WINETASTER ON 11/08/05 WITH 9 JUDGES AND 10 WINES BASED ON RANKS Copyright (c) 1995-2005 Richard E. Quandt, V. 1.65
A Tasting of Opus I and Bordeaux First Growths Matched by Year
FLIGHT 1: Number of Judges = 9 Number of Wines = 10
Identification of the Wine: The judges' overall ranking:
Wine A is Ch. Margaux 1989 tied for 5th place Wine B is Opus One 1990 ........ 4th place Wine C is Ch. Mouton Rothschild 1990 ........ 10th place Wine D is Opus One 1994 ........ 1st place Wine E is Opus One 1989 ........ 3rd place Wine F is Ch. Lafite Rothschild 1982 ........ 9th place Wine G is Opus One 1997 ........ 2nd place Wine H is Ch. Cheval Blanc 1994 ........ 7th place Wine I is Ch. Haut Brion 1997 ........ 8th place Wine J is Opus One 1982 tied for 5th place
The Judges's Rankings
Judge Wine -> A B C D E F G H I J Bob 3. 5. 10. 1. 2. 7. 4. 6. 9. 8. Mark 10. 8. 3. 2. 7. 9. 1. 5. 6. 4. Mike 7. 2. 9. 4. 3. 8. 10. 5. 6. 1. Ed 5. 1. 10. 3. 6. 9. 8. 2. 7. 4. Burt 5. 4. 8. 1. 10. 6. 2. 7. 9. 3. John 5. 4. 10. 1. 3. 8. 2. 7. 6. 9. Alan 4. 10. 8. 9. 1. 6. 2. 7. 3. 5. Orley 9. 3. 10. 1. 2. 4. 5. 6. 7. 8. Dick 3. 7. 8. 2. 4. 5. 1. 10. 6. 9.
Table of Votes Against Wine -> A B C D E F G H I J
Group Ranking -> 5 4 10 1 3 9 2 7 8 5 Votes Against -> 51 44 76 24 38 62 35 55 59 51
( 9 is the best possible, 90 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 = 0.3003
The probability that random chance could be responsible for this correlation is quite small, 0.0038. 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 0.8667 Bob 0.8182 Orley 0.6485 Dick 0.4742 Burt 0.4438 Ed 0.3647 Mike 0.2736 Mark 0.1824 Alan -0.0122
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 D is Opus One 1994 --------------------------------------------------- 2. ........ 2nd place Wine G is Opus One 1997 3. ........ 3rd place Wine E is Opus One 1989 4. ........ 4th place Wine B is Opus One 1990 5. tied for 5th place Wine A is Ch. Margaux 1989 6. tied for 5th place Wine J is Opus One 1982 7. ........ 7th place Wine H is Ch. Cheval Blanc 1994 8. ........ 8th place Wine I is Ch. Haut Brion 1997 9. ........ 9th place Wine F is Ch. Lafite Rothschild 1982 --------------------------------------------------- 10. ........ 10th place Wine C is Ch. Mouton Rothscild 1990 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 24.3212. The probability that this could happen by chance is 0.0038 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 0.65 for significance at the 0.05 level and must exceed 0.56 for significance at the 0.1 level Bob Mark Mike Bob 1.000 -0.042 0.200 Mark -0.042 1.000 -0.164 Mike 0.200 -0.164 1.000 Ed 0.418 -0.115 0.758 Burt 0.370 0.394 0.079 John 0.867 0.164 0.091 Alan 0.115 -0.055 -0.212 Orley 0.673 0.030 0.345 Dick 0.685 0.091 -0.370 Ed Burt John Bob 0.418 0.370 0.867 Mark -0.115 0.394 0.164 Mike 0.758 0.079 0.091 Ed 1.000 0.345 0.333 Burt 0.345 1.000 0.382 John 0.333 0.382 1.000 Alan -0.418 -0.345 0.164 Orley 0.358 0.224 0.721 Dick -0.236 0.358 0.770 Alan Orley Dick Bob 0.115 0.673 0.685 Mark -0.055 0.030 0.091 Mike -0.212 0.345 -0.370 Ed -0.418 0.358 -0.236 Burt -0.345 0.224 0.358 John 0.164 0.721 0.770 Alan 1.000 -0.103 0.370 Orley -0.103 1.000 0.418 Dick 0.370 0.418 1.000 Pairwise correlations in descending order 0.867 Bob and John Significantly positive 0.770 John and Dick Significantly positive 0.758 Mike and Ed Significantly positive 0.721 John and Orley Significantly positive 0.685 Bob and Dick Significantly positive 0.673 Bob and Orley Significantly positive 0.418 Orley and Dick Not significant 0.418 Bob and Ed Not significant 0.394 Mark and Burt Not significant 0.382 Burt and John Not significant 0.370 Alan and Dick Not significant 0.370 Bob and Burt Not significant 0.358 Ed and Orley Not significant 0.358 Burt and Dick Not significant 0.345 Mike and Orley Not significant 0.345 Ed and Burt Not significant 0.333 Ed and John Not significant 0.224 Burt and Orley Not significant 0.200 Bob and Mike Not significant 0.164 John and Alan Not significant 0.164 Mark and John Not significant 0.115 Bob and Alan Not significant 0.091 Mark and Dick Not significant 0.091 Mike and John Not significant 0.079 Mike and Burt Not significant 0.030 Mark and Orley Not significant -0.042 Bob and Mark Not significant -0.055 Mark and Alan Not significant -0.103 Alan and Orley Not significant -0.115 Mark and Ed Not significant -0.164 Mark and Mike Not significant -0.212 Mike and Alan Not significant -0.236 Ed and Dick Not significant -0.345 Burt and Alan Not significant -0.370 Mike and Dick Not significant -0.418 Ed and Alan Not significant
COMMENT (Prepared by Orley Ashenfelter): For the Francophiles in the group this tasting was a shock. The wines were served randomly using an ingenious design due to our host, Burt Malkiel. The tasting consisted of 10 wines: 5 Opus I bottles from 5 different vintages (1982, 1989, 1990, 1994, and 1997) and 5 Bordeaux first growths from each of the same 5 vintages. There were three key results. The Opus I 1994 was the overall top wine (significantly so) and the 1990 Mouton was at the bottom (significantly so also). In addition, in each vintage Opus I gathered more favorable results than the comparable French wine. This was perhaps most especially shocking in the case of the comparison of the 1982 Lafite and Opus I, as the former is rated 100 points by Robert Parker. Moreover, the usual argument is that the Bordeaux wines will fare better when they are older, certainly not the case here. (However, at least one person in the group stated that the 82 Lafite is notoriously unreliable, possibly due to the fact that in 1982 very limited selection within the vineyard was made. This is borne out by the production then which was almost double todays level. Visitors to the chateau have heard this used as a justification as to why todays wines are so much better (and more expensive). Of course it does not offer much consolation to those who are prepared to spend the +$700 at auction for this wine today. We do not have any independent evidence to support this assertion, of course.) Each person was asked to guess whether a wine was an Opus (CA) or First Growth (FR). In fact, only one person was very successful at this, and he (Mark Massad) was the very generous contributor of the Opus wines---showing his familiarity with them and his generosity. In short, no one was able to pinpoint the origin of the wines, though the group generally preferred the Opus wines. It would be fair to say the wines are made in a style very similar to that of a First Growth. The similarity of the wines is perhaps no accident as, historically, the Opus wines were a joint venture of Robert Mondavi and Chateau Mouton-Rothschild. Opus I is now a 50/50 joint venture between Constellation (the purchaser of the Robert Mondavi enterprise) and Baron Philipe de Rothschild S.A. (the owners of Mouton Rothschild). The Opus property is administered by a separate board of directors and has a distinctly separate management team. The winery is the first major American producer to use the Bordeaux negociant system for worldwide distribution (the same method used by first growths). [The information in this paragraph was kindly provided by Ted Hall, who was Chairman of the Robert Mondavi Board at the time of its sale.]
Return to previous page