The American play Solitaire House was released with 9.1 points of praise, and its production team took advantage of nearly 100 years of data, based on audience feedback on existing programs, what kind of actors they like, where they like to play repeatedly, when to click to play, when to stop ... And so the data produced a series of only one senator; however, another team (Amazon) in the same period analyzed almost the same data, and finally they produced a series of four senators starring, with a rating of only 7.6, slightly above the average score.
The same data produces different results.
Even Google has had this problem. Google, according to a large number of data said can predict the outbreak of influenza, the first few years to do well, until a year of failure.
When people do things, they are divided and merged in two steps.
Data and data analysis are very useful, they are the role of the problem from the overall decomposition, every detail becomes clear. So do they have the ability to merge?
The brain can! Data and data analysis are good tools, but you can't make him decide for you!!
Data analysis sometimes results are not satisfactory