The whole point of the method, as far as I am concerned, is that it optimizes the average case under the assumption that you don't have an edge on bull/bear calls.
There are still pathologically pessimal cases, of course, but not under typical assumptions about the nature of a financial time-series.
What I love about aminorex is that he's usually right and I learn one or two new things every time I try to decipher what he's written: http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=0003uY
Could you decipher this one?
https://bitcointalksearch.org/topic/m.11003261
I've always thought of most of cryptoland in terms of the show put on by the Duke and King in Huckleberry Finn--people gets scammed for a sloppily put together show. But instead of riding the con-men out on a rail, they invite their friends and family to see it, so they don't feel alone in the duping. Though, thankfully, they eventually get exposed--ahem Evan, Josh....