Curious Case of Failing Prediction Markets in India (or Not)

Here is an interesting article that summarizes the failure of election predictions in India, at both national the state level.

This report by IDFC Institute says not a single survey since 1996 has been accurate (when defined at 5% confidence interval level). And the worrying trend is rather this :

“Accuracy levels of these surveys show no signs of an improving trend since 1996 though sample sizes and number of surveys have increased substantially.”

Now if more data is not improving your accuracies then one needs to question the sampling procedures these surveys employed.

On a related note, some of the smartest enterpreneurs are already on the task to solve this prediction crisis….truly in an Indian style ofcourse.

Stars Lining up for Indian Astrology startups

All those graduates of the Astrology programs have to be employed some where.

Data Scientist by training with diversions into economics, politics,technology and trivia in no certain order.