A Word In Edgewise: Through a Glass, Darkly…

I read a couple of books recently that, while I’d not like to take a three-hour final on them, nudged my brain out of neutral and set me thinking about subjects I’d previously associated with ten-foot poles.

It started when I opened Nate Silver’s recent “On the Edge: The Art of Risking Everything;” not because I’m planning to make a splash in Vegas, but because the world in general seems to be upping the ante, dancing on the brink, pushing out ever further the parameters of Risk.

I first encountered Silver’s name in 2008 when we ran a blurb on his FiveThirtyEight in Lavender’s “Big Gay News” column. FiveThirtyEight (the number of members of the Electoral College) was a blog that delved into political forecasting. Silver had previously been a professional poker player but moved on when Congress passed the Unlawful Internet Gambling Enforcement Act (UIGEA) in 2000, pretty much drying up that source of income.

A lot of folks, writes Silver, examined his statistical model statistics, and thought him a fool. But as Silver explains, from the standpoint of skilled gamblers, his was a good forecast: his model forecast Candidate T’s chances at 29% while the market price was 17%, so for every $100 placed on T, you’d have a return of $74. If you’d bet on those odds, you’d have won a bundle — without necessarily having to vote for that candidate. “My job,” he stresses here, “is to handicap the race — to tell you how you should bet.”

This volume, as well as his 2012 best-selling “The Signal and the Noise,” explores how to examine and make forecasts for almost any issues requiring decisions: politics, stocks, weather, health, betting on games and sports — poker, chess, baseball, football, what have you. There are levels, and some, particularly earthquakes, are intrinsically less predictable than others.

Silver is not offering a magic key to winning, for there are no certainties, but rather factors more suited to forecasting desired solutions. In this era of proliferating Big Data and AI, data itself is all-important, and the more data the better … to a point. You must know how to use it, and know what percentage is usable signals, and what is merely noise. Winning is in the sifting.

Probabilistic thinking, the method designed by Thomas Bayes, 18th-century English statistician and philosopher and contemporary of economist Adam Smith, is the means Silver advocates. Bayes’ theorem is a simple [sic] mathematical rule allowing one to use “priors” — your already acquired data — as well as incorporating new information (the quarterback is injured and out this game) to update your predicted probabilities. As I understand it, a simpler, more dynamic system than insisting on accumulating all possible surrounding data, much of which won’t be relevant — noise, not signal.

Silver’s texts are briskly written and studded with interviews with sports luminaries, epidemiologists, professional gamblers and even felon Sam Bankman-Fried (aka SBF), bad-boy of cryptocurrency who made “some quite deranged statements about risk.” Silver describes his own guided tour through the art of risking everything as “friendly, informative — and occasionally provocative.” His examples are definitely thought-provoking, though revelations of the darker undercurrents of the River (Texas hold ’em jargon) recall Daddy’s nighttime jingle when his “This little piggy went to market” made your neck hairs horripilate with the sudden knowledge that the little piggy wasn’t destined to shop.

Both books interweave high drama, gaudy characters and intellectual fodder. A heuristic takeaway: overconfidence is a huge problem in any field involving prediction. Or, as Yogi Berra clarified, “It’s difficult to make predictions, especially about the future.”