Read Online Talent Is Overrated by Geoff Colvin - Free Book Online

rate them, the ratings track intelligence moderately well; bosses tend to think that smarter salespeople are better. But when the researchers compared intelligence with actual sales results, they found nothing. Intelligence was virtually useless in predicting how well a salesperson would perform. Whatever it is that makes a sales ace, it seems to be something other than brainpower.
    These results are surprising also because they suggest that sales supervisors are deluding themselves. You’d think they would have every incentive to know the objective performance of their subordinates and rate them on that basis, but apparently they do not. That finding has been supported in at least one other large meta-analysis. It seems our view that intelligence necessarily produces better performance is so deep that it may occasionally even blind us to reality.
    A more detailed investigation of real-world performance focused on an activity that has a lot in common with business: betting on horses. You study the facts, you estimate odds, and you decide where to put your money; it isn’t so different from management. The researchers went to a track and recruited a group of subjects. Based on their ability to forecast post-time odds, these subjects were deemed experts or nonexperts. The experts were by definition a lot better at that task, but except for that difference, the two subsets on average turned out to show no significant differences in several ways that you might expect to matter: years of experience at the track, years of formal education, occupational prestige scores, and IQ. The IQ averages and variabilities of both groups, in addition to being the same, were almost exactly the same as for the overall population. The expert forecasters were no smarter than the nonexperts or than people in general.
    Looking at the data more closely, the researchers found that knowing a particular subject’s IQ was of no use in predicting whether he was a handicapping expert. For example, one of the experts was a construction worker with an IQ of 85 (what one of the early IQ test developers classified as “dull normal”) who had been going to the track regularly for sixteen years; he picked the top horse in ten out of ten races the researchers presented and picked the top three horses in correct order five times out of ten. By contrast, one of the nonexperts was a lawyer with an IQ of 118 (“bright normal,” almost “superior”) who had been going to the track regularly for fifteen years; he picked the top horse in only three of ten races and the top three in only one of ten.
    What makes these results especially interesting is that accurately forecasting odds is highly complex. More than a dozen factors have to be considered, and they relate to one another in complicated ways. In fact, the researchers found that the expert handicappers used models that were far more complex than what the nonexperts used, so-called multiplicative models in which the values of some factors (such as track condition) altered the importance of others (such as last-race speed). In other words, what the experts were accomplishing was extremely demanding. And to repeat, IQ just didn’t seem to matter. “Low-IQ experts always used more complex models than high-IQ nonexperts,” the researchers found. Not only did handicapping expertise fail to correlate with IQ, it didn’t even correlate with performance on the arithmetic subtest of the IQ test.
    The researchers’ conclusion: Their results suggest “that whatever it is that an IQ test measures, it is not the ability to engage in cognitively complex forms of multivariate reasoning.” That last phrase is not one that most of us use very often, but it’s actually a very good description of what most of us do every day in our working lives, and what the best performers do extremely well. You just don’t have to be especially “smart,” as