WINETASTER ON 10-13-06 WITH 13 JUDGES AND 8 WINES BASED ON RANKS,
Copyright (c) 1995-2006
Wine Tasting on Oct 6, 2006 in Montclair, NJ
Number of judges = 13
Number of wines = 8
The wines and their identifying code designations
Wine name Code
Dumas Station, Cabernet, Walla Walla 03 A
Chateau Gloria, St Julien, 03 B
Ch La Vielle Cure, Fronsac, 03 C
Whitman Cellars, Merlot, Walla Walla 03 D
Covey Run, Cabernet, Columbia River 03 E
Ch de la Haute Libarde, Cote de Bourg, 03 F
Woodward Canyon, Cab, Columbia River 03 G
Ch Phelan-Segur, St Julien, 03 H
Rank Table for Judges
Judge Wine -> A B C D E F G H
Paul 1. 4. 3. 6. 2. 8. 5. 7.
Robert 3. 2. 5. 1. 4. 7. 6. 8.
Craig 5. 2. 4. 7. 3. 8. 1. 6.
David 7. 3. 6. 1. 5. 4. 2. 8.
Andy 6. 2. 3. 1. 4. 8. 5. 7.
Jordan 4. 7. 6. 2. 3. 8. 1. 5.
Adam 3. 5. 8. 2. 4. 6. 7. 1.
Loren 5. 3. 6. 1. 2. 8. 7. 4.
Bruce 6. 4. 7. 1. 3. 5. 2. 8.
Ben 6. 5. 7. 1. 4. 8. 3. 2.
Richard 3. 2. 8. 4. 7. 6. 1. 5.
Diego 4. 2. 5. 6. 1. 3. 8. 7.
Karl 3. 2. 5. 1. 7. 8. 4. 6.
Table of Votes Against
Wine -> A B C D E F G H
Rank
1 1. 0. 0. 7. 1. 0. 3. 1.
2 0. 6. 0. 2. 2. 0. 2. 1.
3 4. 2. 2. 0. 3. 1. 1. 0.
4 2. 2. 1. 1. 4. 1. 1. 1.
5 2. 2. 3. 0. 1. 1. 2. 2.
6 3. 0. 3. 2. 0. 2. 1. 2.
7 1. 1. 2. 1. 2. 1. 2. 3.
8 0. 0. 2. 0. 0. 7. 1. 3.
Votes -> 56. 43. 73. 34. 49. 87. 52. 74.
We now measure the amount of correlation (W) among the judges:
W = 1.0 => perfect correlation, W = 0 => no correlation
We also provide a probability measure that this correlation is due to chance
In this case the correlation is = 0.3159
Probability that correlation is due to chance = 0.0002
We compute the correlation of each taster with the average ranking of the oth-
ers and with prices. 1.0 => perfect, 0 => none, -1.0 => total disagreement
Name of Person Correlation R Correlation Price
Robert 0.7857 -0.2036
Andy 0.7619 0.1078
Karl 0.7306 0.3952
Bruce 0.7109 -0.0599
Loren 0.6429 0.0359
Ben 0.5476 0.6467
Jordan 0.5238 0.3234
David 0.4762 0.0838
Richard 0.4048 0.4311
Craig 0.3114 0.0958
Paul 0.2036 -0.4192
Adam 0.1557 0.2275
Diego -0.0714 -0.8743
Rank correlation between the average ranking of wines and the prices
Correlation = 0.0479
Critical value = 0.5240
Table of Aggregate Wine Quality
Wine Ranksum Significance Wine Ranksum Significance
Alphabetic Order Ranksum Order
A 56.0 D 34.0 SIGNIFICANTLY HIGH QUALITY
B 43.0 B 43.0
C 73.0 E 49.0
D 34.0 SIGNIFICANTLY HIGH QUALITY G 52.0
E 49.0 A 56.0
F 87.0 SIGNIFICANTLY LOW QUALITY C 73.0
G 52.0 H 74.0 SIGNIFICANTLY LOW QUALITY
H 74.0 SIGNIFICANTLY LOW QUALITY F 87.0 SIGNIFICANTLY LOW QUALITY
Friedman Test: Chi-square = 28.7436 Probability = 0.0002
Identification of Wines Votes Against
Wine A is Dumas Station, Cabernet, Walla Walla 03 56.
Wine B is Chateau Gloria, St Julien, 03 43.
Wine C is Ch La Vielle Cure, Fronsac, 03 73.
Wine D is Whitman Cellars, Merlot, Walla Walla 03 34.
Wine E is Covey Run, Cabernet, Columbia River 03 49.
Wine F is Ch de la Haut Libarde, Cote de Bourg, 03 87.
Wine G is Woodward Canyon, Cab, Columbia River 03 52.
Wine H is Ch Phelan-Segur, St Julien, 03 74.
Pairwise Rank Correlations
Correlations must exceed in absolute value 0.74 for significance at the 0.0
level and must exceed 0.64 for significance at the 0.1 level
Paul Robert Craig
Paul 1.000 0.476 0.524
Robert 0.476 1.000 0.143
Craig 0.524 0.143 1.000
David -0.262 0.476 0.167
Andy 0.310 0.810 0.333
Jordan 0.238 0.238 0.333
Adam -0.119 0.167 -0.429
Loren 0.238 0.667 0.024
Bruce -0.024 0.548 0.238
Ben -0.190 0.190 0.095
Richard -0.024 0.262 0.405
Diego 0.381 0.333 0.024
Karl 0.238 0.786 0.214
David Andy Jordan
Paul -0.262 0.310 0.238
Robert 0.476 0.810 0.238
Craig 0.167 0.333 0.333
David 1.000 0.548 0.333
Andy 0.548 1.000 0.286
Jordan 0.333 0.286 1.000
Adam -0.238 -0.048 0.214
Loren 0.167 0.667 0.333
Bruce 0.905 0.524 0.595
Ben 0.286 0.357 0.714
Richard 0.429 0.095 0.357
Diego -0.071 0.095 -0.476
Karl 0.452 0.714 0.357
Adam Loren Bruce
Paul -0.119 0.238 -0.024
Robert 0.167 0.667 0.548
Craig -0.429 0.024 0.238
David -0.238 0.167 0.905
Andy -0.048 0.667 0.524
Jordan 0.214 0.333 0.595
Adam 1.000 0.643 -0.048
Loren 0.643 1.000 0.357
Bruce -0.048 0.357 1.000
Ben 0.619 0.643 0.429
Richard 0.119 0.000 0.405
Diego -0.071 0.238 0.024
Karl 0.214 0.476 0.405
Ben Richard Diego
Paul -0.190 -0.024 0.381
Robert 0.190 0.262 0.333
Craig 0.095 0.405 0.024
David 0.286 0.429 -0.071
Andy 0.357 0.095 0.095
Jordan 0.714 0.357 -0.476
Adam 0.619 0.119 -0.071
Loren 0.643 0.000 0.238
Bruce 0.429 0.405 0.024
Ben 1.000 0.357 -0.500
Richard 0.357 1.000 -0.333
Diego -0.500 -0.333 1.000
Karl 0.429 0.619 -0.238
Karl
Paul 0.238
Robert 0.786
Craig 0.214
David 0.452
Andy 0.714
Jordan 0.357
Adam 0.214
Loren 0.476
Bruce 0.405
Ben 0.429
Richard 0.619
Diego -0.238
Karl 1.000
Pairwise correlations in descending order
0.905 David and Bruce Significantly positive
0.810 Robert and Andy Significantly positive
0.786 Robert and Karl Significantly positive
0.714 Jordan and Ben Significantly positive
0.714 Andy and Karl Significantly positive
0.667 Robert and Loren Significantly positive
0.667 Andy and Loren Significantly positive
0.643 Loren and Ben Not significant
0.643 Adam and Loren Not significant
0.619 Richard and Karl Not significant
0.619 Adam and Ben Not significant
0.595 Jordan and Bruce Not significant
0.548 David and Andy Not significant
0.548 Robert and Bruce Not significant
0.524 Andy and Bruce Not significant
0.524 Paul and Craig Not significant
0.476 Paul and Robert Not significant
0.476 Loren and Karl Not significant
0.476 Robert and David Not significant
0.452 David and Karl Not significant
0.429 Ben and Karl Not significant
0.429 Bruce and Ben Not significant
0.429 David and Richard Not significant
0.405 Bruce and Richard Not significant
0.405 Bruce and Karl Not significant
0.405 Craig and Richard Not significant
0.381 Paul and Diego Not significant
0.357 Ben and Richard Not significant
0.357 Jordan and Richard Not significant
0.357 Loren and Bruce Not significant
0.357 Andy and Ben Not significant
0.357 Jordan and Karl Not significant
0.333 Craig and Jordan Not significant
0.333 Jordan and Loren Not significant
0.333 David and Jordan Not significant
0.333 Craig and Andy Not significant
0.333 Robert and Diego Not significant
0.310 Paul and Andy Not significant
0.286 David and Ben Not significant
0.286 Andy and Jordan Not significant
0.262 Robert and Richard Not significant
0.238 Paul and Jordan Not significant
0.238 Paul and Loren Not significant
0.238 Paul and Karl Not significant
0.238 Loren and Diego Not significant
0.238 Robert and Jordan Not significant
0.238 Craig and Bruce Not significant
0.214 Adam and Karl Not significant
0.214 Craig and Karl Not significant
0.214 Jordan and Adam Not significant
0.190 Robert and Ben Not significant
0.167 David and Loren Not significant
0.167 Craig and David Not significant
0.167 Robert and Adam Not significant
0.143 Robert and Craig Not significant
0.119 Adam and Richard Not significant
0.095 Craig and Ben Not significant
0.095 Andy and Diego Not significant
0.095 Andy and Richard Not significant
0.024 Craig and Loren Not significant
0.024 Craig and Diego Not significant
0.024 Bruce and Diego Not significant
0.000 Loren and Richard Not significant
-0.024 Paul and Bruce Not significant
-0.024 Paul and Richard Not significant
-0.048 Adam and Bruce Not significant
-0.048 Andy and Adam Not significant
-0.071 Adam and Diego Not significant
-0.071 David and Diego Not significant
-0.119 Paul and Adam Not significant
-0.190 Paul and Ben Not significant
-0.238 Diego and Karl Not significant
-0.238 David and Adam Not significant
-0.262 Paul and David Not significant
-0.333 Richard and Diego Not significant
-0.429 Craig and Adam Not significant
-0.476 Jordan and Diego Not significant
-0.500 Ben and Diego Not significant
Comments: This was a wine tasting with a total of 26 judges. The group was divided into two sub-groups
(in two different rooms): 13 men and 13 women. Note that the two groups (Report101
and Report102) tasted identical wines. The two groups agreed on which wine was significantly good
and on one of the wines that was significantly bad, but they did not agree on a second wine that
was significantly bad. The rank correlation between the two groups' aggregate evaluations was
0.472, which is not significant statistically at even the 0.1 level.
Return to the previous page