The 72nd Academy Awards, held in March 2000, supplied a striking comparison between two ways of forecasting the same results: a newspaper’s survey of Academy voters and the Hollywood Stock Exchange’s prediction market. The exchange got all six winners in the major Oscar categories right, outperforming a reporting effort that had unusually direct access to the voters themselves. The episode illustrates the book’s broader argument that an aggregated judgment can be more informative than any single expert or source, while remaining a bounded example rather than proof that crowds are invariably accurate.

Competing Forecasts

The Hollywood Stock Exchange (HSX) let participants wager on Oscar outcomes as well as box-office performance. Its 2000 Academy Awards forecast was tested against an intensive reporting project by twelve reporters from The Wall Street Journal, who canvassed Academy members to learn how they had voted. The Academy objected publicly, accusing the newspaper of trying to disclose the results before Oscar night, and urged its members not to speak to reporters.

The newspaper promised anonymity, and 356 members—about six percent of the Academy—disclosed how they had filled out their ballots. On the Friday before the ceremony, the Journal published predictions for the six major categories: Best Picture, Best Director, Best Actor, Best Actress, Best Supporting Actor, and Best Supporting Actress. Its reporting therefore drew on a substantial set of direct responses from voters, even though those respondents made up only a small share of the membership.

Results and Comparison

When the envelopes were opened, the Journal had correctly forecast five of the six winners. The HSX got all six right. The contrast is notable because the newspaper’s method sought information directly from the people casting ballots, whereas the exchange aggregated participants’ wagers into a market forecast. In this instance, the indirect, market-based estimate proved more accurate than the survey of voters.

The exchange also performed well beyond the awards-night comparison: in 2002, it picked 35 of the eventual 40 Oscar nominees. That result extended the case for HSX as a forecasting tool, though the book does not present either outcome as evidence that the market will always succeed.

Significance and Limits

The awards forecast was one example of the wider performance of the HSX, which also traded predictions about box-office returns and opening-weekend results. Its movie forecasts were not as accurate overall as those of the Iowa Electronic Markets for elections, but a comparison by Harvard Business School professor Anita Elberse found that the HSX closing price on the night before a film opened was the best available forecast of its weekend box office. Its owner, Cantor Index Holdings, was consequently marketing the data to Hollywood studios.

The HSX’s wagers used play money rather than real financial stakes. The book notes that rewards can help focus people on a decision, but reports researcher David Pennock’s finding that status and reputation could also motivate active traders to invest serious time and effort. The 2000 awards result thus belongs to a broader account of how prediction markets can combine dispersed judgments: it demonstrates an impressive success, not a guarantee, and its force lies in the specific comparison between six predicted winners and six actual ones.