Ball State University hired Charles Kuhn as an Instructor in Piano in 1966, when he was 37 years old. In 1974 the University promoted Kuhn to Assistant Professor of Music, with tenure. Kuhn sought a further promotion, to associate professor, without success. When he retired in 1994, he was still an assistant professor. This suit under the Age Discrimination in Employment Act challenges the University's decision not to promote him during the 1991-92 academic year (the only decision within the scope of the charge Kuhn filed with the EEOC).

The Promotion and Tenure Committee of the School of Music turned down Kuhn's request in December 1991, stating that his work as a teacher was not superior, his service to the University community was limited, and he had not achieved distinction in performance. Appeals within the University were unavailing. Kuhn believes that the Committee's explanation is a pretext for discrimination, and he offers as evidence the fact that he was viewed as a satisfactory member of the faculty, and that he would have been promoted to associate professor on several earlier occasions but for his lack of a Ph.D., or budgetary shortfalls that derailed many promotions. Brochures the University uses to attract students sing a paean to the members of the faculty, including Kuhn. Yet after the University deemed his experience and contributions equivalent to a Ph.D., making him eligible for promotion, it came up with a new ground to foil his advancement. That shows pretext, Kuhn insists, and entitles him to an inference that the real reason is his age. See Anderson v. Baxter Healthcare Corp., 13 F.3d 1120, 1124 (7th Cir.1994). The district court was not persuaded and granted summary judgment to the University. (The judge assumed that Kuhn had established a prima facie case; we shall do likewise.)

What is missing from Kuhn's submission is any reason to doubt the University's explanation, which is based on the distinction between competent and superior achievement. Satisfactory performance as an assistant professor does not entitle anyone to promotion. Universities prune the ranks--sometimes ruthlessly, so that only the best rise. About all Kuhn has to offer in support of the inference that the reason he fell short is age rather than performance is the fact that he was deemed worthy of promotion in earlier years but was held back by extraneous obstacles. The University says that standards rose between the late 1970s and the early 1990s; Kuhn rejoins that the only thing rising was his age. Because Kuhn as plaintiff bears the burden of persuasion, Taylor v. Canteen Corp., 69 F.3d 773, 780 (7th Cir.1995), he has to do more than show that by his own standards he should have been treated better. The ADEA is not a merit-selection program. Employers may act for many reasons, good and bad; they may err in evaluating employees' strengths; unless they act for a forbidden reason, these errors (more properly, differences in assessment) do not matter. See Russell v. Acme-Evans Co., 51 F.3d 64, 68-69 (7th Cir.1995); Gustovich v. AT & T Communications, Inc., 972 F.2d 845 (7th Cir.1992); Pollard v. Rea Magnet Wire Co., 824 F.2d 557, 560 (7th Cir.1987).

Kuhn observes that a younger person (age 56) was promoted to associate professor in 1991-92; the University replies that an older person (age 65) was promoted in 1988-89. Scattered decisions either way reveal little about the University's processes. The question is whether age tipped the balance. Gehring v. Case Corp., 43 F.3d 340 (7th Cir.1994). Would Kuhn have been promoted had he been younger, and everything else the same? Even an employer who treats age as a decided minus will not fire (or fail to promote) all old people; some will be so superior that they overcome bias. Similarly, the promotion of a handful of youngsters shows nothing of value. Suppose the University acts randomly. Then it will promote some youngsters over more qualified older persons; and it will promote some average oldsters over young stars of the department. Everyone could point to a comparison that would make the University look bad, but the exercise would not show that age played a role. A plaintiff who wants a court to infer discrimination from the employer's treatment of comparable cases has to analyze a goodly sample. One is an anecdote, and several cases are several anecdotes. Judges do not find discrimination on such a thin basis. What a plaintiff in Kuhn's position has to do is subject all of the employer's decisions to statistical analysis to find out whether age makes a difference. Our opinions emphasize the need to get beyond a few comparison cases, e.g., Troupe v. May Department Stores Co., 20 F.3d 734 (7th Cir.1994); Bush v. Commonwealth Edison Co., 990 F.2d 928, 931-32 (7th Cir.1993), and we cannot stress this point enough.

Consider the nature of the parties' arguments. Ball State University says that only superior teachers and scholars are promoted. If so, then many assistant professors should fail to achieve promotion, and others should leave in anticipation of adverse decisions. (We discount the possibility that hiring decisions are so perspicacious that Ball State employs only those who meet the promotion criteria.) If the University is right, adding age as an independent variable in a statistical analysis of initial promotion decisions would not increase its explanatory power. ("Initial" decisions is an important qualifier. Candidates turned down once are likely to be turned down again. Every time they apply they will be older, but successive rejections cannot be attributed to advancing age.) As Kuhn depicts things, however, almost everyone who meets the standards for the post of assistant professor also meets the standards for associate professor. On his view, the promotion rate is high, and the feature distinguishing those promoted from those held back, age. Only knowledge of the promotion rate by age would enable a court to distinguish between Kuhn's hypothesis and that of the University, and therefore to draw the appropriate inference.

Kuhn did not attempt to analyze the University's other promotion decisions. He was content to identify the promotion of a single younger person in 1991-92. The University was scarcely more inclined to come up with data. It produced a list of persons promoted in the Music Department from 1988 to 1993. Such a list is next to worthless, because without knowing how many people, of what age, were not promoted, a court cannot decide which side's proposed inference is superior. Limiting the analysis to the Music Department also is inappropriate; the final decision came from the University-wide Promotion and Tenure Committee. If age discrimination was at work, its effects should be apparent throughout the University. (Departments, with small numbers of recommendations, could show variance either way without revealing anything about the University's criteria.) As it happens, Kuhn did not even try to mine the list of promotions for information; he professed ignorance about where the achievements of those who were promoted ranked in relation to his own.

So we are in the dark. For all we know, Ball State University indeed discriminates against older faculty seeking promotion. But nothing in this record supports such an inference. Once Ball State explained its decision not to promote him, Kuhn had to come forward with evidence to suggest, not that the University was mistaken in failing to promote him, but that it was lying in order to cover up the true reason, his age. Kuhn's evidence does not move far toward suggesting mistake; it does nothing to suggest that the explanation he received was disingenuous. Statistical tools were available--and a faculty member at a university, unlike the ordinary plaintiff, has only to amble over to his colleagues in the statistics department to explore the possibility. Kuhn, however, produced one anecdote and rested. The district court properly held that this dooms his claim.

AFFIRMED.