Nice program. And nice to have a fellow Bayesian in the forum.
I agree that would be ideal but I updated the port ranking system August 2024. I don't think that is enough data. Wish I had more or better.
Harold Jeffreys developed the Bayesian t-test which uses effect size as the prior. To be honest I can read thru the explanation for this and sort of understand it but usually fall asleep. Effect size is the mean which I set to zero above divided by the standard deviation (of the difference in returns for the port and sim). BTW, my original study used a normal distribution for the prior also. Jeffries liked Cauchy priors (which you may see below) because they are "conjugate" when doing the math and generally less informative.
Yes this is a good way. The makers of JASP agree with you and do exactly that. I have included a screenshot of what JASP did with a range of priors (JASP's defaults):
