William A. Sethares

William A. Sethares appears in the book through a study of coordination: with Ann M. Bell, he modeled prospective patrons of El Farol as people who decided whether to return according to whether they had recently enjoyed themselves there. Their comparatively simple rule produced a collectively effective, though imperfect, pattern of attendance and illustrated the value of independent judgment.

The El Farol Experiment

A few years after Brian Arthur formulated the El Farol problem, Bell and Sethares approached it differently. Arthur had assumed that would-be patrons used diverse strategies to anticipate the crowd. Bell and Sethares instead gave every agent the same basic rule: go if recent visits had been enjoyable; stay away if they had been disappointing.

These agents had little information about the crowd. They did not try to infer what other patrons might be thinking, and they did not know how many people had attended on nights when they themselves stayed home. Their decisions rested on the direct, limited evidence of their own experience at the bar.

A Stable but Unequal Outcome

After enough time for patrons to learn from their visits, attendance settled just below 60 percent of the bar’s capacity. This was slightly less successful than the ideal outcome a central planner could have achieved, but it was a stable, collectively intelligent response to the coordination problem. People who focused on their own experience, rather than forecasting everyone else’s behavior, helped the group reach a workable result.

The stability came with a tradeoff. The group divided into frequent regulars and people who went only rarely, making El Farol resemble the television bar Cheers. Assuming that everyone gained equal pleasure from a visit, this arrangement was satisfactory in utilitarian terms: more than half the group could attend nearly every week without usually encountering a crowded bar. But it was not ideal for those who seldom went and generally had a bad time when they did.

Significance

The experiment gives Sethares a specific place in the book’s argument about independent judgment. Bell and Sethares’s patrons were less sophisticated than Arthur’s imagined strategists, yet their reliance on personal experience produced an attendance level close to the desired capacity. The result suggests that people need not explicitly coordinate or correctly predict one another to produce a broadly useful group outcome. It also shows the limit of that success: a stable collective pattern can solve one problem while distributing its benefits unevenly.