It appears to many that the calculation method coded in IceCalc is not consistent with the calculation method specified in USFSA rule 3435. This has generated a lot of discussion on the internet. But interpretation of rule 3435 varies.
It is important to note, however, that the wording of rule 3435 is not the definitive factor in understanding the scoring question. USFSA rules require that IJS be used at the Championships as specified by current ISU rules governing IJS and their interpretations. If ISU rules or their interpretations change after the printing of the rulebook USFSA rules automatically follow, as is directed by the USFSA rulebook. As a result, the printed rulebook is often out of date with respect to the actual current requirements of the rules. Ultimately the governing rule for the calculation method is ISU rule 353.
In 2006 a new step (k) was added to the description of the calculation method in rule 353 which reads, "The factored results are rounded to two decimal places and added. The sum is the program component score."
Now this is rather odd, because factoring is described in the following step (l) and not before. Further, in the factoring step (l) the same text "The factored results are rounded ..." appears again. Taken together (k) and (l) are redundant. So what is going on here? Why was (k) added?
One possibility is that the true intent of step (k) was to say something like the "The trimmed mean results are rounded to two decimal places," but that is not what made it into the ISU rulebook. Alternately, maybe the intent was to was to move the sentence from (l) to (k) and they forgot to delete the sentence in (l). In any event, this change made an already complex rule more confusing.
The fundamental question, however, is this: what is the most current approved description of the calculation method, and how is it interpreted at the ISU?
Peter Krick was quoted in the Chicago Tribune saying, "The programming of the way the results are obtained is consistent, but we must make the language more precise, and that is what we will do." The implication seems to be that the method coded in the software conforms to the current ISU interpretation of rule 353, and the issue is one of drafting and documenting the rules. And yet his comment doesn't exactly say that. The essential who-what-when question is unanswered -- who approved the calculation method, what did they approve, and when did they approve it.
It would be a useful exercise if someone at the ISU would unambiguously say that the calculation method coded in the software was approved by the ISU and is the current official interpretation of rule 353, and not something that was just made up by a programmer trying to decipher the rule on his own. Were that done, it would make crystal clear that the problem is not in the software, but in poorly drafted and badly documented rules at the ISU -- a not uncommon occurrence.
Assuming that is the case, we refer the reader to Rounding, Ties and Tie Breakers, which discusses the implications of the current calculation method on the accuracy of competition results. Of all the rounding methods that the ISU could have chosen to use in the calculation method, the one currently coded in the ISU software is by far the worst, and the least fair to the athletes. Given a choice between the right way, a wrong way and a really wrong way, the ISU software currently codes the really wrong way. How sad for the athletes.
Copyright 2008 by George S. Rossano