The Strategist

Microsoft to predict winner of the UEFA Cup

06/08/2016 - 16:50

Microsoft will be predicting results of each match within the 2016 European Championship which starts on June 10. The forecasts will be made by a team of Microsoft experts using the most modern technologies, such as cloud advanced analytics and machine learning Cortana Intelligence Suite. According to preliminary estimates, the German national team will win over the Spanish national team with a 66% chance in the final championship.

Tsutomu Takasu
Tsutomu Takasu
Microsoft Expert Department with help of information technology will be predicting results of each match of the European Football Championship. For this purpose, the company has developed a special algorithm for analysis of the relevant data array. It processes statistical information on commands of previous games, performance of players, personal injuries, as well as information from the Internet relating to the Championship: news, queries, messages on social networks. The entire dataset will be analyzed by Cortana Intelligence Suite, and the projections will be formed on the basis of the results obtained.

There is a probability of 66% that the German team would win in the final, if the German and Spainish teams get there. It should be noted that the forecast may be adjusted depending on the scenario. In the first match between Romania and France, to be held on June 10, the Microsoft experts predict France's win by a margin of one goal. The probability of such an outcome is estimated at 71.4%, while the probability of Romania’s victory - only at 9%.

The forecasting methodology, built on predictions of results of various sports and cultural activities, is constantly being improved. Experience from previous forecasts is taken into account in future calculations. The predictions are based on statistical analysis of previous games, including information on each player, which allows the system to understand both strengths and weaknesses of each team member, and reflect this in its results as a whole. Analysis of the information field helps identify additional factors affecting a game’s outcome.

The data analysis also takes into account the collective knowledge. As shown by past experience, it adds additional 5% to the accuracy of forecasts. It is possible to identify the most unconscious guesses on the basis of search queries and likes, as well as the information that a huge number of sports fans share on social networks. Surprisingly, they do not always coincide with results of direct surveys of the same people.

This approach to predictions, increasing their accuracy, would not be possible without tools of advanced analytics and machine learning - a set of cloud services Cortana Intelligence Suite. So, they have helped the experts to determine distribution of seats in the show American Idol, and select the most promising players stayed in the team in the qualifying season of the US National Basketball Association NBA.