A total of 42 artists across Europe (and Australia) are gathered in Kiev, Ukraine, and they’re all ready to battle it out at this year’s Eurovision Song Contest. If you’re not one for shocks and surprises, then fear not, for Microsoft has unveiled Bing’s predictions for this year’s competition, slating Italy as the country most likely to walk away from the arena with a victory.
Microsoft’s Bing Predictions team has utilised real-time predictions based on data analysis, shifting preferences, and circumstances to peer into the future to predict which country’s act will walk away victorious.
After all the data dust cleared, Italy’s Francesco Gabbani, with the song Occidentali’s Karma, was predicted as the most likely winner, with a 19.8% chance of success. Sweden’s Robin Bengtsson follows with a 17.1% chance of winning, while Kristian Kostov of Bulgaria follows in third place, with 9.9%.
The contest itself is spread over three days, with the first of two semi-finals set to kick off on May 9, followed by another one on May 11. The final will take place on May 13, and Microsoft’s search engine team does not expect good news for the Czech Republic, Georgia, San Marino, Iceland, and Malta – all of which have just a 0.1% chance of making it into the final.
Behind the crystal ball
These predictions aren’t simply plucked from thin air. In 2015, the Bing Predictions team correctly predicted Sweden as the winner of the competition using data analytics. The same team also went on to correctly predict all top five finishers in the Abu Dhabi Grand Prix in 2016, while it predicted the 2017 Oscar winners with an impressive 71% accuracy.
A combination of three main data streams are analysed in the predictions process– namely web data, social media data, and traffic data – and this information is collected and paired against historical data and past outcomes to create reliable predictions for the outcome of the competition.
The predictions team kicks off with a web data stream which pulls in online content on everyone participating in the competition. Social media data then looks and the online activity and chatter taking place across Europe, drawing from both Twitter and Facebook in real time, using several filters to cut through the huge volume of online conversations.
Content is ordered, for example depending on whether it’s been posted by an influencer with a significant following, if it’s an original post or a retweet, or if it’s positive or negative. Real-time search traffic is then used to understand how interest is waxing or waning. Finally, the model considers where the traffic is coming from. Information is weighted in favour of real-time data like live social media conversations and incoming search traffic, meaning the closer we get to an event, the more accurate the predictions.
All data included in the analytic models are of course anonymous and is impossible to attribute any indicators to an individual user.
The power of predictions
The algorithms that power our predictions play a significant role in the broader digital transformation that is currently happening across industries and continents. The same algorithms can be used by enterprises of all sizes, while these types of data streams offer invaluable insights into what people are searching for and thinking in real time.
It’s even possible for businesses to apply the algorithms to their internal production processes, helping organizations to determine the best time to go to market with a new service, or to learn about product issues as they arise.
Recently, Satya Nadella talked about the four pillars of digital transformation. One of the pillars highlighted is optimizing business operations. In other words, it’s not enough to know what’s happening now in a company’s business – it’s also crucial to anticipate what will happen next and prepare to capitalize on that insight.
Here is where machine learning and the Internet of Things can help. These services use millions of data points from processes and sensors to recognize patterns and generate predictive insights.
If you’re a fan of this year’s Eurovision Song Contest, check the latest predictions to see how the contest continues to unfold as we approach the finals. Will Italy reign supreme on the night, or will another star shine brighter in Kiev?
Until the final results are announced, the only way to know how your favourite contestant is really doing is to keep an eye on their progress – and hope they claim at least a little more than nul points.
