Performance Order and Repertoire in Competitions

puzzleWe sometimes speculate as performers about what effect the relative order of a performance may have on the outcome of a competition, audition, or jury grade. An intriguing study published by Flores and Ginsburgh in The Statistician, available online here, sheds some light on the role performance order has on the outcome of competitions. The authors compiled data from the Queen Elisabeth Musical Competition, beginning in its inception in 1951 and spanning to 1993. This international violin and piano competition, held in Belgium every 4 years, is considered among the best in the world, counting among its winners such luminaries as Leon Fleisher and Vladimir Askenazy.

In evaluating the final rankings, the authors found a statistically significant bias against those competing early (the first day) and in favor of those competing later (the fifth day).

As an adjudicator,  juror, and committee member, I know it is sometimes challenging to compare performances over the course of a full day, let alone multiple days. It may be that early on we hold back a little in an effort to gauge the overall performance level of the field. The authors of the study point out an additional possibility: the familiarity of repertoire. In one segment of the competition they studied, contestants are required to perform a newly-commissioned work unfamiliar to both the judges and the contestants. The authors suggest that judges gradually warm to the new work through the course of the competition as they hear more and more performances of it. Flores and Ginsburgh recommend, “It may thus help the members of the jury to get used to the piece and have it performed once or twice for their own use, before the competition starts.”

Performers do not always have a choice of when they perform (as in the competition in this particular study, which utilizes a random drawing); however, if they do, this study would seem to suggest against picking an early time (all other things being equal, of course). The study would also seem to suggest that, given a choice of repertoire, more familiar repertoire may be advantageous (again, all other things being equal). Finally, judges would be advised to do their best to guard against performance order and repertoire biases.

We’re talking about just part of the puzzle, of course. There’s no replacement for playing well!