WINETASTER ON 05/02/05 WITH 7 JUDGES AND 7 WINES BASED ON RANKS, IDENT=Y Copyright (c) 1995-2005 Richard E. Quandt, V. 1.65

A Tasting of Spanish Wines from Priorat and Toro.
FLIGHT 1: Number of Judges = 7 Number of Wines = 7
Identification of the Wine: The judges' overall ranking:
Wine A is Lo Givot 2001 ........ 2nd place Wine B is Los Mogador 2001 tied for 6th place Wine C is Fra Fulco 2001 ........ 3rd place Wine D is Numanthia 2001 ........ 1st place Wine E is Embruix 2001 ........ 5th place Wine F is Vall Llach 2001 ........ 4th place Wine G is Mas Doix 2001 tied for 6th place
The Judges's Rankings
Judge Wine -> A B C D E F G Bob 1. 4. 7. 2. 6. 5. 3. John 3. 7. 4. 2. 6. 5. 1. Burt 3. 4. 2. 1. 5. 6. 7. Mike 6. 5. 2. 3. 1. 4. 7. Ed 4. 7. 2. 3. 6. 1. 5. Richard 3. 5. 2. 7. 1. 4. 6. Dick 2. 1. 5. 3. 7. 6. 4.
Table of Votes Against Wine -> A B C D E F G
Group Ranking -> 2 6 3 1 5 4 6 Votes Against -> 22 33 24 21 32 31 33
( 7 is the best possible, 49 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 = 0.1283

The probability that random chance could be responsible for this correlation is rather large, 0.4951. 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 with price shows the extent to which each judges ranking are correlated with the price of the wine. A higher correlation with price indicates that the judge in question preferred the more expensive wine.
Correlation Between the Ranks of Each Person With the Average Ranking of Others
Name of Person Correlation R Correlation Price Burt 0.5714 0.0360 John 0.1684 0.3784 Ed 0.0000 -0.0541 Bob -0.1261 0.2162 Dick -0.1261 -0.3964 Mike -0.3214 0.0901 Richard -0.3964 -0.3063

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. None of the wines is significantly better or worse than another.
1. ........ 1st place Wine D is Numanthia 2001 2. ........ 2nd place Wine A is Lo Givot 2001 3. ........ 3rd place Wine C is Fra Fulco 2001 4. ........ 4th place Wine F is Vall Llach 2001 5. ........ 5th place Wine E is Embruix 2001 6. tied for 6th place Wine B is Los Mogador 2001 7. tied for 6th place Wine G is Mas Doix 2001 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 5.3878. The probability that this could happen by chance is 0.4951
We now test whether the group ranking of wines is correlated with the prices of the wines. The rank correlation between them is -0.0545. At the 10% level of significance this would have to exceed the critical value of 0.5710 to be significant.
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 0.79 for significance at the 0.05 level and must exceed 0.71 for significance at the 0.1 level Bob John Burt Bob 1.000 0.536 0.143 John 0.536 1.000 0.071 Burt 0.143 0.071 1.000 Mike -0.679 -0.429 0.393 Ed -0.143 0.321 0.214 Richard -0.607 -0.500 -0.036 Dick 0.679 0.107 0.357 Mike Ed Richard Bob -0.679 -0.143 -0.607 John -0.429 0.321 -0.500 Burt 0.393 0.214 -0.036 Mike 1.000 0.179 0.536 Ed 0.179 1.000 0.000 Richard 0.536 0.000 1.000 Dick -0.607 -0.357 -0.536 Dick Bob 0.679 John 0.107 Burt 0.357 Mike -0.607 Ed -0.357 Richard -0.536 Dick 1.000 Pairwise correlations in descending order 0.679 Bob and Dick Not significant 0.536 Mike and Richard Not significant 0.536 Bob and John Not significant 0.393 Burt and Mike Not significant 0.357 Burt and Dick Not significant 0.321 John and Ed Not significant 0.214 Burt and Ed Not significant 0.179 Mike and Ed Not significant 0.143 Bob and Burt Not significant 0.107 John and Dick Not significant 0.071 John and Burt Not significant 0.000 Ed and Richard Not significant -0.036 Burt and Richard Not significant -0.143 Bob and Ed Not significant -0.357 Ed and Dick Not significant -0.429 John and Mike Not significant -0.500 John and Richard Not significant -0.536 Richard and Dick Not significant -0.607 Bob and Richard Not significant -0.607 Mike and Dick Not significant -0.679 Bob and Mike Not significant

COMMENT: We first wish to note that our colleague Frank was absent for this tasting, as was Orley. Some of the wines were much more tannic than others and the differences among the tasters is partly explained by their tolerance for tannic wines. The wines were all extremely good and drinkable and had delicious bouquets. Basically, B, A, and G were not very tannic, whereas C, E, F were rather more so. One taster noted that the rankings were essentially random and had nothing to do with the tannic nature of the wines. So, the upshot is that not only did we disagree about the wines, we also disagree about what makes us disagree about the wines. Another taster argued that the bouquets were more variable than other aspects of the wines. All in all, all of these wines are remarkably approachable. Note also that our ranking was substantially negatively correlated with the ranking of Parker. The overall correlation among the tasters is one of the lowest in recent memory.
Return to previous page