Why is hypothesis testing bad? Well, it is not bad when the decision requirement is Yes|No. Not everything is black or white. We may never know Schrodinger’s cat living status when it is indeed associated with probability. That leads to the next one, the confidence interval is somewhat helpful to understand the limit (or bound). However it is still will not tell you the probability of one vs another.

That is where comes I’m not keen of hacking the p value, or scatching head to calculate the confidence interval. Rather building an bayes model to explain the phenomenon.