Bordeaux wines: ccvobg.org, Scores' system,
standard deviation
Comparative scales and calculation process

The main issue in averaging scores consists in comparing the existing different
rating systems (there is unfortunately no ISO norm governing this delicate subject for the moment!). Basically, there are two different approaches in the notation of wines: the American approach which is extremely precise, often based on a scholar evaluation scale base of 100 points, which allows a more subtle differentiation
of winetasting. The European approach is more adaptable, the objective of which is to communicate a general impression of the wine in comparison to a precise expertise of the product. A wine noted 86.5 is it really much better than it’s cousin at 85? It is a question of flavors and colors…
To simplify our calculation approach, we have decided to adopt the American approach, so as to proceed in a more efficient manner in different comparative calculations. The European scales are consequently transposed to a scale of 100 points
i.e. a wine rated with a 4 stars value will be transposed in our study in a bracket of notation of between 90 and 92 points, thus we have systematically adopted in our data a mean value of 91
points. The concern in knowing why the winetaster, after having rated a wine at 4 stars, would have chosen 90 instead of 92 points remains a mystery
deeply anchored in the thoughts of the expert. 

It must be noted however that Contributors who have the habit of noting their wines on a scale of 20 points, show a tendency to assign more easily a lower rating in comparison to winetasters using other grading systems, more particularly for wines that are rated disappointing. Consequently, when using a rating system on 20 points, it is not rare to find notations of 12/20 which transposes to a 60/100 rating, despite the fact that winetasters using the 100 point scale would rarely have the occasion to evaluate wines under a value of 70/100, this value representing the extreme lower limit in acceptable quality criteria. 

ccvobg's approach
The table of scores is established by using the mathematical average of the results of all the
Contributors. However, you should avoid falling in certain traps when reading the table. Take for example two wines with an average
rating of 85/100.


Score 1

Score 2 
Score 3 
Average 

Wine A

83 / 100

86 / 100 
86 / 100 
85 / 100

Wine B

76 / 100

96 / 100 
83 / 100 
85 / 100 



For Wine A we are confronted with a group of homogeneous
scores where three wine critics have a very similar opinion of this particular wine.
For Wine B, even if the final average result is equivalent to Wine A, the conclusions to be drawn in this example are totally different.
Here we have the first winetaster who obviously did not appreciate this wine,
the second who loved it and the third critic who has a rather neutral opinion. As a result, the notation of 85/100 cannot be considered as
representative of true value of Wine B. Thus, we decided to add to the average
score another statistical value: the standard deviation.


Standard deviation :
The standard deviation measures the dispersion of values of a numerical variable around the average of these values. In our case, the higher is the value, the greater is the dispersion
or distribution around the average and viceversa.
The standard deviationis commonly used in the domain of finance as a measure of risk, but also in the scientific and social sciences. In the above example, the value of the standard deviation for Wine A is valued at 1.73, a result relatively close to a value of zero (zero meaning that there is no difference in the value of the three notations). For Wine B, the standard deviation is valued at 10.15, a result far from the value zero, which means that the dispersion of the notations around the average of 85 is high. It certainly does not mean that the Wine B’s quality is poorer than Wine A’s, but that under some specific circumstances, (i.e. bottles presenting a flaw in a winetasting session, or a winetaster which appreciates less a type of wine, and so forth) a more thorough examination will be necessary. For wines whose standard deviations are high, we have developed inserts containing the tasters’ original comments on some scores presenting a high differential. This will help you to determine the reasons for some of the differences in the notations.

TOP 
ccvobg's comparative table 
100 
25 
20 
***** 
99.5 

19.8 
***** 
99.0 

19.6 
***** 
98.5 

19.4 
***** 
98.0 
24.5 
19.2 
***** 
97.5 

19.0 
***** 
97.0 

18.8 
***** 
96.5 

18.6 
***** 
96.0 
24 
18.4 
***** 
95.5 

18.2 
***** 
95.0 

18.0 
***** 
94.5 

17.8 
****(*) 
94.0 
23.5 
17.6 
****(*) 
93.5 

17.4 
****(*) 
93.0 

17.2 
****(*) 
92.5 

17.0 
****(*) 
92.0 
23 
16.8 
**** 
91.5 

16.6 
**** 
91.0 

16.4 
**** 
90.5 
22.5 
16.2 
**** 
90.0 

16.0 
**** 
89.0 

15.6 
**** 
88.0 
22 
15.2 
***(*) 
87.0 

14.8 
***(*) 
86.7 

14.7 
*** 
86.0 
21 
14.4 
*** 
85.0 

14.0 
*** 
84.0 
20 
13.6 
**(*) 
83.3 

13.3 
**(*) 
83.0 

13.2 
**(*) 
82.0 
19 
12.8 
** 
81.0 

12.4 
** 
80.3 

12.1 
** 
80.0 
18 
12.0 
** 
79.0 

11.6 
*(*) 
78.0 
17 
11.2 
*(*) 
77.8 

11.1 
*(*) 
77.0 

10.8 
* 
76.0 
16 
10.4 
* 
75.3 

10.1 
* 
75.0 

10.0 
* 
74.0 
15 
9.6 
(*) 
73.0 

9.2 
(*) 
72.8 

9.1 
(*) 
72.0 
14 
8.8 
 
71.0 

8.4 
 
70.3 

8.1 
 
70.0 
13 
8.0 
 
69.0 

7.6 
 
68.4 

7.4 
 
68.0 
12 
7.2 
 
67.0 

6.8 
 
66.0 
11 
6.4 
 
65.0 

6.0 
 
64.0 
10 
5.6 
 
63.0 

5.2 
 
62.0 
9 
4.8 
 
61.0 

4.4 
 
60.0 
8 
4.0 
 
Paul McCartney Miami however her bank adjust looks measly contrasted with those objection with its office since
Pink, otherwise called Pink, is a wine fan. Credit: Wiki Commons MediaSan Antonio Stock Show And Rodeo: Josh Turner
The new session of David Geffen In New York Philharmonic


Our stats Since Jan. 2002, ccvobg.org has welcomed +1'500,000
Bordeaux wine lovers
around the world.
Thank you for making ccvobg one of the most successful websites dedicated to
Bordeaux wines!
