Friday, October 08, 2010

Casino Math

Casino Math

Tom Siegfried, editor in chief of Science News, kicked off our the 2010-2011 Linus Pauling Memorial Lecture series last night, at Portland's central Broadway venue, the Arlene Schnitzer Concert Hall, his name in lights.

Most of Tom's talk was a walk through the rise of statistics, from its inception in the world of gambling (hence my title). Cardano got the ball rolling, with Pascal and Fermat picking up the thread, then de Moivre and Adolphe Quetelet (a lesser known but one of Tom's favorite thinkers).

Quetelet was pioneering the notion of "human physics" i.e. predicting social trends and phenomena using statistical methods. Applications to molecular dynamics, stochastic analysis, was more a derivative of nascent sociological sciences than vice versa, a point often overlooked by those claiming the soft sciences are always trying to imitate the so-called hard sciences.

The arbitrary P=0.05 correlation threshold required to prove "statistical significance" has its own history, which was more spelled out in the printed program than in Tom's talk. Safe to say, a lot of studies misapply statistical reasoning, meaning drugs that shouldn't be out there are, and some that should be are not -- among other consequences. The process of aggregating studies to get a larger sample size and perhaps a more significant outcome is also regularly botched.

In general, people have a poor understanding of probability theory, which demands improbable events, flukes.

Then Tom went off the deep end in a culturally approved way, by discussing a "multi-verse" picture of branching universes. The discourse was especially interesting to those five-to-ten of us in the world who read Synergetics, as Boltzmann gets credit for "eternally regenerative Universe" ideas in Fuller's work, and here it was all about "Boltzmann brains" and their likelihood in alternative universes. I disagree that "infinite time" is an antidote to "never" i.e. I don't agree that what's "infinitely improbable" is somehow bound to occur given infinite time.

Ascribing meaning or sense to "infinitely improbable" occurrences is problematic to begin with i.e. how does one know if one is confronting such an outcome even if it does occur? If a monkey types a work by William Shakespeare (say Macbeth), why would we agree this was "random"? Perhaps the monkey was suddenly possessed by some well-versed ghost? Remote controlled by hidden grad student by means hitherto unsuspected? Who has authority to give the "true account" in this alternative world? In a multi-verse, anything would seem possible, including far-fetched explanations for incredible events. Sounds familiar doesn't it?

The entire exercise seems more like a safety valve, a licensed "letting off steam" in an otherwise pressurized system (alluding to Boltzmann again, and how his thinking is featured in Fuller's philo). Language needs to exercise its grammatical capabilities sometimes. Philosophy is a playground, largely unsupervised these days (so watch out for rusty nails, hidden pitfalls -- most your grownups are on vacation).

After the Heathman dinner of cod with eerily moving parts (some kind of garnish -- very eXistenZ), Tom answered some additional questions. He expressed a "Google is evil" point of view, i.e. here's a company dumping hard-won value added writings into the commons at no cost, undermining publishing as we know it.

No wonder science has become so much the sex slave of moneyed interests with axes to grind. There's almost no way to sustain oneself as an independent, self-respecting, objective private investigator anymore.

Science journalism is likewise under pressure to sensationalize, to hype murky findings, simply to stay in the game. The economics seem unsustainable, which makes the job difficult, sometimes discouraging.

I also entitled this journal entry "Casino Math" because of my Heuristics for Teachers on Wikieducator. Puritanical readers may have a problem with this nomenclature because of strictures against gambling, but it's true to the history of the field, as Tom's talk well illustrated. Universe has casino-like aspects regardless of what games people play, as one of the questioners pointed out, saying "Even if God does not play dice, He seems to have a serious gambling problem" (audience laughter).

I was happy Tara could join us for this event. We took the bus to and fro. She gamely ate the cod, if not the mushrooms, and shared many insightful and intelligent remarks about the talk and proceedings.

The next day, Terry phoned me at American Dream Pizza, where I was waiting (on chauffeur duty). He wanted me to accompany him in treating Tom to an intimate dinner with some PSU students interested in further discussion. Twas my distinct privilege.

We spent the evening at the McMenamins on the Park Blocks, quaffing beers, munching on snacks, and discussing the pros and cons of various philosophical positions, such as determinism versus macroscopic randomness.

Terry drove me home in his brand new Prius after leaving his distinguished guest at The Heathman.

Take Two