So, the Euro 2016 started this past Friday. We’ll have matches almost everyday until July 10th.
And here at Ubiquis we decided to start a little project around the Euro. The basic idea is: suppose England gets to half time level with Slovakia. Given both teams’ past history, what’s the most likely scenario? A win for England? A draw? A win for Slovakia? We got score change information from all Euro finals matches since 1960 from Wikipedia and parsed it with PDI. And using a real time scores API called xmlscores.com, built a couple queries that answer one basic question: “out of all matches in which team A was leading/trailing by X goals or more, at the same point in time, how many ended up in wins, losses and draws?”.
So, for example, when yesterday Germany was leading Ukraine by 1 goal at Half time, the dashboard would give us some idea of how often Germany managed to keep a lead and win the match, or how often Ukraine managed to turn the tables around when losing by 1 goal or more and ended up drawing or winning the match.
The dataset is quite small and the data model very simple, but it serves to show how Pentaho C-tools and PDI can be leveraged to create a real time information system, fed by external data sources, in a reliable maner.
Here’s the link: Euro 2016 Analysis dashboard (no longer available).
Hopefully we’ll use the learnings from this project to expand its scope and reliability and use it for other sports competitions. Stay tuned.