Assorted links


I once bought an umbrella for $5 to get me from the West Side Highway to Penn Station. When I got to Penn Station, dressed in professional business attire and mostly dry, I offered the umbrella to several passers-by. None would even look me in the eye, assuming I was trying to sell them something. I eventually left it on top of a trash can, only to see it be picked up by an umbrella salesman.

3. Humans might be able to tell the difference between real charts and randomly generated ones, but I doubt the field of "technical analysis" could.

The beard article reminded me of the advice to a bald person who is going to start wearing a hair piece.

First, grow a beard and wear it for a while so people get use to you in a beard.

The first day you wear the hair piece shave off your beard.

People will notice something different. But if you point out it is the beard they will not notice it is also the hair piece.

Albert's main argument against Bayesian rationality seems to be that anything goes because priors are left undetermined by the theory. But doesn't the same issue apply to his proposed solution of critical rationality? You have to decide what counts as criticism and what doesn't, which means anything could happen.

So Brigham Young would've had to apply to grow his beard?

I just found a much better paper by the same author (Max Albert) on the same topic. I still don't agree with him, but this paper at least discusses non-informative priors, and there's an interesting "chaotic clock" prior that leads to an inability to learn anything through Bayesian inference.

And for comparison, here's Jaynes on Bayesian priors:

Mr. Albert fails to point out that the reason the chaotic clock prior cannot learn is that it contains no inductive bias (such as Occam's Razor):

People are too used to "There is no such
thing as a free lunch" to trust free umbrellas.
As for umbrellas forgotten in public places,
some will take them and others will not.
"Are you a therapist or are you on probation?"
in the conversation between strangers
cuold be a line in a novel about a modern city.

On the BYU beard thing from a BYU alum - BYU's stance on beards changed in the 1960s and really hasn't changed back. Beards are still largely associated with rebellion outside BYU though. Try to name even one lead actor in a romantic movie with a goatee. I wrote a post last year asking if anyone could name a goody-two-shoes in a goatee, reposted over here. So far, I have no satisfactory answers - any good guy in a goatee (e.g. Three Musketeers) is there to kick trash. As a clean-shaven guy whose wife has him wear a goatee much of the year, I would like a few more good role models.

As far as the financial time series go, the only thing proven. by the experiment is that the stochastic process followed by the real data is different (in a distinguishable way) from whatever model the experimenters used (presumably either an arithmetic or geometric Brownian motion). This in no way proves that markets are inefficient; simply that the experimenters are poor modelers.

Comments for this post are closed