Report

WINETASTER ON 09/08/08 WITH 8 JUDGES AND 4 WINES BASED ON RANKS, IDENT=N
Copyright (c) 1995-2008 Richard E. Quandt, V. 1.65

FLIGHT 1:
Number of Judges = 8
Number of Wines = 4

Identification of the Wine: The judges' overall ranking:

Wine A is 1997 Merryvale Profile ........ 2nd place
Wine B is 1995 Ch. Musar ........ 1st place
Wine C is 1996 Ch. Lascombes ........ 4th place
Wine D is 1997 Brunello di Montalcino, Fanti ........ 3rd place

The Judges's Rankings

Judge Wine -> A B C D

Eamonn 2. 1. 3. 4.
Hago 4. 2. 1. 3.
David 3. 1. 4. 2.
Scott 3. 1. 2. 4.
Kirk 2. 1. 3. 4.
Tom 1. 3. 4. 2.
Derek 3. 1. 4. 2.
Dick 2. 1. 4. 3.

Table of Votes Against

Wine -> A B C D

Group Ranking -> 2 1 4 3
Votes Against -> 20 11 25 24

( 8 is the best possible, 32 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 = 0.3812

The probability that random chance could be responsible for this correlation
is quite small, 0.0274. 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

Dick 0.9487
Eamonn 0.8000
Kirk 0.8000
Scott 0.4000
David 0.4000
Derek 0.4000
Tom 0.0000
Hago -0.4000

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 1995 Ch. Musar
---------------------------------------------------
2. ........ 2nd place Wine A is 1997 Merryvale Profile
3. ........ 3rd place Wine D is 1997 Brunello di Montalcino, Fanti
---------------------------------------------------
4. ........ 4th place Wine C is 1996 Ch. Lascombes

We now test whether the ranksums AS A WHOLE provide a significant ordering.
The Friedman Chi-square value is 9.1500. The probability that this could
happen by chance is 0.0274

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 1.00 for significance at the 0.05
level and must exceed 1.00 for significance at the 0.1 level

Eamonn Hago David

Eamonn 1.000 0.000 0.400
Hago 0.000 1.000 -0.200
David 0.400 -0.200 1.000
Scott 0.800 0.600 0.200
Kirk 1.000 0.000 0.400
Tom 0.000 -1.000 0.200
Derek 0.400 -0.200 1.000
Dick 0.800 -0.400 0.800

Scott Kirk Tom

Eamonn 0.800 1.000 0.000
Hago 0.600 0.000 -1.000
David 0.200 0.400 0.200
Scott 1.000 0.800 -0.600
Kirk 0.800 1.000 0.000
Tom -0.600 0.000 1.000
Derek 0.200 0.400 0.200
Dick 0.400 0.800 0.400

Derek Dick

Eamonn 0.400 0.800
Hago -0.200 -0.400
David 1.000 0.800
Scott 0.200 0.400
Kirk 0.400 0.800
Tom 0.200 0.400
Derek 1.000 0.800
Dick 0.800 1.000

Pairwise correlations in descending order

1.000 Eamonn and Kirk Significantly positive
1.000 David and Derek Significantly positive
0.800 Scott and Kirk Not significant
0.800 Eamonn and Dick Not significant
0.800 Kirk and Dick Not significant
0.800 David and Dick Not significant
0.800 Eamonn and Scott Not significant
0.800 Derek and Dick Not significant
0.600 Hago and Scott Not significant
0.400 Eamonn and Derek Not significant
0.400 David and Kirk Not significant
0.400 Eamonn and David Not significant
0.400 Tom and Dick Not significant
0.400 Scott and Dick Not significant
0.400 Kirk and Derek Not significant
0.200 Tom and Derek Not significant
0.200 David and Tom Not significant
0.200 Scott and Derek Not significant
0.200 David and Scott Not significant
0.000 Kirk and Tom Not significant
0.000 Eamonn and Hago Not significant
0.000 Eamonn and Tom Not significant
0.000 Hago and Kirk Not significant
-0.200 Hago and Derek Not significant
-0.200 Hago and David Not significant
-0.400 Hago and Dick Not significant
-0.600 Scott and Tom Not significant
-1.000 Hago and Tom Significantly negative

COMMENT:

We first note the host's ingenious design. Each flight had a '97 American
Cabernet, a '96 Bordeaux, a '97 Brunello and one "mystery wine" from the
1990s, that remained unidentified until the end of the tasting. All the
wines were in impeccable condition and had been decanted. The degree of
agreement among the judges was remarkable and statistically highly signi-
ficant. Everybody realized that wine B was the mystery wine but nobody
identified it. One taster thought that B might have been a Burgundy
because of its color, nose and taste; but Ch. Musar does that to you at
times.

Return to previous page

WINETASTER ON 09/09/08 WITH 8 JUDGES AND 4 WINES BASED ON RANKS, IDENT=N
Copyright (c) 1995-2003 Richard E. Quandt, V. 1.65

FLIGHT 2:
Number of Judges = 8
Number of Wines = 4

Identification of the Wine: The judges' overall ranking:

Wine A is 1996 Lynch Bages ........ 1st place
Wine B is 1997 Brunello di Montalcino, Carmigliano tied for 2nd place
Wine C is 1992 Penfolds Grange ........ 4th place
Wine D is 1997 Martin Ray Diamond Mountain tied for 2nd place

The Judges's Rankings

Judge Wine -> A B C D

Eamonn 1. 3. 2. 4.
Hago 3. 1. 4. 2.
David 1. 3. 4. 2.
Scott 1. 2. 4. 3.
Kirk 1. 2. 3. 4.
Tom 1. 3. 2. 4.
Derek 1. 4. 3. 2.
Dick 2. 4. 3. 1.

Table of Votes Against

Wine -> A B C D

Group Ranking -> 1 2 4 2
Votes Against -> 11 22 25 22

( 8 is the best possible, 32 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 = 0.3562

The probability that random chance could be responsible for this correlation
is quite small, 0.0359. 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

David 0.8000
Scott 0.8000
Kirk 0.4000
Derek 0.4000
Tom 0.2000
Eamonn 0.2000
Dick 0.0000
Hago -0.1054

1. ........ 1st place Wine A is 1996 Lynch Bages
---------------------------------------------------
2. tied for 2nd place Wine B is 1997 Brunello di Montalcino, Carmigliano
3. tied for 2nd place Wine D is 1997 Martin Ray Diamond Mountain
---------------------------------------------------
4. ........ 4th place Wine C is 1992 Penfolds Grange

We now test whether the ranksums AS A WHOLE provide a significant ordering.
The Friedman Chi-square value is 8.5500. The probability that this could
happen by chance is 0.0359

Pairwise Rank Correlations

Correlations must exceed in absolute value 1.00 for significance at the 0.05
level and must exceed 1.00 for significance at the 0.1 level

Eamonn Hago David

Eamonn 1.000 -0.600 0.200
Hago -0.600 1.000 0.200
David 0.200 0.200 1.000
Scott 0.400 0.400 0.800
Kirk 0.800 0.000 0.400
Tom 1.000 -0.600 0.200
Derek 0.400 -0.400 0.800
Dick -0.200 -0.200 0.600

Scott Kirk Tom

Eamonn 0.400 0.800 1.000
Hago 0.400 0.000 -0.600
David 0.800 0.400 0.200
Scott 1.000 0.800 0.400
Kirk 0.800 1.000 0.800
Tom 0.400 0.800 1.000
Derek 0.400 0.200 0.400
Dick 0.000 -0.400 -0.200

Derek Dick

Eamonn 0.400 -0.200
Hago -0.400 -0.200
David 0.800 0.600
Scott 0.400 0.000
Kirk 0.200 -0.400
Tom 0.400 -0.200
Derek 1.000 0.800
Dick 0.800 1.000

Pairwise correlations in descending order

1.000 Eamonn and Tom Significantly positive
0.800 Eamonn and Kirk Not significant
0.800 David and Derek Not significant
0.800 David and Scott Not significant
0.800 Kirk and Tom Not significant
0.800 Scott and Kirk Not significant
0.800 Derek and Dick Not significant
0.600 David and Dick Not significant
0.400 David and Kirk Not significant
0.400 Eamonn and Derek Not significant
0.400 Eamonn and Scott Not significant
0.400 Hago and Scott Not significant
0.400 Scott and Tom Not significant
0.400 Scott and Derek Not significant
0.400 Tom and Derek Not significant
0.200 Hago and David Not significant
0.200 Kirk and Derek Not significant
0.200 Eamonn and David Not significant
0.200 David and Tom Not significant
0.000 Scott and Dick Not significant
0.000 Hago and Kirk Not significant
-0.200 Eamonn and Dick Not significant
-0.200 Hago and Dick Not significant
-0.200 Tom and Dick Not significant
-0.400 Hago and Derek Not significant
-0.400 Kirk and Dick Not significant
-0.600 Eamonn and Hago Not significant
-0.600 Hago and Tom Not significant

COMMENT:

The degree of agreement among the judges continued in this second flight.
Ch. Lynch Bages was the easy winner and it is surpising that the Penfolds
showed up as badly as it did.

Return to previous page

WINETASTER ON 09/09/08 WITH 8 JUDGES AND 4 WINES BASED ON RANKS, IDENT=N
Copyright (c) 1995-2003 Richard E. Quandt, V. 1.65

FLIGHT 3:
Number of Judges = 8
Number of Wines = 4

Identification of the Wine: The judges' overall ranking:

Wine A is 1997 Spotswood (Napa) ........ 1st place
Wine B is 1999 Bodegas y Vinedos Alion ........ 4th place
Wine C is 1997 Brunello di Montalcino Altesino,Montosoli ........ 3rd place
Wine D is 1996 Ch. d'Armailhac ........ 2nd place

The Judges's Rankings

Judge Wine -> A B C D

Eamonn 1. 3. 4. 2.
Hago 1. 3. 2. 4.
David 1. 3. 4. 2.
Scott 3. 4. 2. 1.
Kirk 3. 4. 1. 2.
Tom 4. 2. 1. 3.
Derek 2. 4. 3. 1.
Dick 1. 4. 2. 3.

Table of Votes Against

Wine -> A B C D

Group Ranking -> 1 4 3 2
Votes Against -> 16 27 19 18

( 8 is the best possible, 32 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 = 0.2188

The probability that random chance could be responsible for this correlation
is rather large, 0.1544. 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

Dick 0.6325
Derek 0.4000
Scott 0.3162
Kirk 0.2000
Eamonn 0.1054
David 0.1054
Hago -0.2000
Tom -0.8000

1. ........ 1st place Wine A is 1997 Spotswood (Napa)
2. ........ 2nd place Wine D is 1996 Ch. d'Armailhac
3. ........ 3rd place Wine C is 1997 Brunello di Montalcino Altesi
---------------------------------------------------
4. ........ 4th place Wine B is 1999 Bodegas y Vinedos Alion

We now test whether the ranksums AS A WHOLE provide a significant ordering.
The Friedman Chi-square value is 5.2500. The probability that this could
happen by chance is 0.1544

Pairwise Rank Correlations

Correlations must exceed in absolute value 1.00 for significance at the 0.05
level and must exceed 1.00 for significance at the 0.1 level

Eamonn Hago David

Eamonn 1.000 0.200 1.000
Hago 0.200 1.000 0.200
David 1.000 0.200 1.000
Scott 0.000 -0.400 0.000
Kirk -0.400 0.000 -0.400
Tom -1.000 -0.200 -1.000
Derek 0.600 -0.200 0.600
Dick 0.400 0.800 0.400

Scott Kirk Tom

Eamonn 0.000 -0.400 -1.000
Hago -0.400 0.000 -0.200
David 0.000 -0.400 -1.000
Scott 1.000 0.800 0.000
Kirk 0.800 1.000 0.400
Tom 0.000 0.400 1.000
Derek 0.800 0.400 -0.600
Dick 0.200 0.400 -0.400

Derek Dick

Eamonn 0.600 0.400
Hago -0.200 0.800
David 0.600 0.400
Scott 0.800 0.200
Kirk 0.400 0.400
Tom -0.600 -0.400
Derek 1.000 0.400
Dick 0.400 1.000

Pairwise correlations in descending order

1.000 Eamonn and David Significantly positive
0.800 Scott and Derek Not significant
0.800 Hago and Dick Not significant
0.800 Scott and Kirk Not significant
0.600 Eamonn and Derek Not significant
0.600 David and Derek Not significant
0.400 David and Dick Not significant
0.400 Kirk and Dick Not significant
0.400 Kirk and Tom Not significant
0.400 Kirk and Derek Not significant
0.400 Derek and Dick Not significant
0.400 Eamonn and Dick Not significant
0.200 Eamonn and Hago Not significant
0.200 Hago and David Not significant
0.200 Scott and Dick Not significant
0.000 David and Scott Not significant
0.000 Eamonn and Scott Not significant
0.000 Hago and Kirk Not significant
0.000 Scott and Tom Not significant
-0.200 Hago and Derek Not significant
-0.200 Hago and Tom Not significant
-0.400 Hago and Scott Not significant
-0.400 David and Kirk Not significant
-0.400 Eamonn and Kirk Not significant
-0.400 Tom and Dick Not significant
-0.600 Tom and Derek Not significant
-1.000 David and Tom Significantly negative
-1.000 Eamonn and Tom Significantly negative

COMMENT:

The agreement among the judges was slightly less strong in Flight 3. The
Napa wine was the winner in this flight and the Spanish wine, from
Ribera del Duero, the loser.

Overall, this was a most interesting and rewarding tasting.

Return to previous page