WINETASTER ON 02/04/13 WITH 8 JUDGES AND 8 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-2013 Richard E. Quandt, V. 1.65
FLIGHT 1: Number of Judges = 8 Number of Wines = 8
Identification of the Wine: The judges' overall ranking:
Wine A is Heitz Cellar 2002 tied for 6th place Wine B is Twomey Merlot 2002 ........ 8th place Wine C is Phelps Insignia 2005 ........ 4th place Wine D is BV Georges de Latour Priv.Res. 2004 ........ 5th place Wine E is Overture NV ........ 3rd place Wine F is Silver Oak Napa 2003 ........ 1st place Wine G is Opus One 2006 ........ 2nd place Wine H is Dominus Estate 2006 tied for 6th place
The Judges's Rankings
Judge Wine -> A B C D E F G H Mike 4. 8. 2. 7. 1. 3. 5. 6. Bob 5. 8. 6. 2. 7. 4. 1. 3. Frank 5. 6. 7. 8. 3. 2. 1. 4. Burt 8. 3. 5. 6. 4. 1. 2. 7. Zaki 5. 8. 6. 3. 1. 2. 4. 7. Ed 7. 8. 4. 5. 6. 3. 1. 2. Orley 6. 8. 4. 1. 3. 2. 5. 7. Dick 3. 8. 2. 5. 4. 1. 6. 7.
Table of Votes Against Wine -> A B C D E F G H
Group Ranking -> 6 8 4 5 3 1 2 6 Votes Against -> 43 57 36 37 29 18 25 43
( 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 = 0.3847
The probability that random chance could be responsible for this correlation is quite small, 0.0030. 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 Zaki 0.7381 Burt 0.6108 Orley 0.5988 Frank 0.5868 Dick 0.4762 Mike 0.4762 Ed 0.3952 Bob 0.1905
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 F is Silver Oak Napa 2002 --------------------------------------------------- 2. ........ 2nd place Wine G is Opus One 2006 3. ........ 3rd place Wine E is Overture NV 4. ........ 4th place Wine C is Phelps Insignia 200 5. ........ 5th place Wine D is BV Georges de Latour Priv.Res. 200 6. tied for 6th place Wine H is Dominus Estate 2006 7. tied for 6th place Wine A is Heitz Cellar 2002 --------------------------------------------------- 8. ........ 8th place Wine B is Twomey Merlot 2002 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 21.5417. The probability that this could happen by chance is 0.0030 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 Mike Bob Frank Mike 1.000 -0.238 0.333 Bob -0.238 1.000 0.262 Frank 0.333 0.262 1.000 Burt 0.119 -0.024 0.548 Zaki 0.571 0.214 0.381 Ed 0.119 0.762 0.524 Orley 0.405 0.310 -0.048 Dick 0.762 -0.048 0.071 Burt Zaki Ed Mike 0.119 0.571 0.119 Bob -0.024 0.214 0.762 Frank 0.548 0.381 0.524 Burt 1.000 0.310 0.262 Zaki 0.310 1.000 0.143 Ed 0.262 0.143 1.000 Orley 0.214 0.833 0.190 Dick 0.095 0.548 0.071 Orley Dick Mike 0.405 0.762 Bob 0.310 -0.048 Frank -0.048 0.071 Burt 0.214 0.095 Zaki 0.833 0.548 Ed 0.190 0.071 Orley 1.000 0.619 Dick 0.619 1.000 Pairwise correlations in descending order 0.833 Zaki and Orley Significantly positive 0.762 Bob and Ed Significantly positive 0.762 Mike and Dick Significantly positive 0.619 Orley and Dick Not significant 0.571 Mike and Zaki Not significant 0.548 Zaki and Dick Not significant 0.548 Frank and Burt Not significant 0.524 Frank and Ed Not significant 0.405 Mike and Orley Not significant 0.381 Frank and Zaki Not significant 0.333 Mike and Frank Not significant 0.310 Burt and Zaki Not significant 0.310 Bob and Orley Not significant 0.262 Burt and Ed Not significant 0.262 Bob and Frank Not significant 0.214 Bob and Zaki Not significant 0.214 Burt and Orley Not significant 0.190 Ed and Orley Not significant 0.143 Zaki and Ed Not significant 0.119 Mike and Burt Not significant 0.119 Mike and Ed Not significant 0.095 Burt and Dick Not significant 0.071 Ed and Dick Not significant 0.071 Frank and Dick Not significant -0.024 Bob and Burt Not significant -0.048 Frank and Orley Not significant -0.048 Bob and Dick Not significant -0.238 Mike and Bob Not significant
COMMENT: These wines were all terrific and similar, but one wine had a technical flaw that many of us felt indicated that it was corked. One taster felt that if we had not known what these wines were, we would have identified them as New World cabernets. In fact, three were cabernets, one was a merlot, and the rest Bordeaux blends. We all agreed that these wines will improve for a number of years. Note that in every tasting we have had in which Opus One was involved, it came in at or near the top. It is interesting that our third favorite was Overture, which is the second wine of Opus One and non-vintage, available only at the chateau. This tasting also permits us to examine, at least partially, the judges' intertemporal consistency. A tasting on February 1, 2010 had five of the wines in the present tasting. Their "votes against" and their overall ranks in the two tastings were as follows: February 1, 2010 February 4, 2013 Wine Votes Against Rank Votes Against Rank Twomey 2002 35 6 57 8 Silver Oak 2003 29 3 18 1 Opus One 2002 22 1 25 2 Phelps Insignia 2005 30 4 36 4 Dominus 2006 36 7 43 6 (tied) The votes are more tightly bunched in the February 2010 tasting, ranging from 22 to 36, while in the 2013 tasting they range from 18 to 57. But the two tastings impose essentially the same rank order among these five wines, which suggests a fairly significant temporal consistency in ranking these wines.
Return to previous page