He is making a wise point
If you get to pick:
The dataset
Code the exposure yourself
Pick the analytic plan
Let 10,000 people run the study in parallel, inc believers
Publish the results that fit your narrative
You don't have science
You can get any answer you want
🚨BREAKING. Inspired by this NEJM study, I now used the difference-in-difference method to assess the impact of lifting all lockdown measures in England (aka 'freedom day'). It turns out that lifting all restriction measures reduces case rate by 14.5%!