AAWE, Economics Dept, New York University, 19 W 4th St, 6Fl., New York NY 10012aawe@wine-economics.org

AAWE Rheinhessen Riesling Tasting with Stuart Pigott

Post 215 of 227

AAWE Rheinhessen Riesling Tasting with Stuart Pigott
November 28, 2012, at Trestle on Tenth, New York City

On November 28, 2012, we held the 3rd AAWE Wine Tasting, this time with Berlin-based English-born wine writer Stuart Pigott at the Manhattan restaurant Trestle on Tenth.  Stuart organized the tasting and put together an impressive line of Rheinhessen Rieslings. Here you will find an introductory text on Rheinhessen by Stuart. Here is a 10 min video (produced by Marcarthur Baralla).

We tasted 15 dry Rieslings (A-O) in five flights (F1–F5) of three wines each (see Table 1).  All wines were from the 2011 vintage. In order to compare the Rheinhessen wines with Rieslings from other regions, we included three ringers (wine A from the Finger Lakes in upstate New York, F from the Rheingau, M from the Wachau region in Austria). The judges had to assess the wines on the 100-point scale.

 

Table 1
The Wines

A 2011 Wiemer Riesling Dry, Starkey, NY, Finger Lakes
F1 B 2011 Rothenberg GG, Gunderloch, Nackenheim
C 2011 ‘Rabenturm,’ Fritz Ekkehard Huff, Schwabsburg
D 2011 Heerkretz GG, Weingut Wagner-Stempel, Siefersheim
F2 E 2011 Scharlachberg ‘Turm,’ Weingut Riffel, Bingen-Büdesheim
F 2011 Schloss Johannisberg Erstes Gewächs, Domäne Schl Jberg, Rheingau
G 2011 Hundertgulden, Weingut Hofmann, Appenheim
F3 H 2011 Leckerberg, Weingut Winter, Dittelsheim-Hessloch
I 2011 Kirchenstück GG, Weingut Battenfeld-Spanier, Hohen-Sülzen
J 2011 Kirchspiel GG , Weingut Groebe, Westhofen
F4 K 2011 Morstein GG, Weingut Wittmann, Westhofen
L 2011 Hubacker GG, Weingut Keller in Flörsheim-Dalsheim
M 2011 Kellerberg Smaragd, Weingut F.X. Pichler, Loiben, Wachau
F5 N 2011 Geyersberg, Weingut Dreissigacker, Bechtheim
O 2011 Hölle, Weingut Thörle, Saulheim

 

The results are displayed in Table 2.

Table 2
Results Based on Points Awarded

 

Flight 1

Flight 2

Flight 3

Flight 4

Flight 5

 

 

Judge

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

Avg

Std
Dev

T. Joyce

89

85

85

86

91

87

88

90

89

89

89

92

87

90

90

88.5

2.10

D. Jaeger

90

85

89

87

93

92

90

93

92

95

92

91

90

96

92

91.1

2.83

O. Ashenfelter

82

80

75

80

82

83

81

83

85

84

81

89

79

84

86

82.3

3.28

J. Czerwinski

91

88

87

92

90

87

87

88

90

89

86

90

85

86

92

88.5

2.23

V. Tan

84

83

87

84

83

89

87

80

80

85

93

91

81

79

80

84.4

4.27

D. Rosengarten

89

86

92

87

92

91

88

84

92

87

89

94

85

86

89

88.7

2.96

C. Harrison

85

92

80

92

86

80

89

90

93

90

92

98

89

88

90

88.9

4.74

H. Goldberg

88

85

87

90

91

92

90

85

88

88

86

89

91

84

85

87.9

2.55

J. Frost

88

86

88

82

82

82

84

84

83

83

83

87

72

83

85

83.5

3.80

R. Gray

85

86

88

89

87

90

89

85

86

87

86

88

89

85

85

87.0

1.73

M. Reiser

82

78

88

93

85

87

86

85

74

90

88

82

95

98

85

86.4

6.21

S. Bitterolf

79

82

85

90

87

90

87

86

87

85

84

83

82

85

82

84.9

3.06

J. Haeger

86

82

84

90

85

86

82

79

85

87

83

88

82

80

84

84.2

2.98

K. Storchmann

75

80

83

92

90

91

93

92

94

95

90

94

92

93

96

90.0

5.96

M. Baralla

92

90

93

91

93

90

90

92

88

90

90

92

91

91

91

90.9

1.33

G. Lauer

87

85

91

88

89

90

89

90

90

92

92

93

87

85

87

89.0

2.48

C. Rzeszewski

86

84

84

87

85

88

80

81

81

81

82

86

84

86

84

83.9

2.46

S. Pigott

86

90

92

96

89

92

90

91

92

91

94

92

88

91

90

90.9

2.37

S. Lambiase

88

90

89

91

94

92

92

90

93

92

92

93

90

92

91

91.3

1.62

L. Gevirtz

87

89

86

89

90

84

88

88

88

89

90

97

95

92

90

89.5

3.27

Avg

86.0

85.3

86.7

88.8

88.2

88.2

87.5

86.8

87.5

88.5

88.1

90.5

86.7

87.7

87.7

87.6

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The bottom row reports the average score per wine. Accordingly, wine L received the highest marks (90.5 points) while wine B averaged at only 85.3 pts. The two columns on the far right provide each judge’s average score and the corresponding standard deviations.

 

The judge-specific average scores range from 82.3 (Ashenfelter) to 91.3 points (Lambiase). Most importantly, we notice a wide range from 1.33 (Baralla) to 6.21 (Reiser) in the corresponding standard variations from. As a result, if we simply calculated the average points received for each wine to find out what wine was first, second etc., our results would be biased. In fact, the overall results would be above-proportionally influenced by judges that exploit a wider range of the point scale (particularly, Reiser and Storchmann) compared to those with a small standard deviation.

 

In order to give all judges equal weights we, therefore, convert all point into ranks. For m (20) judges and n (15) wines, the maximum ranksum is n*m=300. The minimum ranksum equals n=15. Note that the ranksum are “points against.” Hence, 15 is the best possible ranksum while 300 is the worst possible ranksum. Ranks and ranksums are reported in Table 3.


Table 3
Ranks and Ranksums

 

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

Joyce

7.5

14.5

14.5

13

2

11.5

10

4

7.5

7.5

7.5

1

11.5

4

4

Jaeger

11

15

13

14

3.5

6.5

11

3.5

6.5

2

6.5

9

11

1

6.5

Ashenfelter

8.5

12.5

15

12.5

8.5

6.5

10.5

6.5

3

4.5

10.5

1

14

4.5

2

Czerwinski

3

8.5

11

1.5

5

11

11

8.5

5

7

13.5

5

15

13.5

1.5

Tan

7.5

9.5

4.5

7.5

9.5

3

4.5

13

13

6

1

2

11

15

13

Rosengarten

7

12.5

3.5

10.5

3.5

5

9

15

2

10.5

7

1

14

12.5

7

Harrison

13

4

14.5

4

12

14.5

9.5

7

2

7

4

1

9.5

11

7

Goldberg

8

13

10

4.5

2.5

1

4.5

13

8

8

11

6

2.5

15

13

Frost

1.5

4

1.5

13

13

13

6.5

6.5

9.5

9.5

9.5

3

15

9.5

5

Gray

13.5

10

5.5

3

7.5

1

3

13.5

10

7.5

10

5.5

3

13.5

13.5

Reiser

12.5

14

5.5

3

10

7

8

10

15

4

5.5

12.5

2

1

10

Bitterolf

15

13

8

1.5

4

1.5

4

6

4

8

10

11

13

8

13

Haeger

4.5

12.5

8.5

1

6.5

4.5

12.5

15

6.5

3

10

2

11

14

8.5

Storchmann

15

14

13

8

11.5

10

5.5

8

3.5

2

11.5

3.5

8

5.5

1

Baralla

4

12

1.5

7.5

1.5

12

12

4

15

12

12

4

7.5

7.5

7.5

Lauer

12

14.5

4

10

8.5

6

8.5

6

6

2.5

2.5

1

12

14.5

12

Rzeszewski

4

8.5

8.5

2

6

1

15

13

13

13

11

4

8.5

4

8.5

Pigott

15

11

4.5

1

13

4.5

11

8

4.5

8

2

4.5

14

8

11

Lambiase

15

12

14

9.5

1

6

6

12

2.5

6

6

2.5

12

6

9.5

Gevirtz

13

8

14

8

5

15

11

11

11

8

5

1

2

3

5

Ranksum

190.5

223

174.5

135

134

140.5

173

183.5

147.5

136

156

80.5

196.5

171

158.5

 

Concordance among the Wine Judges

In order to measure the concordance among the judges we report Kendall’s coefficient of concordance (Kendall’s W). The coefficient ranges from 0 (no correlation) to 1 (perfect correlation). For the entire sample of all 15 wines we find that W=0.1432. The probability that random chance could be responsible for this correlation is quite small, 0.0002 (0.02%). Most analysts would say that the judges’ preferences are only weakly related.

Table 4
Kendall’s W

Kendall’s W Significance
p-value
Flight 1 0.0944 0.1514
Flight 2 0.0131 0.7691
Flight 3 0.0675 0.2592
Flight 4 0.2294 0.0102
Flight 5 0.0644 0.2760
All 0.1432 0.0002

 

Table 4 reports the judge concordance within each flight. As we see, the judges exhibit more agreement regarding the wines in Flight 4 than within other Flights.

 

Preferred Wines

Based on the ranksums Table 5 reports the following order of preferences (which is identical with the Borda count method)

Table 5
Ranksums and Preferred Wines

Rank Code Wine Name

Ranksum

1st L 2011 Hubacker GG, Weingut Keller in Flörsheim-Dalsheim  

80.5

2nd E 2011 Scharlachberg ‘Turm,’ Weingut Riffel, Bingen-Büdesheim

134

3rd D 2011 Heerkretz GG, Weingut Wagner-Stempel, Siefersheim

135

4th J 2011 Kirchspiel GG , Weingut Groebe, Westhofen

136

5th F 2011 Schloss Johannisberg Erstes Gewächs, Domäne Schl Jberg, Rheingau

140.5

6th I 2011 Kirchenstück GG, Weingut Battenfeld-Spanier, Hohen-Sülzen

147.5

7th K 2011 Morstein GG, Weingut Wittmann, Westhofen

156

8th O 2011 Hölle, Weingut Thörle, Saulheim

158.5

9th N 2011 Geyersberg, Weingut Dreissigacker, Bechtheim

171

10th G 2011 Hundertgulden, Weingut Hofmann, Appenheim

173

11th C 2011 ‘Rabenturm,’ Fritz Ekkehard Huff, Schwabsburg

174.5

12th H 2011 Leckerberg, Weingut Winter, Dittelsheim-Hessloch

183.5

13th A 2011 Wiemer Riesling Dry, Starkey, NY, Finger Lakes

190.5

14th M 2011 Kellerberg Smaragd, Weingut F. X. Pichler in Loiben, Wachau

196.5

15th B 2011 Rothenberg GG, Gunderloch, Nackenheim

223

 

The wines in bold are significantly different from the other wines at the 5% significance level. That is, we find 4 significantly better wines and three wines that are significantly less preferred. It is noteworthy to mention that Wine L is the distant winner of the tasting.

 

We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 40.09. The probability that this could happen by chance is 0.0002. We can, therefore, conclude with considerable confidence that the observed differences among the mean rankings for the 15 wines reflect something more than mere random variability, something more than mere chance coincidence among the judgments of the wine judges.
References

Quandt, R.E. (2006). Measurement and inference in wine tasting. Journal of Wine Economics, 1(1), 7-30.

Quandt, R.E. (2007). A note on a test on the sum of ranksums. Journal of Wine Economics, 2(1), 98-102.

 

This article was written by Karl Storchmann

1 comment:

A Video Interview with Stuart Pigott | Appellation Nature ContrôléeJanuary 13, 2013 at 8:35 pmReply

[…] American Association of Wine Economists (for more information concerning this tasting and results, click here). A week later, we interviewed Stuart Pigott at the new Terroir wine bar in Park Slope, Brooklyn […]

Menu