx is it’s quick to prototype and iterate so why not break something and see if you find a breakthrough.

 

In Moneyball, Billy Beane is taking loads of flak for his approach as the benefits are not immediately obvious. There’s a great line from John Henry, the Oakland A’s owner, as he tries to reassure Beane:

 

“I know you’re taking it in the teeth out there, but the first guy through the wall always gets bloodied. Always”

 

Credit Daniel on flickr - https://www.flickr.com/photos/57511216@N04/Credit Daniel on flickr - https://www.flickr.com/photos/57511216@N04/

 

Perhaps the greatest example of challenging the status quo in sport comes from the 1968 Olympics in Mexico City. A certain man whose name will eternally remain in sporting folklore changed an entire discipline. Dick Fosbury was awful at jumping and so set about taking his failure as inspiration to innovate. He took his innovation to the Olympics and won the gold medal. “The Fosbury Flop” is now the accepted method for high jumping and every gold medal since 1972 has been won using his method.

 

Can you do it on a rainy Tuesday night at Stoke?

This is our call to action. Get your boots on and get out on the field. Take your theories and run with them.

 

Details of the hackathon are below, get involved and we will see what we’ve got in the next webinar!

 

Hackathon

Be bold. Be brilliant. Be breakthrough.

 

But seriously, take what we have built and demonstrated and see what you can do with it. See this as a weekly challenge where you can use the finished result down the bookmakers!

 

  • Download the attached files and work on them. Add your own features, expertise, and modelling prowess to improve on our basic model.
  • Submit your packaged workflows in a post below with any information on what you tried and your findings.
  • Entries will close on Sunday 20th June so we can assess the submissions.
  • Prizes will be awarded after the tournament for the best and most accurate submissions.

 

LukeM_0-1623329127603.png

 

 

Further Resources

I’ve collated a list of references from our research as well as some useful articles that may help you with your Data Science Development.

 

Previous Analytics in Sport Blogs

Analytics in Sport - Introduction 

Analytics in Sport - Get Ready for the Euros