geoffrey_miller's picture
Evolutionary psychologist, NYU Stern Business School and University of New Mexico; author of The Mating Mind and Spent

The Internet changes every aspect of thinking for the often-online human: perception, categorization, attention, memory, spatial navigation, language, imagination, creativity, problem-solving, Theory of Mind, judgment, and decision-making. These are the key research areas in cognitive psychology, and constitute most of what the human brain does. BBC News and The Economist Website extend my perception, becoming my sixth sense for world events. Gmail structures my attention through my responses to incoming messages: delete, respond, or star for response later? Wikipedia is my extended memory. An online calendar changes how I plan my life. Google Maps change how I navigate through my city and world. FaceBook expands my Theory of Mind — better understanding the beliefs and desires of others.

But for me, the most revolutionary change is in my judgment and decision-making — the ways I evaluate and choose among good or bad options. I've learned that I can offload much of my judgment on to the large samples of peer ratings available on the Internet. These, in aggregate, are almost always more accurate than my individual judgment. To decide which Blu-ray disks to put in my Netflix cue, I look at the average movie ratings on Netflix, IMDB, and Metacritic. These reflect successively higher levels of expertise among the raters — movie renters on Netflix, film enthusiasts on IMDB, and film critics on Metacritic. Any film with high ratings across all three sites is almost always exciting, beautiful, and thoughtful.

My fallible, quirky, moody judgments are hugely enhanced by checking average peer ratings: book and music ratings on Amazon, used car ratings on Edmunds, foreign hotel ratings on Tripadvisor, and citations to scientific papers on Google scholar. We can finally harness the Law of Large Numbers to improve our decision-making: the larger the sample of peer ratings, the more accurate the average. As ratings accumulate, margins of error shrink, confidence intervals get tighter, and estimates improve. Ordinary consumers have access to better product-rating data than market researchers could hope to collect.

Online peer ratings empower us to be evidence-based about almost all of our decisions. For most goods and services, and indeed most domains of life, they offer the consumer a kind of informal meta-analysis — an aggregation of data across all the analyses already performed by other like-minded consumers. Judgment becomes socially distributed and statistical rather than individual and anecdotal.

Rational-choice economists might argue that sales figures are a better indication than online ratings of real consumer preferences, insofar as people vote with their dollars to reveal their preferences. This ignores the problem of buyer's remorse: consumers buy many things that they find disappointing. Their post-purchase product ratings mean much more than their pre-purchase judgments. Consumer Reports data on car owner satisfaction ('Would you buy your car again?') are much more informative than new-car sales figures. Metacritic ratings of the Twilight movies are more informative about quality than first-weekend box office sales. Informed peer ratings are much more useful guides to sensible consumer choices than popularity-counts, sales volumes, market share, or brand salience.

You might think that post-purchase ratings would be biased by rationalization — I bought product X, so it must be good, or I'd look like a fool. No doubt that happens when we talk with friends and neighbors, but the anonymity of most online ratings reduces the embarrassment effect of admitting one's poor judgments and wrong decisions.

Of course, peer ratings of any product, like votes for a politician, can be biased by stupidity, ignorance, fashion cycles, mob effects, lobbying, marketing, and vested interests. The average online consumer's IQ is only a little above 100 now, and their average education is just a couple of years of college. Runaway popularity can be mistaken for lasting quality. Clever ads, celebrity endorsements, and brand reputations can bias the judgment of even the most independent-minded consumers. Rating sites can be gamed and manipulated by retailers. Nonetheless, online peer ratings remain more useful than any other consumer-empowerment movement in the last century.

To use peer ratings effectively, we have to let go of our intellectual and aesthetic pretensions. We have to recognize that some of our consumer judgments served mainly as conspicuous displays of our own intelligence, openness, taste, or wealth, and are not really the best way to choose the best option. We have to learn some humility. My best recent movie-viewing experiences have all come from valuing the Metacritic ratings over my own assumptions, prejudices, and pre-judgments. In the process, I've learned a new-found respect for the collective wisdom of our species. This recognition that my own thinking is not so different from, or better than, everyone else's, is one of the Internet's great moral lessons. Online peer ratings reinforce egalitarianism, mutual respect, and social capital. Against the hucksterism of marketing and lobbying, they knit humanity together into collective decision-making systems of formidable power and intelligence.