While analyzing this year's results for input into the Year End Awards, the Chief Scorer noticed that the High Points were not calculated correctly. The Race Instruction states: "The High Point Series consists of 31 scheduled races of which a yacht must compete in 10 races to be eligible for High Point standing. One throw-out is allowed for every 6 races. Throw-outs will be determined by the Chief Scorer to achieve the highest possible score for the yacht. Each yacht is scored one point for starting, plus one point for finishing, and one point for each starting yacht it beats. Each yacht's total score is divided by the sum of total number of yachts starting in each race the scored yacht started plus one point for each scheduled race (i.e., 31). The highest score at the end of the series wins. Awards will be presented for both fleet and overall standing in the High Point Series. ".

There were two problem areas:

(1) In the denominator, there is supposed to be the number of boats that started plus the number of races scheduled (31). However, the logic in the scoring algorithm only used the number of races completed (25). This resulted in position swaps in several areas.

(2) The algorithm for determining the throwouts to achieve the best possible score was not 100% correct. While it did do a good job, there are cases where it does not maximize the score. Due to the nature of the formula [ (number of points ) / (number of boats + races scheduled) ], a single closed form solution for picking the throwouts is not possible. To complicate matters, the race instruction states "One throw-out is allowed for every 6 races." It is not required that all throwouts be used. Depending on the actual yacht's scores (different for overall and class) and number of races competed, the best score may result from using a number of throwouts less than the maximum allowed for each yacht. This appear to be "counter-intuitive" and required a "brute force" check of all possible combinations to actually determine the number and values for the throwouts.

These issues were discussed with the Race Committee 11/3/2003 to verify the understanding of the instructions. The revised scores resulted. Hopefully there are no other problems in the scoring for this year.

Please note that the throw out values are enclosed in "{}". Note that in almost all cases, a single point change in either the numerator or denominator would affect the 2nd decimal place. There are several positions in the overall standings where the difference in score is much less than a single point change. In one case (10 & 11) is it only 0.0013, in another (2 & 3) it is only 0.0025, and yet in another (7 & 8) is it only 0.0090!