Test skis were reset and prepared appropriately for the given temperature, humidity, and snow type conditions. At the testing site, each pair of skis (e.g. [S1, S2], [S3, S4], [S5, S6], [S7, S8]) was initially skied for approximately 2 km which served as a break-in period. The pairs of skis were then split-up and sorted into randomized pairs (e.g. [S1, S8], [S3, S7], [S4, S6], [S2, S5]). For each randomized pair, a “winner” and a “loser” was chosen using a “feel test”. This “feel test” involved skiing a given pair of skis for approximately 1 km and evenly rotating the skis from one foot to the other over the course of that distance in order to determine which ski “felt” the best in regards to multiple factors (including overall speed, ease of release, and general slickness).1 Following testing of the initial set of randomized pairs of skis, a new set of randomized pairs of skis was selected from the set of “winners,” and an additional new set of randomized pairs of skis was selected from the set of “losers”. Using the same methodology that led to the first sets of “winners” and “losers”, the new sets of randomized pairs of skis were “feel tested” and “winners” and “losers” were chosen. This process was repeated until an overall “winner” and “loser” were confirmed. Following the confirmation of an overall “winner” and “loser”, an order for the remaining skis was deduced from the proceeding test results and a full ranking for the entire set of skis was produced (e.g. [S4 > S7 > S1 > S3 > S6 > S2 > S5 > S8]).
1 In an effort to remove random error from each “feel test”, each randomized pair of skis was skied by a minimum of two other testers in order to achieve a group consensus.