WINETASTER ON 06/06/11 WITH 6 JUDGES AND 6 WINES BASED ON RANKS, IDENT=Y Copyright (c) 1995-2011 Richard E. Quandt, V. 1.65

FLIGHT 1: Number of Judges = 6 Number of Wines = 6

Identification of the Wine: The judges' overall ranking: Grape:

Wine A is Mer etSoleil Chardonnay 2007 ........ 3rd place Chardonnay, unoaked, US Wine B is Clos de la Bolliataz 2009 ........ 1st place Chasselas, Switzerland Wine C is Cuma 2009 ........ 6th place Torrontes, Argentina Wine D is Céleste Saint Joseph 2008 ........ 4th place Rousanne, Rhone Wine E is Furmint Royal Tokaji 2006 ........ 2nd place Furmint, Hungary Wine F is Quinta de Muradella 2006 ........ 5th place Dona Blanca, Spain

The Judges's Rankings

Judge Wine -> A B C D E F Alexa 5. 3. 6. 4. 2. 1. Mike 1. 2. 3. 5. 4. 6. Orley 4. 5. 6. 2. 1. 3. Burtr 1. 3. 2. 6. 5. 4. John 6. 3. 5. 2. 1. 4. Dick 2. 1. 3. 4. 5. 6.

Table of Votes Against Wine -> A B C D E F

Group Ranking -> 3 1 6 4 2 5 Votes Against -> 19 17 25 23 18 24

( 6 is the best possible, 36 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 = 0.0921

The probability that random chance could be responsible for this correlation is rather large, 0.7366. 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.

The correlation I measures the degree to which the identification of each judge is correlated with the truth. Here a 1.0 means that the judge identified the wines perfectly, and a 0 means that he identified none of them.

Correlation Between the Ranks of Each Person With the Average Ranking of Others

Name of Person Correlation R Correlation I Dick 0.0857 1.0000 Mike 0.0304 1.0000 Alexa -0.2648 0.0000 John -0.3479 1.0000 Orley -0.3479 1.0000 Burtr -0.4857 1.0000

Next, we show the correlation among the wine identifications of the judges, which also ranges between 1.0 and 0.0:

C = 0.6667

The probability that random chance could be responsible for this correlation is rather large: > 10%. Most people would say that unless this probability is less than 0.1, the judges' identifications are not highly related.

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 B is Clos de la Bolliataz 2009 2. ........ 2nd place Wine E is Furmint Royal Tokaji 3. ........ 3rd place Wine A is Mer et`Soleil Chardonnay 2007 4. ........ 4th place Wine D is Céleste Saint Joseph 2008 5. ........ 5th place Wine F is Quinta de Muradella 6. ........ 6th place Wine C is Cuma 2009 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 2.7619. The probability that this could happen by chance is 0.7366 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 0.89 for significance at the 0.05 level and must exceed 0.83 for significance at the 0.1 level Alexa Mike Orley Alexa 1.000 -0.600 0.600 Mike -0.600 1.000 -0.543 Orley 0.600 -0.543 1.000 Burtr -0.543 0.771 -0.771 John 0.543 -0.486 0.714 Dick -0.600 0.886 -0.657 Burtr John Dick Alexa -0.543 0.543 -0.600 Mike 0.771 -0.486 0.886 Orley -0.771 0.714 -0.657 Burtr 1.000 -0.886 0.600 John -0.886 1.000 -0.371 Dick 0.600 -0.371 1.000 Pairwise correlations in descending order 0.886 Mike and Dick Significantly positive 0.771 Mike and Burtr Not significant 0.714 Orley and John Not significant 0.600 Alexa and Orley Not significant 0.600 Burtr and Dick Not significant 0.543 Alexa and John Not significant -0.371 John and Dick Not significant -0.486 Mike and John Not significant -0.543 Alexa and Burtr Not significant -0.543 Mike and Orley Not significant -0.600 Alexa and Mike Not significant -0.600 Alexa and Dick Not significant -0.657 Orley and Dick Not significant -0.771 Orley and Burtr Not significant -0.886 Burtr and John Significantly negative

COMMENT: The objective of this winetasting was to get a wide cross section of grapes that were used to make these wines. The wines were arranged into three flights, since it was deemed difficult to cope with tasting 18 wines in a single flight. The division of the 18 wines into the three flights was not based on any scientific principles. One wine had some residual sugar and at least one person loved it and one person did not. It turned out to be the unoaked California chardonnay. These wines were all of high quality and which you would want to drink would depend on what you are eating and where you are eating it. There was basically a bipolar distribuition by which three people liked the driest and least fruity wines and the other three preferred the fruiter ones.

Return to previous page

Report

WINETASTER ON 06/06/11 WITH 6 JUDGES AND 6 WINES BASED ON RANKS, IDENT=Y Copyright (c) 1995-2011 Richard E. Quandt, V. 1.65

FLIGHT 2: Number of Judges = 6 Number of Wines = 6

Identification of the Wine: The judges' overall ranking: Grape:

Wine A is Spies Gruner Veltliner 2008 ........ 5th place Grüer Veltliner, Austria Wine B is Cour Cheverney 2005 tied for 1st place Sauvignon blanc, Loire Wine C is Schloss Gobelsburg 2007 tied for 1st place Riesling, Austria Wine D is Qupe 2009 tied for 3rd place Marsanne 85%, Rousane 15%, US Wine E is Bourgogne Aligote2008 tied for 3rd place Aligoté, Burgundy Wine F is In Fine 2008 ........ 6th place Clairette 80%, Bourboulenc 20%, Côtes Ventoux

The Judges's Rankings

Judge Wine -> A B C D E F Alexa 3. 4. 2. 6. 1. 5. Mike 5. 1. 2. 3. 4. 6. Orley 5. 1. 4. 2. 3. 6. Burtr 5. 3. 1. 4. 2. 6. John 4. 1. 3. 2. 5. 6. Dick 1. 6. 4. 3. 5. 2.

Table of Votes Against Wine -> A B C D E F

Group Ranking -> 5 1 1 3 3 6 Votes Against -> 23 16 16 20 20 31

( 6 is the best possible, 36 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 = 0.2476

The probability that random chance could be responsible for this correlation is rather large, 0.1907. 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.

The correlation I measures the degree to which the identification of each judge is correlated with the truth. Here a 1.0 means that the judge identified the wines perfectly, and a 0 means that he identified none of them.

Correlation Between the Ranks of Each Person With the Average Ranking of Others

Name of Person Correlation R Correlation I Mike 0.8857 0.0000 Burtr 0.6377 1.0000 Orley 0.5218 1.0000 John 0.4928 1.0000 Alexa -0.0290 0.0000 Dick -0.8857 1.0000

Next, we show the correlation among the wine identifications of the judges, which also ranges between 1.0 and 0.0:

C = 0.4667

The probability that random chance could be responsible for this correlation is rather large: > 10%. Most people would say that unless this probability is less than 0.1, the judges' identifications are not highly related.

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 B is Cour Cheverney 2005 2. tied for 1st place Wine C is Schloss Gobelsburg 2007 3. tied for 3rd place Wine D is Qupe 2009 4. tied for 3rd place Wine E is Bourgogne Aligote2008 5. ........ 5th place Wine A is Spies Gruner Veltliner 2008 --------------------------------------------------- 6. ........ 6th place Wine F is In Fine 2008 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 7.4286. The probability that this could happen by chance is 0.1907 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 0.89 for significance at the 0.05 level and must exceed 0.83 for significance at the 0.1 level Alexa Mike Orley Alexa 1.000 0.086 -0.086 Mike 0.086 1.000 0.829 Orley -0.086 0.829 1.000 Burtr 0.657 0.714 0.486 John -0.257 0.886 0.829 Dick -0.314 -0.771 -0.771 Burtr John Dick Alexa 0.657 -0.257 -0.314 Mike 0.714 0.886 -0.771 Orley 0.486 0.829 -0.771 Burtr 1.000 0.371 -0.714 John 0.371 1.000 -0.486 Dick -0.714 -0.486 1.000 Pairwise correlations in descending order 0.886 Mike and John Significantly positive 0.829 Mike and Orley Not significant 0.829 Orley and John Not significant 0.714 Mike and Burtr Not significant 0.657 Alexa and Burtr Not significant 0.486 Orley and Burtr Not significant 0.371 Burtr and John Not significant 0.086 Alexa and Mike Not significant -0.086 Alexa and Orley Not significant -0.257 Alexa and John Not significant -0.314 Alexa and Dick Not significant -0.486 John and Dick Not significant -0.714 Burtr and Dick Not significant -0.771 Mike and Dick Not significant -0.771 Orley and Dick Not significant

COMMENT: In this tasting we had one oxidized wine, namely the one rated last. The high quality riesling, though not significant, was the highest scorer along with wine B,from the Côte de Ventoux.

Return to previous page

Report

WINETASTER ON 06/06/11 WITH 6 JUDGES AND 6 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-2011 Richard E. Quandt, V. 1.65

FLIGHT 3: Number of Judges = 6 Number of Wines = 6

Identification of the Wine: The judges' overall ranking: Grape:

Wine A is Efeso Val di Neto 2004 ........ 6th place Efeso, Italy Wine B is Domaine de Villargeau 2009 ........ 1st place Sauvignon blanc. Loire Wine C is Pannonhalmi Tricollis 2008 ........ 2nd place Riesling 40%, Welsh Riesling 40%, Traminer 20% Wine D is Apremont 2008 tied for 4th place Jacquère, Savoie Wine E is Dom Bardo 2007 tied for 4th place Albariño, Spain Wine F is Trimbach Gewürztraminer 2006 ........ 3rd place Gewürztraminer, Alsace

The Judges's Rankings

Judge Wine -> A B C D E F Alexa 6. 1. 4. 3. 2. 5. Mike 4. 1. 3. 6. 5. 2. Orley 6. 3. 2. 4. 5. 1. Burt 5. 1. 4. 3. 2. 6. John 6. 3. 2. 4. 5. 1. Dick 2. 3. 1. 5. 6. 4.

Table of Votes Against Wine -> A B C D E F

Group Ranking -> 6 1 2 4 4 3 Votes Against -> 29 12 16 25 25 19

( 6 is the best possible, 36 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 = 0.3270

The probability that random chance could be responsible for this correlation is quite small, 0.0808. 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 Mike 0.7143 Orley 0.7143 John 0.7143 Alexa 0.3769 Burt 0.3143 Dick 0.1739

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 B is Domaine de Viullargeau 2009 --------------------------------------------------- 2. ........ 2nd place Wine C is Pannonhalmi TRricollis 2008 3. ........ 3rd place Wine F is Efeso Val di Neto 2004 4. tied for 4th place Wine D is Apremont 2008 5. tied for 4th place Wine E is Dom Baredo 2007 --------------------------------------------------- 6. ........ 6th place Wine A is Efeso Val di Neto 2004 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 9.8095. The probability that this could happen by chance is 0.0808 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 0.89 for significance at the 0.05 level and must exceed 0.83 for significance at the 0.1 level Alexa Mike Orley Alexa 1.000 0.086 0.029 Mike 0.086 1.000 0.600 Orley 0.029 0.600 1.000 Burt 0.943 -0.029 -0.257 John 0.029 0.600 1.000 Dick -0.429 0.486 0.200 Burt John Dick Alexa 0.943 0.029 -0.429 Mike -0.029 0.600 0.486 Orley -0.257 1.000 0.200 Burt 1.000 -0.257 -0.314 John -0.257 1.000 0.200 Dick -0.314 0.200 1.000 Pairwise correlations in descending order 1.000 Orley and John Significantly positive 0.943 Alexa and Burt Significantly positive 0.600 Mike and John Not significant 0.600 Mike and Orley Not significant 0.486 Mike and Dick Not significant 0.200 Orley and Dick Not significant 0.200 John and Dick Not significant 0.086 Alexa and Mike Not significant 0.029 Alexa and John Not significant 0.029 Alexa and Orley Not significant -0.029 Mike and Burt Not significant -0.257 Orley and Burt Not significant -0.257 Burt and John Not significant -0.314 Burt and Dick Not significant -0.429 Alexa and Dick Not significant

COMMENT: The clear winner was the sauvignong blanc, but close behind was the Pannonhalmi Tr1eicollis.

Return to previous page