Thanks for the thoughtful response Alex. The “flaw in my argument” is not a flaw at all — in the FAQ, I speculate about whether there is an optimal response to a “sufficiently funded” commitment contract and I say it’s likely that there is not. You agree.

Perhaps you think I should have been more explicit about what “sufficiently-funded” means. I have written about the 1/2 + epsilon solution on reddit and elsewhere but, well, there’s a reason I didn’t… Do you think that’s the likely outcome of the game?