Bettors and Pundits: Never Wrong, Just Unlucky

By Shankar Vedantam
Monday, March 26, 2007

The NCAA men's college basketball championship game was on the line. People in office pools around the country were holding their breath. Louisville was down by four points with a few minutes left on the clock. A UCLA player stole a pass and raced down the court where, after being bumped by a Louisville player, he missed an easy basket. UCLA fans cried foul. Louisville fans saw only incidental contact -- and the referees agreed. Moments later, Louisville stormed back and won the game.

The moral of this story goes well beyond basketball and sports, although anyone who is part of an NCAA office pool could take away a lesson from it. After the game, a psychologist conducted an experiment. He asked UCLA and Louisville fans whether they thought their team would win a rematch of the 1980 final. Only 23 percent of UCLA fans said that UCLA would win.

The psychologist then asked another group of fans of the two teams the same question, but slipped in a reference to the controversial foul call: "By the way, what did you think of the breakaway layup that Kiki Vandeweghe missed near the end? Do you think he was fouled like he said he was? When you consider that if he had made that basket UCLA would have led by six with only about two minutes to play, there was a lot of money riding on that one basket."

Surprisingly, fully 70 percent of UCLA fans now said they thought their team would win a rematch. Reminding the fans about the controversial play, in other words, prompted large numbers of UCLA fans to conclude their side was really the better team and, minus a fluke event, would win.

There was little difference among Louisville fans whether or not they were reminded of the play. Nearly all said their team would win a rematch -- in other words, foul call or no foul call, they believed that Louisville was simply the better team.

The point here is not whether the foul call mattered. In real life, you don't get to wind the clock back and replay a different scenario. The point the psychologist was trying to make is that winners discount information about lucky breaks and chalk up their right calls to superior judgment, whereas losers tend to emphasize the role of bad luck -- rather than bad judgment -- when their predictions go wrong.

Thomas Gilovich, the Cornell University psychologist who conducted the experiment, said this is why a lot of water-cooler conversations among NCAA office-pool participants feature people explaining to others why their predictions went wrong. People don't talk very much about predictions that went right, Gilovich said, because they automatically chalk up those results to brilliant insight. Wrong calls, however, are invariably seen to be caused by fluke events -- which is why they need explaining.

The psychological truth here goes beyond sports fans and NCAA office pools. Our ability to blame fluke events for our errors in judgment affects many aspects of our lives. This phenomenon is perhaps most vividly on display when policymakers and pundits remain endlessly confident about their pronouncements even after they have been proved breathtakingly wrong on major questions such as the outcome of the war in the Iraq. (You know who you are.) "People have explanations for the things that went badly so they can be set aside -- 'I made the right prediction but for that fumble in the fourth quarter' -- so they walk out with their confidence intact," Gilovich said. Experts say, "Okay, I was wrong about the Soviet Union, but everything I said makes perfect sense if it wasn't for this one unpredictable factor."

"Experts are very wrong, but they don't think they are wrong because a lot of mistakes are coded as near misses," the psychologist added. "So their intuitive hit rate is higher than the actual hit rate."

Basketball turns out to be a surprisingly useful subject for studying other elements of human behavior and decision making. When psychologists once asked sports fans to visualize how a particular team might win a game, the volunteers became far more likely to later believe that the team would win -- and to bet money on it. Essentially, psychologist Bryan Gibson of Central Michigan University said, focusing on some part of a conundrum makes it much more difficult for people to keep in mind how much they do not know about all the other variables involved.

"Focusing on one potential outcome leads to the overestimation of the probability that this focal outcome will occur," Gibson and his colleagues concluded. "This inflated probability estimate . . . leads to a greater willingness to gamble."

Gibson has also found that gambling is one domain in which it may be wiser to have a pessimistic personality rather than an optimistic personality. Gamblers tend to be optimists in that they inherently believe good things are likely to happen to them. When a pessimist wins a bet, he is likely to walk away with his winnings because he can't believe that his lucky streak will continue.

By contrast, Gibson found in one series of experiments, optimists were far more likely to throw good money after bad because they believed that, sooner or later, things would break their way.

"As losses built up, pessimists were likely to say, 'That is enough,' whereas optimists were more likely to continue despite the losses," Gibson said.

© 2007 The Washington Post Company