WINETASTER ON 02/06/2006 WITH 8 JUDGES AND 9 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-2006 Richard E. Quandt, V. 1.65
FLIGHT 1: Number of Judges = 8 Number of Wines = 9
Identification of the Wine: The judges' overall ranking:
Wine A is Ch. Latour 1998 ........ 1st place Wine B is Ch. Latour 1997 ........ 2nd place Wine C is Ch. Clinet 1996 ........ 7th place Wine D is Ch. Clinet 1997 ........ 5th place Wine E is Ch. Clinet 1998 tied for 3rd place Wine F is Ch.Pape Clement 1996 tied for 8th place Wine G is Ch. Latour 1996 tied for 8th place Wine H is Ch.Pape Clement 1997 ........ 6th place Wine I is Ch.Pape Clement 1998 tied for 3rd place
The Judges's Rankings
Judge Wine -> A B C D E F G H I Burt 7. 3. 8. 5. 2. 4. 6. 9. 1. Alexa 2. 7. 3. 5. 4. 6. 8. 9. 1. Mike 2. 7. 9. 8. 1. 5. 4. 3. 6. Bob 3. 1. 8. 4. 6. 9. 7. 2. 5. Frank 2. 6. 4. 5. 3. 1. 7. 8. 9. Ed 8. 3. 2. 1. 4. 9. 5. 6. 7. John 4. 5. 9. 3. 6. 8. 7. 1. 2. Dick 4. 1. 5. 6. 8. 9. 7. 2. 3.
Table of Votes Against Wine -> A B C D E F G H I
Group Ranking -> 1 2 7 5 3 8 8 6 3 Votes Against -> 32 33 48 37 34 51 51 40 34
( 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 = 0.1302
The probability that random chance could be responsible for this correlation is rather large, 0.4016. 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 Bob 0.4686 John 0.3405 Dick 0.0921 Alexa -0.0588 Burt -0.1255 Mike -0.2000 Ed -0.3361 Frank -0.3866
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 A is Ch. Latour 1998 2. ........ 2nd place Wine B is Ch. Latour 1997 3. tied for 3rd place Wine E is Ch. Clinet 1998 4. tied for 3rd place Wine I is Ch.Pape Clement 1998 5. ........ 5th place Wine D is Ch. Clinet 1997 6. ........ 6th place Wine H is Ch.Pape Clement 1997 7. ........ 7th place Wine C is Ch. Clinet 1996 8. tied for 8th place Wine F is Ch.Pape Clement 1996 9. tied for 8th place Wine G is Ch. Latour 1996 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 8.3333. The probability that this could happen by chance is 0.4016 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 Burt Alexa Mike Burt 1.000 0.350 0.017 Alexa 0.350 1.000 -0.100 Mike 0.017 -0.100 1.000 Bob -0.067 -0.183 0.117 Frank -0.050 0.217 0.183 Ed -0.067 -0.100 -0.550 John 0.033 -0.017 0.200 Dick -0.150 -0.033 -0.200 Bob Frank Ed Burt -0.067 -0.050 -0.067 Alexa -0.183 0.217 -0.100 Mike 0.117 0.183 -0.550 Bob 1.000 -0.400 0.150 Frank -0.400 1.000 -0.183 Ed 0.150 -0.183 1.000 John 0.750 -0.583 -0.100 Dick 0.817 -0.600 0.117 John Dick Burt 0.033 -0.150 Alexa -0.017 -0.033 Mike 0.200 -0.200 Bob 0.750 0.817 Frank -0.583 -0.600 Ed -0.100 0.117 John 1.000 0.600 Dick 0.600 1.000 Pairwise correlations in descending order 0.817 Bob and Dick Significantly positive 0.750 Bob and John Significantly positive 0.600 John and Dick Significantly positive 0.350 Burt and Alexa Not significant 0.217 Alexa and Frank Not significant 0.200 Mike and John Not significant 0.183 Mike and Frank Not significant 0.150 Bob and Ed Not significant 0.117 Ed and Dick Not significant 0.117 Mike and Bob Not significant 0.033 Burt and John Not significant 0.017 Burt and Mike Not significant -0.017 Alexa and John Not significant -0.033 Alexa and Dick Not significant -0.050 Burt and Frank Not significant -0.067 Burt and Bob Not significant -0.067 Burt and Ed Not significant -0.100 Ed and John Not significant -0.100 Alexa and Mike Not significant -0.100 Alexa and Ed Not significant -0.150 Burt and Dick Not significant -0.183 Alexa and Bob Not significant -0.183 Frank and Ed Not significant -0.200 Mike and Dick Not significant -0.400 Bob and Frank Not significant -0.550 Mike and Ed Not significant -0.583 Frank and John Not significant -0.600 Frank and Dick Significantly negative
COMMENT: It was a great tasting with very good wines. There was an unusual amount of disagreement in the group and there were no significantly great wines, nor significantly bad ones. There were two separate groupings of three wines each: by the three vintages (1996, 1997, 1998) and by the three Chateaux, representing Pauillac, Pomerol and Graves. While we have no significance test for the sum of the rank sums, they were as follows 1996: 150, 1997: 110, 1998: 100; whereas the ranksums for the regions were Pauillac: 116, Pomerol: 125, Graves: 119. The differences among the regions are inconsequential, but those for the vintages seem a bit greater. The 1996 wines have a way to go to not contradict their high standings in the ratings. The 1997s more than held their own---a trio in which the Parker ratings say that they should have been drubbed as of now. The same could be said to a lesser extent of the 1998s.
Return to previous page