Mathias Brandewinder on .NET, F#, VSTO and Excel development, and quantitative analysis / machine learning.
by Mathias 8. December 2009 13:58

How to value life is a thorny issue for economists and non-economists, but I hadn’t expected the question to be also a matter of lively discussion among dinosaurs. And considering the state of decision theory at the time, they do a pretty good job at it.

 comic2-1631

by Mathias 28. June 2009 12:41

I finished “The Drunkard’s Walk – how randomness rules our lives”, by Leonard Mlodinow, a few days after Sway; I have postponed writing about it, because a minor storm of work has hit my parts – but I really loved this book, and thought I should say a few words about it.

The book’s point is that humans are very bad at drawing conclusions from observations of random phenomena. They routinely make gross mistakes when dealing with conditional probability (92% of Americans, some of them with pretty solid mathematical credentials, get the Monty Hall problem wrong), and fall prey to the Law of Small Numbers, seeing patterns where there is none, and refusing to admit the importance of randomness in shaping our fates. In the end, the message is pretty upbeat – I loved this quote from Thomas Watson:

If you want to succeed, double your failure rate.

I really enjoyed how the book builds up following the history of probability and statistics; some of the individuals who contributed to its development are truly remarkable (just lookup Cardano for instance), and the book contains a fair share of anecdotes about them. If anything, it gave me my first introduction to Benford’s law, which I am still digesting – and which has been weirdly prominent in the news, via the Iran election issue.

In short, I strongly recommend this book if you are interested in either history of sciences, probability, or decision making.

by Mathias 14. June 2009 16:44

While travelling to Seattle, I went shopping to the bookstore and purchased “Sway – the irresistible pull of irrational behavior”, by Ori and Rom Brafman. Each of the book’s chapter illustrates a specific bias which leads individuals to make poor decisions, through multiple experiments or real-life stories.

If you are familiar with decision theory, and specifically with decision biases – the way our brain plays tricks with us and processes information poorly, misleading our judgment – you will probably not learn much from the book. Even then, it’s a pretty entertaining read; the style is light, and the anecdotes and cases funny and very clearly explained. If you are not familiar with the topic, I recommend the book: it is a good primer to recognize traps and make better decisions!

The part which was new to me was the chapter “Compensation and Cocaine”. It seems that monetary decisions and altruism are processed in completely different parts of the brain (no surprise so far), but also that when one is active, it shuts down the other. Furthermore, the monetary decision brain is processed in the same place as the center of pleasure derived from sex, drugs or gambling (not sure what this says about the financial sector…), which has very practical implication in designing incentives. From what I gather, altruism provides a stronger motivation than money – but surprisingly, mixing both doesn’t add up: the pleasure center will completely override altruistic motivation, which is then ignored – and reduces overall motivation.

Overall, fun read – recommended!

by Mathias 7. December 2008 20:42

The November issue of msdn magazine has a nice piece by Dr. James McCaffrey on determining a best option from a list of choices. It's an interesting departure from his usual topic, testing, and it proposes a discussion and implementation of the Schultze method. As far as I remember, Arrow's impossibility theorem demonstrates that

no voting system can convert the ranked preferences of individuals into a community-wide ranking while also meeting a certain set of reasonable criteria with three or more discrete options to choose from.

Therefore, I have to assume that there are limitations to this method. If I have time, I'll try to figure out in what cases these show up. However, the exposition of the problem is worth reading.

by Mathias 21. July 2008 17:45

Via Wired, came across this great website, with a more rockin' version of the "to be or not to be" dilemma:

funny graphs
more graph humor and song chart memes

Comments

Comment RSS