WINETASTER ON 09/11/06 WITH 9 JUDGES AND 8 WINES BASED ON RANKS, Copyright (c) 1995-2006 Richard E. Quandt, V. 1.65
FLIGHT 1: Number of Judges = 9 Number of Wines = 8
Identification of the Wine: The judges' overall ranking:
Wine A is Vieux Telegraph 2000 ........ 8th place Wine B is Ch. de Beaucastel 2001 tied for 5th place Wine C is Ch. de Beaucastel 1999 ........ 7th place Wine D is Clos des Papes 1999 ........ 2nd place Wine E is Vieux Telegraph 2000 ........ 3rd place Wine F is Clos des Papes 2001 ........ 1st place Wine G is Vieux Telegraph 1998 tied for 5th place Wine H is Domaine Bois de Boursan 1998 ........ 4th place
The Judges' Rankings
Judge Wine -> A B C D E F G H Clemens 7. 5. 8. 6. 1. 3. 4. 2. John 3. 8. 2. 7. 6. 5. 4. 1. Bob 6. 8. 7. 3. 5. 2. 4. 1. Burt 7. 6. 5. 2. 1. 3. 4. 8. Orley 3. 2. 1. 4. 5. 8. 6. 7. Ed 6. 3. 8. 5. 7. 1. 2. 4. Mike 7. 1. 4. 2. 3. 5. 6. 8. Frank 7. 3. 8. 2. 1. 4. 5. 6. Dick 8. 4. 7. 3. 6. 1. 5. 2.
Table of Votes Against Wine -> A B C D E F G H
Group Ranking -> 8 5 7 2 3 1 5 5 Votes Against -> 54 40 50 34 35 32 40 39
( 9 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 = 0.1235
The probability that random chance could be responsible for this correlation is rather large, 0.3526. 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. 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
Frank 0.5476 Dick 0.4551 Clemens 0.4048 Bob 0.2994 Burt 0.2275 Ed 0.1190 Mike -0.0838 John -0.7186 Orley -0.7306
Table of Aggregate Wine Quality
Wine Ranksum Significance Wine Ranksum Significance Alphabetic Order Ranksum Order A 54.0 SIGNIFICANTLY LOW QUALITY F 32.0 B 40.0 D 34.0 C 50.0 E 35.0 D 34.0 H 39.0 E 35.0 G 40.0 F 32.0 B 40.0 G 40.0 C 50.0 H 39.0 A 54.0 SIGNIFICANTLY LOW QUALITY
Friedman Test: Chi-square = 7.7778 Probability = 0.3526
Identification of Wines Votes Against Wine A is Vieux Telegraph 2000 54. Wine B is Ch. Beaucastel 2001 40. Wine C is Ch. de Beaucastel 1999 50. Wine D is Clos des Papes 1999 34. Wine E is Vieux Telegraph 2000 35. Wine F is Clos des Papes 2001 32. Wine G is Vieux Telegraph 1998 40. Wine H is Domaine Bois de Boursan 1998 39.
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
Clemens John Bob Clemens 1.000 -0.095 0.548 John -0.095 1.000 0.286 Bob 0.548 0.286 1.000 Burt 0.262 -0.571 0.095 Orley -0.762 -0.143 -0.881 Ed 0.357 -0.238 0.429 Mike -0.143 -0.857 -0.500 Frank 0.548 -0.833 0.119 Dick 0.500 -0.190 0.714 Burt Orley Ed Clemens 0.262 -0.762 0.357 John -0.571 -0.143 -0.238 Bob 0.095 -0.881 0.429 Burt 1.000 -0.167 -0.048 Orley -0.167 1.000 -0.643 Ed -0.048 -0.643 1.000 Mike 0.548 0.476 -0.119 Frank 0.714 -0.238 0.190 Dick 0.095 -0.690 0.714 Mike Frank Dick Clemens -0.143 0.548 0.500 John -0.857 -0.833 -0.190 Bob -0.500 0.119 0.714 Burt 0.548 0.714 0.095 Orley 0.476 -0.238 -0.690 Ed -0.119 0.190 0.714 Mike 1.000 0.643 0.024 Frank 0.643 1.000 0.357 Dick 0.024 0.357 1.000
Pairwise correlations in descending order
0.714 Bob and Dick Significantly positive 0.714 Ed and Dick Significantly positive 0.714 Burt and Frank Significantly positive 0.643 Mike and Frank Not significant 0.548 Clemens and Frank Not significant 0.548 Clemens and Bob Not significant 0.548 Burt and Mike Not significant 0.500 Clemens and Dick Not significant 0.476 Orley and Mike Not significant 0.429 Bob and Ed Not significant 0.357 Frank and Dick Not significant 0.357 Clemens and Ed Not significant 0.286 John and Bob Not significant 0.262 Clemens and Burt Not significant 0.190 Ed and Frank Not significant 0.119 Bob and Frank Not significant 0.095 Bob and Burt Not significant 0.095 Burt and Dick Not significant 0.024 Mike and Dick Not significant -0.048 Burt and Ed Not significant -0.095 Clemens and John Not significant -0.119 Ed and Mike Not significant -0.143 John and Orley Not significant -0.143 Clemens and Mike Not significant -0.167 Burt and Orley Not significant -0.190 John and Dick Not significant -0.238 John and Ed Not significant -0.238 Orley and Frank Not significant -0.500 Bob and Mike Not significant -0.571 John and Burt Not significant -0.643 Orley and Ed Not significant -0.690 Orley and Dick Significantly negative -0.762 Clemens and Orley Significantly negative -0.833 John and Frank Significantly negative -0.857 John and Mike Significantly negative -0.881 Bob and Orley Significantly negative .
Comments: First, we should say that the astute reader of these notes may have noticed that there were two identical bottles of Vieux Telegraph 2000 included in the tasting. This was an accident that not even the host of the tasting was aware of until the covers were removed from the wines. The host's intention had been to include a bottle of Ch. de Beaucastel 2000 instead of a Vieux Telegraph 2000 again. All the wines were in fine condition and no bottle was corked or otherwise spoiled; under ideal circumstances the two identical bottles of Vieux Telegraph 2000 should have been ranked very close to each other in the tasting. This did not happen: one of these bottles was ranked worst, and the other was ranked third best in the aggregate. In fact, four judges ranked these wines either adjacent to each other or next to adjacent; the other five ranked them farther apart.
The 2001 vintage in Chateauneuf du Pape is reputed to be not quite ready and rather tannic; we did not find this to be the case and all the wines were drinking well. They were sufficiently similar in quality that the ranking was not statistically sifgnificant. But it is noteworthy that the two highest rankings went to the two bottles of Clos des Papes. All the wines with one exception were in the same price range (around $50 a bottle) with the exception of Domaine Bois de Boursan, which cost $125; while it ranked among the top six wines (which were nearly indistinguishable in terms of the rankings), it was not the clear winner.
Return to previous page