The Gambler and the Scientist
A day at the races teaches us about our unscientific notions of “science literacy.”
For reasons that are somewhat obscure to me, my son, for his 11th birthday, wanted to go to the races. So on a beautiful May Saturday evening we headed off to Charlestown Race Track in West Virginia. Upon arrival we commandeered a bench along the rail right at the finish line. Picking horses is not the sort of applied reasoning that you learn in fifth-grade science or math, so I proceeded to give my kid a short tutorial on how to assess the horses for each race based on the past performance charts that were printed in the program for that evening's races. By evening’s end, I had also come to realize what we really mean when we talk about “science literacy.”
The past performance charts are a marvel of condensed information. For each race, we were presented with incredible detail about every contesting horse, from its pedigree to the jockey's record to the performance specifics of recent races: the purse and track conditions, post position, position at different points in the race with split times, and position at the finish, of course, but also an assessment of the finish that is often a model of poetic density—"faltered under pressure," or "fast on the outside but too late," or "drove through the pack." From all this information, one can make informed comparisons about all of the horses in each race—and informed bets on the outcome.
Using the race charts to place smart bets requires sophisticated knowledge of how the innumerable variables that influence a horse's past performances translate into a plausible prediction of that horse's present chances against a slate of other horses (each of which must be similarly assessed).
Now, the stereotypical view of your average racetrack bettor is not entirely unfair. Few bobos or yuppies or Ivy degrees here. Lots of cigarettes and hats with heavy equipment logos and VFW patches, lots of RVs and American cars in the parking lot. A man, perhaps in his 60s, sat down next to me with a groan, declared that his foot was killing him, and launched into a discussion of his medical woes—diabetes, hypertension, and mental health issues. "My psychiatrist almost killed me by giving me a drug that caused my blood pressure to spike. Over 700. My wife tried to call me to tell me to go to the hospital, but I was out with a friend bringing in firewood. If I had been on a tractor or watching TV I probably would have died. I fired all my doctors the next day." This guy had a sophisticated sense of the limits of applying scientific knowledge to a complex system—the human body—when expertise is divided up into narrow specialties.
Later, a woman exhaled a lungful of smoke into my face. I must have involuntarily recoiled because she apologized and explained with some embarrassment that "all smokers aren't rude." She went on to say, "I'm just an addict. I've tried to stop, but that just makes me get even fatter." She was poignantly well aware of the lethal dilemma that her culture, her habits, and her physiology had created.
Moreover, both people were able to decipher the racing charts using that combination of factual knowledge, tacit experience, and informed intuition that characterizes the expertise of a real-world practitioner. They probably didn't take classes in statistics or risk analysis, and if they had, it wouldn't have been of much help to them. In this context, my Ph.D. in geology is a silly irrelevance. I am barely literate in the horse-racing domain, able to more or less break even after an evening's modest betting by being boringly conservative, but almost never able to come up with an analytical insight that allows me to turn favorable odds into a big win.
Daniel Sarewitz co-directs the Consortium for Science, Policy, and Outcomes at Arizona State University. He also writes a monthly column on science and technology policy for Nature. He is based in Washington, D.C.