WINETASTER ON 01/03/11 WITH 7 JUDGES AND 8 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-2011 Richard E. Quandt, V. 1.65 A Tasting of Bonnes-Mares, Comparing four vintages of Comte de Vogüé with the Same Vintages of Other Wine-makers


FLIGHT 1: Number of Judges = 7 Number of Wines = 8
Identification of the Wine: The judges' overall ranking:
Wine A is Comte de Vogüé 1996 ........ 6th place Wine B is Moillard 1990 ........ 8th place Wine C is Jadot 1996 ........ 5th place Wine D is Arlaud 2001 ........ 2nd place Wine E is Comte de Vogüé 2002 tied for 3rd place Wine F is Comte de Vogüé 1990 ........ 7th place Wine G is Moine-Hudelot 2002 ........ 1st place Wine H is Comte de Vogüé 2001 tied for 3rd place
The Judges's Rankings
Judge Wine -> A B C D E F G H Alexa 3. 7. 5. 6. 4. 8. 2. 1. Mike 2. 3. 4. 1. 8. 7. 6. 5. Burt 5. 4. 8. 1. 2. 3. 6. 7. Bob 1. 7. 6. 2. 8. 3. 4. 5. Zachy 7. 8. 2. 6. 3. 5. 1. 4. John 6. 7. 4. 5. 1. 8. 3. 2. Dick 8. 6. 1. 7. 3. 2. 4. 5.
Table of Votes Against Wine -> A B C D E F G H
Group Ranking -> 6 8 5 2 3 7 1 3 Votes Against -> 32 42 30 28 29 36 26 29
( 7 is the best possible, 56 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 = 0.0914

The probability that random chance could be responsible for this correlation is rather large, 0.7236. 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.
Correlation Between the Ranks of Each Person With the Average Ranking of Others
Name of Person Correlation R John 0.2303 Zachy 0.1464 Alexa 0.1446 Bob -0.3473 Dick -0.4431 Mike -0.5714 Burt -0.6261

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 G is Moine-Hudelot 2002 2. ........ 2nd place Wine D is Arlaud 2001 3. tied for 3rd place Wine E is Comte de Vogüé 2002 4. tied for 3rd place Wine H is Comte de Vogüé 2001 5. ........ 5th place Wine C is Jadot 1996 6. ........ 6th place Wine A is Comte de Vogüé 1996 7. ........ 7th place Wine F is Comte de Vogüé 1990 --------------------------------------------------- 8. ........ 8th place Wine B is Moillard 1990 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 4.4762. The probability that this could happen by chance is 0.7236 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 0.74 for significance at the 0.05 level and must exceed 0.64 for significance at the 0.1 level Alexa Mike Burt Alexa 1.000 -0.095 -0.524 Mike -0.095 1.000 0.024 Burt -0.524 0.024 1.000 Bob 0.024 0.500 0.119 Zachy 0.452 -0.595 -0.429 John 0.738 -0.381 -0.214 Dick -0.190 -0.714 -0.286 Bob Zachy John Alexa 0.024 0.452 0.738 Mike 0.500 -0.595 -0.381 Burt 0.119 -0.429 -0.214 Bob 1.000 -0.286 -0.452 Zachy -0.286 1.000 0.667 John -0.452 0.667 1.000 Dick -0.500 0.690 0.190 Dick Alexa -0.190 Mike -0.714 Burt -0.286 Bob -0.500 Zachy 0.690 John 0.190 Dick 1.000 Pairwise correlations in descending order 0.738 Alexa and John Significantly positive 0.690 Zachy and Dick Significantly positive 0.667 Zachy and John Significantly positive 0.500 Mike and Bob Not significant 0.452 Alexa and Zachy Not significant 0.190 John and Dick Not significant 0.119 Burt and Bob Not significant 0.024 Alexa and Bob Not significant 0.024 Mike and Burt Not significant -0.095 Alexa and Mike Not significant -0.190 Alexa and Dick Not significant -0.214 Burt and John Not significant -0.286 Burt and Dick Not significant -0.286 Bob and Zachy Not significant -0.381 Mike and John Not significant -0.429 Burt and Zachy Not significant -0.452 Bob and John Not significant -0.500 Bob and Dick Not significant -0.524 Alexa and Burt Not significant -0.595 Mike and Zachy Not significant -0.714 Mike and Dick Significantly negative



COMMENT:
The wines were all delicious and none of them had any fault. This in part explains why the overall correlation was so low (0.0914), although one wine turned out to be consistently less liked than the others (Moillard 1990). The sum of the ranks over the Comte de Vogüé wines was 126, exactly the same as for the other wines; hence nothing can be said about these two groupings relative to each other. However, the Comte de Vogüé wines were quite a bit more expensive than the others, and hence these could not be said to represent good value. Interestingly, however, the younger wines were somewhat better liked than the older ones. Computing the R-statistic over the two younger and two older vintages (See Richard E. Quandt, "A Note on a Test for the Sum of Ranksums,", Journal of Wine Economics, 2/1 (2007), 98-102) yields R= 0.8, which just misses being significant at the 0.05 level of significance (the critical value being 0.78). After the rankings were obtained, we opened a ninth bottle, a 1983 Comte de Vogüé, which by consensus was declared to be the best of the lot, although no completely objective comparison could be made with the other wines, since there was not enough left of the others for a detailed comparison.
Return to previous page