Big Data and Analytics – World Cup Use Case
Everyone says it was the best world cup, in the history. However, it was certainly not predictable. Brazil was a favourite of many people and they capitulated 7-1 to the winner, Germany. Spain and England got knocked out only after playing two games and Costa Rica made to the quarterfinals. For a lot of people, the results of the World cup were evidence that football is the exact opposite of sports analytics. The notion was summarised by Nate Silver who used to be an analyst in New York Times and now ESPN and got popularity during the 2012 presidential election. The guy was famous for never making a wrong prediction; he seemed to possess the power of prediction.
The Soccer Power Index (SPI) gave Germany a 35% chance of a win and a paltry 0.022% probability that they will score seven or more goals (almost 1 in 4,500) and Nate silver the prediction champion swore by it. There is nothing extraordinary about a 65% chance of losing, problems arise and data should not be judged on accuracy of every single result because it is just a prediction tool and not a guaranteed way to know the outcome of a game. The problem that the result was so inaccurate, according to the ELO rating shift it was the most unanticipated scoreline ever. It does lead to the question whether Big Data need to be ostracised from football or not?
Germany would be the team to be 100% against excluding the Big Data and analytics from the sport. A lot of people refer the crowd as the 12th man, however for Germany the 12th man was data. German’s pre-match training was altered because of analytics. They analysed every player’s performance and calculate key indicators such as the number of touches, movement speed and average possession time. In 2010, German’s set an objective to decrease their average possession time in which Big Data played an important role. Their average possession time in 2010 was 3.4 seconds and it went down to 1.1 seconds in 2014. For a lot of people, this played an essential role which lead them to win the World Cup and why Brazil and Spain which played evidently slower games and failed to live up to expectations. The analysts may have had it wrong with their pre-world cup predictions; however, to go to the extent to say that Big Data is not welcomed in the football world is a risky precedent for the professionals looking to get the best from their players. If the analytics can work for Germany, there is no reason why it cannot work for everyone else!?
Google Trends Missing The Point In Big Data?
Maybe the biggest and most renowned Big Data flop happened in 2013 with Google Flue trend. In 2008, Google launched this service with an aim to predict flu outbreaks in 25 countries. There was a simple logic: analyze the Google search queries about the flue in a given region. After that, compare the search results with a historical record of flu activity in the geographical region. On the basis of these results, the activity level was classified as low, medium, high or extreme. Although at initial glance, this might look like a cool idea, it, in fact, was not. In 2013 when the flue was at a height, Google Flue trend flopped. In reality, it was off by an astonishing 140%. The cause was the algorithm was faulty and didn’t consider a lot of factors.
Data Governance has been lacking
A lot of time organisation doesn’t completely comprehend the data that they possess but they still settle on undertaking a new project based on it. There is definitely a lack of proper documents, storage, policy and other processes regarding data management. It is a better idea to go to a big data consulting company in order to get your business a clear roadmap and instructions on how you can manage the data that you have and just after that try to overcome the challenges of big data.
Unclear Strategies & Blurry Goals
There are a number of IT terminologies and marketing terms available, and it can be a lot hard to make sense out of them. In addition, there are several big data products available out there in the market and it is actually hard to pick the right one. Before you select anything, it is significant that you figure out what services and technologies you require to achieve your goals.
Is Data Science too hard to understand?
Data science and big data are a difficult mixture of domain knowledge, mathematical, and statistical knowledge and programming skills. However, simultaneously, it has to make business sense. Generally, the IT department will create modifications that management may not understand and vice-versa. Ensure that your big data actions seem sensible to both IT and business leaders. Make a bridge between IT and the business in the big data plan. Business people must be totally involved in every stage of a big data project.