Report

WINETASTER ON 10/07/13 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 Ponzi Vineyards 2010 ........ 4th place
Wine B is Vosne Romanée Les Suchots 1995 tied for 6th place
Wine C is Peter Michael Le Caprice 2010 ........ 8th place
Wine D is Mt. Difficulty 2008 tied for 1st place
Wine E is Echezeaux Grand Cru Forey 1997 ........ 5th place
Wine F is Pommard 1er cru Potinet-Ampeau 200 tied for 6th place
Wine G is Papietro Perry 2004 ........ 3rd place
Wine H is Kistler Cuvée Elizabeth 2009 tied for 1st place

The Judges's Rankings

Judge Wine -> A B C D E F G H

Ed 2. 6. 7. 1. 3. 4. 5. 8.
Mike 6. 2. 8. 1. 4. 3. 7. 5.
Zaki 3. 8. 7. 4. 5. 6. 1. 2.
Burt 4. 8. 2. 3. 7. 6. 5. 1.
Bob 6. 8. 1. 3. 5. 7. 4. 2.
Alexa 4. 8. 5. 2. 6. 7. 3. 1.
Allan 3. 1. 8. 5. 6. 7. 2. 4.
Dick 7. 3. 8. 6. 5. 4. 1. 2.

Table of Votes Against

Wine -> A B C D E F G H

Group Ranking -> 4 6 8 1 5 6 3 1
Votes Against -> 35 44 46 25 41 44 28 25

( 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 = 0.2083

The probability that random chance could be responsible for this correlation
is rather large, 0.1121. 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

Alexa 0.8095
Zaki 0.7306
Burt 0.2619
Bob 0.2275
Allan 0.1190
Dick 0.0952
Ed -0.0599
Mike -0.2619

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. tied for 1st place Wine D is Mt. Difficulty 2008
2. tied for 1st place Wine H is Kistler Cuvée Elizabeth 2009
3. ........ 3rd place Wine G is Papietro Perry 2004
4. ........ 4th place Wine A is Ponzi Vineyards 2010
5. ........ 5th place Wine E is Echezeaux Grand Cru Forey 1997
6. tied for 6th place Wine B is Vosne Romanée Les Suchots 1995
7. tied for 6th place Wine F is Pommard 1er cru Potinet-Ampeau 200
8. ........ 8th place Wine C is Peter Michael Le Caprice 2010

We now test whether the ranksums AS A WHOLE provide a significant ordering.
The Friedman Chi-square value is 11.6667. The probability that this could
happen by chance is 0.1121

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

Ed Mike Zaki

Ed 1.000 0.429 0.119
Mike 0.429 1.000 -0.310
Zaki 0.119 -0.310 1.000
Burt -0.262 -0.405 0.429
Bob -0.310 -0.476 0.333
Alexa 0.000 -0.214 0.810
Allan -0.024 0.143 0.310
Dick -0.381 0.119 0.405

Burt Bob Alexa

Ed -0.262 -0.310 0.000
Mike -0.405 -0.476 -0.214
Zaki 0.429 0.333 0.810
Burt 1.000 0.857 0.810
Bob 0.857 1.000 0.714
Alexa 0.810 0.714 1.000
Allan -0.310 -0.429 0.071
Dick -0.238 -0.214 0.119

Allan Dick

Ed -0.024 -0.381
Mike 0.143 0.119
Zaki 0.310 0.405
Burt -0.310 -0.238
Bob -0.429 -0.214
Alexa 0.071 0.119
Allan 1.000 0.571
Dick 0.571 1.000

Pairwise correlations in descending order

0.857 Burt and Bob Significantly positive
0.810 Burt and Alexa Significantly positive
0.810 Zaki and Alexa Significantly positive
0.714 Bob and Alexa Significantly positive
0.571 Allan and Dick Not significant
0.429 Ed and Mike Not significant
0.429 Zaki and Burt Not significant
0.405 Zaki and Dick Not significant
0.333 Zaki and Bob Not significant
0.310 Zaki and Allan Not significant
0.143 Mike and Allan Not significant
0.119 Mike and Dick Not significant
0.119 Ed and Zaki Not significant
0.119 Alexa and Dick Not significant
0.071 Alexa and Allan Not significant
0.000 Ed and Alexa Not significant
-0.024 Ed and Allan Not significant
-0.214 Bob and Dick Not significant
-0.214 Mike and Alexa Not significant
-0.238 Burt and Dick Not significant
-0.262 Ed and Burt Not significant
-0.310 Mike and Zaki Not significant
-0.310 Burt and Allan Not significant
-0.310 Ed and Bob Not significant
-0.381 Ed and Dick Not significant
-0.405 Mike and Burt Not significant
-0.429 Bob and Allan Not significant
-0.476 Mike and Bob Not significant

COMMENT:

The provenance of the wines was as follows: B, E and F were French,
C, G and H were from California, A ws from Oregon and D was from New Zealand.

The tasters' top four ranked wines were new world selections.
The Burgundies were the three oldest wines tasted and not from
the most distinguished vintages. There was clearly a stylistic preference
for different winemaking styles. The New Zealand wine seemed to bridge
the sweeter American wines and the drier French wines. All the
wines were drinking very well. Comparing the sum of the rank sums for the three
French wines with the corresponding sum of the non-French wines yields a test statistic
of 1.352, indicating that the non-French wines were significantly preferred to
the French ones. (See Quandt, R. E., "A Note on a Test for the Sum of Ranksums,"
Journal of Wine Economics, Vol. 2, No. 1, 2007, pp. 98-112.) It is also worth noting
that the agreement in the group was not very strong, yielding a Kendall W of 0.2083, and the
dispersion among the tasters is indicated by the fact that there were three wines each
of which received one or more first place votes as well as one or more last place votes.

Return to previous page