I had an email exchange with Patrick Steigler the other day that amused me, so I’ll share it with you.
Steigler started with the subject line “Taleb and STDEV vs. MAD” and the message: I came across http://www.edge.org/response-detail/25401 and was wondering what your thoughts on the issue might be.
My reply: How come they never ask me for my opinions in these surveys?
Steigler: Maybe you need to write a book of aphorisms.
me: There’s this:
and then this, which some students collected from a course I taught:
http://www.stat.columbia.edu/~gelman/book/gelman_quotes.pdf
Steigler: So, what would you opinion be on the scientific idea that is ready for retirement?
me: I guess it’s a good thing they didn’t ask me!
Steigler: How about the retirement of p value < .05 for significance? me: Yeah, but that's too easy. It's like if someone asks for a pop music recommendation and you say The Beatles. Steigler: True, but the idea could still use some drum beating about since it does not seem to be as fashionable as The Beatles yet (but hopefully some day!).
Steigler: How about the retirement of p value < .05 for significance?
me: Yeah, but that’s too easy. It’s like if someone asks for a pop music recommendation and you say The Beatles.
Ha. My best laugh of the day so far (of course, it is only 10 AM), but still, it’s likely to stay on top for the rest of the day).
Instead of asking which ideas need to be retired, they should have asked “which people need to retire?” The latter has far more to do with the advancement of knowledge than the former.
You have a good answer? Say, in Statistics.
Yes.
A colleague of mine used to say “Where there’s death, there’s hope.” (I was never comfortable standing close to him.)
Also, I read that list of quotes and was surprised how pro-Bayesian many were. You come across less partisan on the blog.
After one deployment to Iraq, my Marines presented me with a similar list of my quotes. One of my favorites was “Being Canadian is un-American”. apparently this was said with extreme certitude, but I don’t remember saying it all. I can’t even imagine what the context was since unlike Afghanistan, there were no Canadian troops in Iraq.
Another favorite was after I saw the effects of a mission one of the Marines had completed, I said “This turned out no where near as sh***y as I though it would be!” The marine was extremely proud of that quote because it was the only compliment I had ever given him.
One of the benefits of pessimism is that every once in a while you end up pleasantly surprised.
Taleb has to be one of the most overhyped quacks of recent years.
Really? I know nothing of him but using MAD over STD is motherhood and apple pie as far as I’m concerned. (Actually, I like NMAD = MAD/0.6745 as then NMAD=STD when the data is normally-distributed and n->inf.)
Yes, so either you repeat the obvious making it sound revolutionary or you say sensational stuff that’s not obvious at all.
Ah, got it.
I looked through about 20 of the answers and see that most respondents have sort of cheated. They presented two ideas A and B, A is more popular, but they prefer B. OK. But why should it be the grounds to retire A? It’s not like every idea you disagree with should be taken out and shot.
Two things stand out in the article by Taleb.
1) Danny Kahneman. Do they know eachother on personally and is it important that this is conveyed to the reader. I know that Andrew is sometimes referred to as Andy. Even, if I could refer to him as Andy, I am not sure why I would do it in writing for a general readership…
2) The whole article seemed more or less reasonable. I am not sure whether it is indeed the case that people misinterpret the statistic the way he claims, but this could well be the case. I am however baffled by the last few sentences… I have no idea what it really means (what is the connection between skin-in-the-game and using some kind of statistic) and why he feels compelled to provoke.
Re: 1), as far as I could tell the thesis of The Black Swan was something like “Experts are morons and frauds, and you should listen to me because I’m best friends with Mandelbrot.” So, I dunno if it’s important, but I’d bet ten bucks it’s important to him, which maybe means it should be important to the reader too…
Mandelbrot was much closer to being one those expert frauds than Taleb would ever admit.
I’ve always found this quote attributed to Taleb rather annoying:
Any person, who can find only one other person he can intellectually respect exudes arrogance & hubris to an excess.
Hey, stop that, everyone! I like Taleb and I found his books to be thought provoking. See here and here.
People can have good or interesting criticisms and still have no clue what to do about it. Talebs solutions are much lamer than his criticisms.
Tyler Cowen’s book “The Great Stagnation” struck me as very similar. Most of it was great, but when he got around to the solutions part his “best” idea was to raise the social status of nerds. Yep – if only geeks got more girlfriends science would progress much faster. Super lame. It makes Taleb’s “replace StDev with MAD” look positively normal lame.
Cowen really is in a different league. Cowen has breadth. Taleb keeps flogging a small set of theses over & over again.
Taleb was faced with his very own black swan event as author/thinker, when the explosion of the derivatives market gave him an opportunity to spread his ideas about probability theory, which as he repeatedly insists in the Black Swan is a deadly boring pursuit. Given his pre-explosion “Fooled by Randomness” , he could now earn a living in the non-scalable world of book sales by “juicing up” (now with 3/2 times more fraud-busting fractals!) his earlier work. Don’t skim over the boring, broad spectrum name-dropping nor entire chapters that Taleb himself recommends that you not read. He warned you!
His invocation of chaos theory is a perfect example of his laziness. He highlights Mandelbrot’s glories in the main text while burying the problems with applying it to the real world in the endnotes( the Butterfly effect e.g.). Without careful implementation, systems theory is just another Platonic prison that Taleb bitterly protests against. Having said that, I did find his core ideas stimulating and it caused me to think how to deal with rare events in my own work. Black Swan would have been an insightful at 100 pages, but instead felt very bloated. As for the solutions,I have not read Antifragile. Is it more of the same?
Kahnemann and Taleb at the NYPL talking about Anti-Fragile: http://youtu.be/MMBclvY_EMA