Great piece in the FT (below) on the mis-use of mathematics in finance. The target of criticism are those who over-fit past data to profess a successful trading strategy for the future.
It is not a long shot to extend this to explanations of entrepreneurial performance. To a large degree, we tend to look for patterns in past performance data, but are these really useful as guides for the future?
Because we know what actually happened, we can always find an explanation for it. But since what happened cannot be assessed probabilistically among its many alternative paths (which did not happen and which cannot be really enumerated) there is no weight to our explanation in guiding us towards an open ended future.