A month and a half ago I announced a $500 bounty/prize to explain infrabayes work in the language of game theory. Others added another $980 to the prize pool.

There were no entries which did exactly the thing I asked for, but we did get one post which I think provided an excellent 80/20. Or, well, almost 80/20, as implied by the payout: I have decided to pay out 75% of the prize money to David Matolcsi's A very non-technical explanation of the basics of infra-Bayesianism.

David's post is the best conceptual explanation I have seen to date of the core ideas of infra-bayes. It is the only intro I've seen which presents things in a significantly different conceptual frame/language than the original, which I think is very much necessary in order to understand what's going on. After reading it, I feel like I have a decent intuitive idea of what infrabayes does/doesn't do, and roughly how/why it does/doesn't do those things.

The main thing David's post does not do is explicitly tie any of the concepts to the math. Or use any math at all. Which is a reasonable choice - as I said, 80/20(ish).

Thankyou David for writing a very useful post! Again, the post is A very non-technical explanation of the basics of infra-Bayesianism, and I strongly recommend it for anyone who wants an intuitive understanding of what's going on with infrabayes.

New Comment