I get all the statistical egghead theory on statistical...

I get all the statistical egghead theory on statistical principles on trawling RCT until you find something you like. Or just adding 100 endpoints until you strike gold.

However, I am not ideological enough to ignore data and signals on large population subsets. Endpoints, which had they pre specified would have had met overwhelmingly approval requirements for FDA.

However, that isn't enough for me. I need to understand why did they not pre specify it. In MSB's case, I am very comfortable as to why they got it wrong which I have partly elaborated previously and will just add the ppl responsible for the trial design were also in error, naive, inexperienced, incompetent and or over confident. But ultimately, they were lacking the raw data from large population sizes to maximise their chances of success. This will no longer be an excuse.