WINETASTER ON 12/07/09 WITH 8 JUDGES AND 9 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-2009 Richard E. Quandt, V. 1.65
FLIGHT 1: A 1997 Napa Tasting Number of Judges = 8 Number of Wines = 9
Identification of the Wine: The judges' overall ranking:
Wine A is Stags Leap Cask 23 ........ 4th place Wine B is Paul Hobbs Michael Black Vineyard ........ 3rd place Wine C is Kendall Jackson Grand Reserve ........ 5th place Wine D is Chimney Rock ........ 8th place Wine E is Anderson Conn Valley Reserve ........ 9th place Wine F is Lewis L Reserve ........ 2nd place Wine G is Beringer Private Reserve ........ 6th place Wine H is Beaulieu Vineyard Georges de Latour ........ 7th place Wine I is Joseph Phelps Insignia ........ 1st place
The Judges's Rankings
Judge Wine -> A B C D E F G H I Mike 1. 9. 2. 8. 4. 5. 3. 7. 6. Jerry 5. 1. 6. 7. 9. 4. 3. 8. 2. Greg 5. 2. 7. 8. 9. 3. 6. 4. 1. Burt 4. 3. 2. 7. 8. 1. 9. 6. 5. Bob 7. 1. 5. 3. 9. 2. 8. 6. 4. Ed 6. 5. 2. 8. 9. 4. 1. 7. 3. John 4. 3. 9. 8. 6. 7. 5. 2. 1. Dick 1. 8. 4. 9. 7. 2. 5. 6. 3.
Table of Votes Against Wine -> A B C D E F G H I
Group Ranking -> 4 3 5 8 9 2 6 7 1 Votes Against -> 33 32 37 58 61 28 40 46 25
( 8 is the best possible, 72 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 = 0.3365
The probability that random chance could be responsible for this correlation is quite small, 0.0059. 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 Correlation Price Greg 0.7782 0.4667 Jerry 0.7113 0.3000 Burt 0.5607 -0.0833 Ed 0.5294 0.2167 Dick 0.5167 0.2000 Bob 0.3025 -0.0333 John 0.0753 0.5667 Mike -0.0502 -0.0667
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 I is Joseph Phelps Insignia --------------------------------------------------- 2. ........ 2nd place Wine F is Lewis L Reserve 3. ........ 3rd place Wine B is Paul Hobbs Michael Black Vineyard 4. ........ 4th place Wine A is Stags Leap Cask 23 5. ........ 5th place Wine C is Kendall Jackson Grand Reserve 6. ........ 6th place Wine G is Beringer Private Reserve 7. ........ 7th place Wine H is Beaulieu Vineyard Georges de Latour --------------------------------------------------- 8. ........ 8th place Wine D is Chimney Rock 9. ........ 9th place Wine E is Anderson Conn Valley Reserve We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 21.5333. The probability that this could happen by chance is 0.0059
We now test whether the group ranking of wines is correlated with the prices of the wines. The rank correlation between them is 0.1833. At the 10% level of significance this would have to exceed the critical value of 0.4830 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 0.70 for significance at the 0.05 level and must exceed 0.60 for significance at the 0.1 level Mike Jerry Greg Mike 1.000 -0.167 -0.350 Jerry -0.167 1.000 0.750 Greg -0.350 0.750 1.000 Burt 0.033 0.333 0.483 Bob -0.650 0.517 0.567 Ed 0.333 0.667 0.383 John -0.300 0.383 0.700 Dick 0.683 0.250 0.367 Burt Bob Ed Mike 0.033 -0.650 0.333 Jerry 0.333 0.517 0.667 Greg 0.483 0.567 0.383 Burt 1.000 0.650 0.267 Bob 0.650 1.000 0.117 Ed 0.267 0.117 1.000 John -0.150 -0.017 0.000 Dick 0.467 -0.133 0.467 John Dick Mike -0.300 0.683 Jerry 0.383 0.250 Greg 0.700 0.367 Burt -0.150 0.467 Bob -0.017 -0.133 Ed 0.000 0.467 John 1.000 0.117 Dick 0.117 1.000 Pairwise correlations in descending order 0.750 Jerry and Greg Significantly positive 0.700 Greg and John Significantly positive 0.683 Mike and Dick Significantly positive 0.667 Jerry and Ed Significantly positive 0.650 Burt and Bob Significantly positive 0.567 Greg and Bob Not significant 0.517 Jerry and Bob Not significant 0.483 Greg and Burt Not significant 0.467 Ed and Dick Not significant 0.467 Burt and Dick Not significant 0.383 Greg and Ed Not significant 0.383 Jerry and John Not significant 0.367 Greg and Dick Not significant 0.333 Mike and Ed Not significant 0.333 Jerry and Burt Not significant 0.267 Burt and Ed Not significant 0.250 Jerry and Dick Not significant 0.117 John and Dick Not significant 0.117 Bob and Ed Not significant 0.033 Mike and Burt Not significant 0.000 Ed and John Not significant -0.017 Bob and John Not significant -0.133 Bob and Dick Not significant -0.150 Burt and John Not significant -0.167 Mike and Jerry Not significant -0.300 Mike and John Not significant -0.350 Mike and Greg Not significant -0.650 Mike and Bob Significantly negative
COMMENT: The date that will live in infamy! This tasting represented a cross section of the 1997 vintage. Overall the wines were fresh, enjoyable with fruit and they had a lot of life in them. The alcohol levels of these wines were from 13.3 to 14.5 % alcohol, which appears to be lower than these same vineyards are producing. The winner, wine I, was 13.8% and the loser was 13.3%. The Merlot finished third and shows that merlot has greater longevity than usually expected. Everybody agreed that all the wines in the tasting were of extremely high quality. We also computed the Spearman Rho rank correlation between the group's ranking and Robert Parker's ranking. In order to compute this we had to omit the one merlot in the group (Stags Leap Cask 23) for which we had no Parker rating. The correlation coefficient was 0.205, which is not statistically significant.
Return to previous page