Abstract: Instead of ranking law schools through statistical aggregations of expert judgments, or by combining a list of heterogeneous factors, it would be possible to rely on a market test, simply by examining student choices. This tournament-type approach would have the large advantage of relying on the widely dispersed information that students actually have; it would also reduce reliance on factors that can be manipulated (and whose manipulation does no good other than to increase rankings). On the other hand, a market test has several problems as a measure of law school quality, partly because cognitive biases and social influences may lead some or many students to make bad choices and thus to participate in the production of inaccurate rankings.