WINETASTER ON 02/04/02 WITH 8 JUDGES AND 8 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-2002 Richard E. Quandt
FLIGHT 1: Number of Judges = 8 Number of Wines = 8
Identification of the Wine: The judges' overall ranking:
Wine A is Ch. Beychevelle 1970 ........ 6th place Wine B is Clos Fourtet 1970 ........ 8th place Wine C is Ch. La Lagune 1970 ........ 3rd place Wine D is Ch. Brane Cantenac 1970 ........ 5th place Wine E is Ch. Haut Brion 1970 ........ 7th place Wine F is Ch. La Lagune 1970 ........ 4th place Wine G is Ch. Leoville Poyferre 1970 ........ 2nd place Wine H is Ch. Mouton Rothschild 1970 ........ 1st place
The Judges's Rankings
Judge Wine -> A B C D E F G H Burt 3. 6. 4. 1. 8. 5. 7. 2. Ed 3. 5. 4. 8. 6. 7. 2. 1. John 2. 8. 6. 7. 5. 4. 3. 1. Frank 7. 8. 3. 1. 6. 2. 5. 4. Dick 3. 7. 1. 8. 4. 2. 6. 5. Bob 7. 6. 5. 8. 4. 2. 1. 3. Orley 8. 5. 6. 2. 1. 7. 3. 4. Norton 7. 6. 4. 2. 8. 5. 1. 3.
Table of Votes Against Wine -> A B C D E F G H
Group Ranking -> 6 8 3 5 7 4 2 1 Votes Against -> 40 51 33 37 42 34 28 23
( 8 is the best possible, 64 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 = 0.1949
The probability that random chance could be responsible for this correlation is rather large, 0.1423. 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 0.4910 Norton 0.4671 Bob 0.4286 Ed 0.3593 Frank 0.3234 Burt -0.0482 Dick -0.1317 Orley -0.1796
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 H is Ch. Mouton Rothschild --------------------------------------------------- 2. ........ 2nd place Wine G is Ch. Leoville Poyferre 3. ........ 3rd place Wine C is Ch. La Lagune 1970 4. ........ 4th place Wine F is Ch. La Lagune 1970 5. ........ 5th place Wine D is Ch. Brane Cantenac 6. ........ 6th place Wine A is Ch. Beychevelle 1970 7. ........ 7th place Wine E is Ch. Haur Brion 1970 --------------------------------------------------- 8. ........ 8th place Wine B is Clos Fourtet 1970 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 10.9167. The probability that this could happen by chance is 0.1423 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 Burt Ed John Burt 1.000 0.000 0.143 Ed 0.000 1.000 0.690 John 0.143 0.690 1.000 Frank 0.500 -0.405 -0.048 Dick -0.119 0.119 0.310 Bob -0.524 0.381 0.476 Orley -0.238 -0.190 -0.238 Norton 0.357 0.214 0.095 Frank Dick Bob Burt 0.500 -0.119 -0.524 Ed -0.405 0.119 0.381 John -0.048 0.310 0.476 Frank 1.000 0.095 0.071 Dick 0.095 1.000 0.262 Bob 0.071 0.262 1.000 Orley 0.119 -0.595 0.071 Norton 0.571 -0.381 0.262 Orley Norton Burt -0.238 0.357 Ed -0.190 0.214 John -0.238 0.095 Frank 0.119 0.571 Dick -0.595 -0.381 Bob 0.071 0.262 Orley 1.000 0.238 Norton 0.238 1.000 Pairwise correlations in descending order 0.690 Ed and John Significantly positive 0.571 Frank and Norton Not significant 0.500 Burt and Frank Not significant 0.476 John and Bob Not significant 0.381 Ed and Bob Not significant 0.357 Burt and Norton Not significant 0.310 John and Dick Not significant 0.262 Bob and Norton Not significant 0.262 Dick and Bob Not significant 0.238 Orley and Norton Not significant 0.214 Ed and Norton Not significant 0.143 Burt and John Not significant 0.119 Frank and Orley Not significant 0.119 Ed and Dick Not significant 0.095 Frank and Dick Not significant 0.095 John and Norton Not significant 0.071 Frank and Bob Not significant 0.071 Bob and Orley Not significant 0.000 Burt and Ed Not significant -0.048 John and Frank Not significant -0.119 Burt and Dick Not significant -0.190 Ed and Orley Not significant -0.238 John and Orley Not significant -0.238 Burt and Orley Not significant -0.381 Dick and Norton Not significant -0.405 Ed and Frank Not significant -0.524 Burt and Bob Not significant -0.595 Dick and Orley Not significant
COMMENT: We have been tasting these wines for 12 years. We all agree that the wines are lighter and have less fruit than in the past. However, none is over hill. The winetasting was unusual, because two of the wines were identical--- taken from the same magnum, so there was no possible bottle variability. Of the seven people who did not know which were the identical wines, two people were able to tell which were the same wines (which is a little better than what you would expect by chance). However, in the tasters' rankings, the gap in the ranking between the identical wines was only one rank unit for four of the seven tasters and no more than three rank units for anyone (and this is also slightly better than chance). One of the oddities is that wine D (the Brane Cantenac) had a bimodal distribution of ranks, with several people ranking it first and several ranking it last.
Return to previous page