Scores of wine competitions are held annually to bestow gold, silver or bronze medals upon a select number of bottles. Only within the past few years has the reliability of the awards come under rigorous scrutiny and been found wanting. The focus of these studies has been on the decision makers. For major competitions, these are typically expert judges who may or may not have been trained and who may or may not have been screened for consistency. Since ultimately the credibility of the decisions stems from the credibility of the judges, how they are selected is critical. But once reliable judges have been empanelled, how their opinions are recorded and aggregated becomes of paramount importance. It is well-known that an outcome of any vote depends as much on the choice of method used to combine the votes as on any other factor (Saari 2001b). After examining a number of procedures for comparing wines, Amerine and Roessler (1983) concluded that “[r]anking procedures are then usually preferred” (p. 168). Ashenfelter and Quandt (1999) used rank values, which they called “Points Against,” introduced in Amerine and Roessler (1983) to reassess the famous Judgment of Paris red wine competition. This method is equivalent to the Borda Count which Hulkower (2009, 2011) emphasized is the most mathematically defensible for combining individual rankings of wines to arrive at an aggregate ranking. In addition to the unique properties summarized in the third section of this paper, the Borda Count avoids distortions introduced by summing or averaging points assigned by individual judges which can diminish the influence of tougher graders thereby violating “one judge, one vote.” The purpose of this paper is to offer a ranking procedure based on the Borda Count that can be used to award medals in a manner that most reliably and completely reflects each judge’s opinion while preserving “one judge, one vote.”
A summary of recent studies exposing problems with wine competitions is presented in the next section. The case for the Borda Count is made in the third section and a method for awarding medals based on it comprises the fourth section. The fifth section is a discussion that compares the method proposed in this paper to an alternative in the literature. Conclusions are contained in the sixth section.