WINETASTER ON 04/29/01 WITH 3 JUDGES AND 3 WINES BASED ON RANKS, IDENT=Y Copyright (c) 1995-2000 Richard E. Quandt
FLIGHT 1: Number of Judges = 3 Number of Wines = 3
Identification of the Wine: The judges' overall ranking:
Wine A is Tokaji Aszu Citadella 1993 3 puttonyos ........ 1st place Wine B is Tokaji Aszu Citadella 1993 4 puttonyos ........ 3rd place Wine C is Tokaji Aszu Citadella 1993 5 puttonyos ........ 2nd place
The Judges's Rankings
Judge Wine -> A B C Gina 1. 3. 2. Orley 1. 3. 2. Dick 1. 2. 3.
Table of Votes Against Wine -> A B C
Group Ranking -> 1 3 2 Votes Against -> 3 8 7
( 3 is the best possible, 9 is the worst)
Here is a measure of the correlation in the preferences of the judges which ranges between 1.0 (perfect correlation) and 0.0 (no correlation):
W = 0.7778
The probability that random chance could be responsible for this correlation is quite small, 0.0970. Most analysts would say that unless this probability is less than 0.1, the judges' preferences are not strongly related. We now analyze how each taster's preferences are correlated with the group preference. A correlation of 1.0 means that the taster's preferences are a perfect predictor of the group's preferences. A 0.0 means no correlation, while a -1.0 means that the taster has the reverse ranking of the group. This is measured by the correlation R.
The correlation I measures the degree to which the identification of each judge is correlated with the truth. Here a 1.0 means that the judge identified the wines perfectly, and a 0 means that he identified none of them.
Correlation Between the Ranks of Each Person With the Average Ranking of Others
Name of Person Correlation R Correlation I Gina 0.8660 0.0000 Orley 0.8660 0.0000 Dick 0.5000 0.3333
Next, we show the correlation among the wine identifications of the judges, which also ranges between 1.0 and 0.0:
C = 0.5556
The probability that random chance could be responsible for this correlation is quite small: < 5 %. Most people would say that unless this probability is less than 0.1, the judges' identifications are not highly related.
The wines were preferred by the judges in the following order. When the preferences of the judges are strong enough to permit meaningful differentiation among the wines, they are separated by -------------------- and are judged to be significantly different.
1. ........ 1st place Wine A is Tokaji Aszu Citadella 1993 3 putto --------------------------------------------------- 2. ........ 2nd place Wine C is Tokaji Aszu Citadella 1993 5 putto --------------------------------------------------- 3. ........ 3rd place Wine B is Tokaji Aszu Citadella 1993 4 putto We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 4.6667. The probability that this could happen by chance is 0.0970 We now undertake a more detailed examination of the pair-wise rank correla- tions that exist between pairs of judges. First, we present a table in which you can find the correlation for any pair of judges, by finding one of the names in the left hand margin and the other name on top of a column. A second table arranges these correlations in descending order and marks which is significantly positive significantly negative, or not significant. This may allow you to find clusters of judges whose rankings were particularly similar or particularly dissimilar. Pairwise Rank Correlations Correlations must exceed in absolute value 1.00 for significance at the 0.05 level and must exceed 1.00 for significance at the 0.1 level Gina Orley Dick Gina 1.000 1.000 0.500 Orley 1.000 1.000 0.500 Dick 0.500 0.500 1.000 Pairwise correlations in descending order 1.000 Gina and Orley Significantly positive 0.500 Gina and Dick Not significant 0.500 Orley and Dick Not significant
The goal of this tasting was to sample three Tokaji Aszu wines at various "puttonyos" levels. In principle, the sweeter wines have a larger puttony level and have a longer life in the bottle. The three wines bore the identical label and vintage and differed only in the fraction of the late harvest wine added to the blend. Our identification of the wines could not have been much worse---we assumed that the darker, more caramelized wines had higher puttonyos numbers, when the precise opposite was the case. Unless there was some mistake in the labeling at the winery (and at first, we were strongly inclined to believe this!), which is unlikely, we simply tasted the wines as an early stage of development. If this is the case, then it may well be that the sensible consumer decision is to drink the wines with the lower puttonyos number early in their lives, and to cellar the wines with the higher puttnyos numbers. No doubt this rule, which we seem to have rediscovered, was known to Peter the Great, who was a great fan of these wines!
Before we even started to taste the wines, it was obvious that there were
significant color differences; A was darkest, B was lighter, and C was the
lightest, although the color difference between B and C was small. We
(erroneously) believed that the darker the color, the higher the puttonyos number would be. In fact, A was most caramelized, B next, and C least,
which confirmed this impression. In fact C had no caramel flavor at all,
and as time passed, it turned out to be the sweetest of the three wines. While
the degree of concordance among the judges was high, none of them did well
in identifying the particular wines.
The wines were all produced by the Bodrog Varhegy Kft and were all quite