45-1,p18

The Structure & Judging of Jugging Competitions

By Arthur Lewbel

In a poll a few years ago, a large majority of IJA'ers expressed support for having juggling competitions, though there were passionate disputes about how they should be judged. Having been involved in both writing the rules and judging IJA competitions for years, I have received reactions ranging from congratulations to abuse in response to competition rules and outcomes.

At the other extreme are the stage championships for seniors, juniors and teams. The rules are essentially to do any juggling show, and the "best" show wins, where quality is judged along a large number of vaguely defined dimensions like difficulty, originality, technique, execution and showmanship. The result is basically a

talent show, in which equally qualified judges can (and do) arrive at radically different rankings of the performers. This is true regardless of whether the competitors are ranked ordinally against each other, as was the case up until two years ago, or whether they are ranked against some more broadly defined (cardinal) standard of excellence as in the current modified Kappell system. (Those readers familiar with measure theory will recognize the mathematical significance of referring to these systems as ordinal vs cardinal).

Intrinsically, cardinal rankings are more difficult to arrive at than ordinal ones, just as it is easier to tell from the ground which of two mountains is taller than it is to determine the exact height of each.

However, the lJA's Kappell system only requires a coarse cardinal ranking into gold, silver, bronze or no medal (with, as currently implemented, almost no act being good enough for gold). So, under the old system it was necessary to tell which of two almost equally good acts was better, while under Kappell both acts can be given a silver. In the mountain analogy, if two mountains have similar heights it may be easier to say both are very high than to determine which is higher.

The difficulty in judging the stage championships is not primarily whether the ranking is ordinal or cardinal, but rather in the subjective nature and, more importantly, the number of different criteria that must be weighed together in determining the outcome. The problem can be viewed as one of "estimating" the true quality of each act. In statistical analyses, the more variables a problem has, and the greater the measurement errors in those variables, the less accurate are the resulting estimates. The only solutions are either to accept the ambiguity of the results or to reduce the dimensionality of the problem.

Accepting the ambiguity means that as long as competition performances are allowed to be any juggling show, there will a ways be legitimate disputes about who should have won, regardless of how the rules for judging are structured. Reducing the dimensionality of the problem means designing a new type of competition, more like judged Olympic sports. The problems in judging sports like gymnastics and figure skating are analogous to judging juggling, and the Olympic solution is to reduce the dimensions of the problem (i.e. the number of criteria considered) to a manageable number, and to make each of the criterion be as objectively measurable as possible.

A difficulty in implementing an Olympic style juggling competition, which could certainly exist in addition to those already in place, is that jugglers prize originality, which almost by definition is incomparable to what has been seen (and judged) before. Nevertheless, an Olympic style juggling competition could be designed to reduce the dimension of the judging problem as much as possible, thereby making performances more comparable (or "rankable") while leaving as much scope for creating new juggling tricks as possible.

There are many ways to reduce dimensions. For example, a competition could be restricted to a single prop (say, just balls). Juggling tricks could be classified roughly by type (e.g., numbers, body throws, bounce moves, intricate patterns, contact moves, transitions, etc.,.) and by difficulty. Once a group of judges had codified a number of standard moves along these lines, new tricks that appear in competitor's performances could be classified by analogy to existing moves, with extra points being awarded for originality. This is essentially what is done in rhythmic gymnastics, where competitors at the junior level do only already classified moves and tricks (which obviously simplifies judging), with the pros creating new moves within preexisting classifications.

Other ways of eliminating dimensions would include restricting each performance to a precise number of each type of move, and placing limits on costumes, music (say, none with lyrics), or even on the type of prop stand.

Anything that removes differences between acts, especially differences that are only tangentially related to the juggling itself, results in performances that can more objectively ranked, and hence reduces ambiguity about who the winner should be.

There is a fundamental tradeoff. The more we restrict performances to "look the same," the more accurately they can be judged, compared and ranked. But the more constraints we impose on performances, the less is the scope for originality. Ultimately, the current free form of competitions results in the most entertaining theatre and the most debatable judging, while at the other extreme the current numbers competitions are the most boring to watch and the most objectively measured. In addition to these extremes, I believe that many jugglers would like to see (and enter) Olympic style juggling events, each of which could be judged with some degree of objectivity while still allowing a great deal of inventiveness, entertainment value, and originality. Certainly Olympic style competitions are very popular for many other movement skills, from diving to figure skating to gymnastics.

"The Academic Juggler" is an occasional feature of Jugglers World, and is devoted to all kinds of formal analyses of juggling. Anybody who has suggestions, comments, or potential contributions for this feature is encouraged to write to me (Arthur Lewbel, Lexington, MA).