Tuesday, March 14, 2006

"Fooled by randomness" reviewed by a "Statistical Modeling, Causal Inference, and Social Science" Blog

Here's a blog entry book review of Taleb's "Fooled by Randomness" [1,2] by a statistician. The main page for the blog is here and the author statistician Dr. Andrew Gelman's homepage has links to his papers. There's a lot to be learned from this blog entry:

"I prefer writing books to writing journal articles because books are written for the reader (and also, in the case of textbooks, for the teacher), whereas articles are written for referees. Taleb definitely seems to be writing to the reader, not the referee. There is risk in book-writing, since in some ways referees are the ideal audience of experts, but I enjoy the freedom in book-writing of being able to say what I really think."

[Another good argument for peer review. Shouldn't it be qualified as "experts within a discipline or school of thought". For instance, both a novelist like Balzac and a statistician might touch on the theme of contingency and human action and even address the same points, but since they don't speak the same theoretical meta-language the parallels in the note would be only rarely noticed or noted down?

The beauty of Taleb's book is that his ideas cut across disciplines. Most historians will recognize similarities between Taleb's little potted biographies of traders blowing-up and losing their entire accumulated trading profits in a few weeks and the real-life historical narratives of generals and political leaders and recognize the relevance of formal probability and statistics models, problems, and theories to separating the role of contingent human agency from deterministic long-term structure that exists in most historical narratives.]

"Taleb's general points--about variation, randomness, and selection bias--will be familiar with statisticians and also to readers of social scientists and biologists such as Niall Ferguson, A.J.P. Taylor, Stephen J. Gould, and Bill James who have emphasized the roles of contingency and variation in creating the world we see."

[It's nice to find names you're not familiar with: Niall Ferguson has relevance for the structure vs. human agency problem cited above:

"Another controversial aspect of the Pity of War was Ferguson's use of counter-factual history. Ferguson presented a counter-factual version of Europe under Imperial German domination that was peaceful, prosperous, democratic and without ideologies like Communism and fascism."

"Ferguson is the leading academic champion of virtual history. Ferguson likes to imagine alternative outcomes to history as a way of stressing the contingent aspects of history. For Ferguson, great forces don't make history; individuals do and nothing is pre-determined. Thus, for Ferguson there are no paths in history which will determine how things will work out. The world is neither progressing nor regressing; only the actions of individuals will determine whatever we live in a better or worse world."]

The papers at the bottom of the blog entry are certainly hard-going. I took a graduate course in stochastic processes from Stanford many moons ago, but it wasn't until I read Taleb's book that all those abstractions like ergodicity came alive for me. Bridging the gap between the clarity of math and the messiness of life is not easy and Taleb has found a way of bridging this gap with his prose. The clarity of mathematics can clarify patterns in history, but is inaccessible in its usual dry form. Take the work of Turchin for instance. His books go a long way towards making mathematics more accessible, but Turchin fits long-term historical data to the model, whereas Taleb has simply noticed fits between the data and real life events that he has witnessed in his trading career. The work of both Taleb and Turchin is probably prescient of future work to come that bridges the gap between humanities hermeneutics textual-based insights into human history [http://en.wikipedia.org/wiki/Max_Weber] and math-scientific law-based insights.]