WINETASTER ON 04/25/99 WITH 11 JUDGES AND 10 WINES BASED ON RANKS, IDENT=N Copyright (c) 1995-99 Richard E. Quandt FLIGHT 1: Number of Judges = 11 Number of Wines = 10 Identification of the Wine: The judges' overall ranking: Wine A is J. Saffirio, Barolo 90 ........ 2nd place Wine B is R. Voerzio, Brunate, Barolo90 ........ 1st place Wine C is Ceretto, Barbaresco, Bricco Asili90 ........ 6th place Wine D is Gromis, Conteisa cerequio 90 ........ 3rd place Wine E is F. Rinaldi&Figli, Barolo90 ........ 4th place Wine F is D. Clerico, Ciablot Mentin90 ........ 10th place Wine G is Giacosa, Barbaresco90 ........ 9th place Wine H is Gaja, Sperrs90 ........ 5th place Wine I is B. Rocca, Rabaja90 ........ 8th place Wine J is Borgogno, Brunate90 ........ 7th place The Judges's Rankings Judge Wine -> A B C D E F G H I J wolfgang 4. 2. 3. 6. 7. 5. 8. 1. 9. 10. kai 3. 1. 4. 2. 6. 9. 7. 5. 8. 10. edgar 3. 2. 4. 5. 1. 10. 7. 9. 8. 6. fabrizio 5. 10. 6. 2. 3. 4. 9. 7. 8. 1. claude 9. 10. 8. 6. 7. 5. 3. 4. 1. 2. herbert 10. 9. 8. 7. 4. 5. 6. 1. 2. 3. helmut 2. 1. 6. 3. 5. 7. 4. 8. 10. 9. annette 2. 9. 4. 7. 5. 6. 3. 10. 1. 8. juergen 1. 3. 7. 8. 9. 6. 5. 2. 4. 10. karl 3. 1. 6. 2. 7. 8. 9. 5. 10. 4. bernd 8. 1. 7. 4. 2. 9. 10. 5. 6. 3. Table of Votes Against Wine -> A B C D E F G H I J Group Ranking -> 2 1 6 3 4 10 9 5 8 7 Votes Against -> 50 49 63 52 56 74 71 57 67 66 (11 is the best possible, 110 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 = 0.0720 The probability that random chance could be responsible for this correlation is rather large, 0.6240. 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 kai 0.7134 Karl 0.6220 edgar 0.5680 helmut 0.5574 bernd 0.3293 wolfgang 0.1951 juergen 0.0000 fabrizio -0.2796 annette -0.5714 herbert -0.6342 claude -0.7455 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 R. Voerzio, Brunate, Barolo90 2. ........ 2nd place Wine A is Saffirio, Barolo 90 3. ........ 3rd place Wine D is Gromis, Conteisa cereqiuo 90 4. ........ 4th place Wine E is F. Rinaldi&Figli, Barolo90 5. ........ 5th place Wine H is Gaja, Sperrs 6. ........ 6th place Wine C is Ceretto, Barbaresc, Bricco Asili90 7. ........ 7th place Wine J is Borgogno, Brunate90 8. ........ 8th place Wine I is B. Rocca, Rabaja90 9. ........ 9th place Wine G is Giacosa, Barbaresco90 10. ........ 10th place Wine F is D. Clerico, Ciablot Mentin90 We now test whether the ranksums AS A WHOLE provide a significant ordering. The Friedman Chi-square value is 7.1256. The probability that this could happen by chance is 0.6240 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 0.65 for significance at the 0.05 level and must exceed 0.56 for significance at the 0.1 level wolfgang kai edgar wolfgang 1.000 0.673 0.115 kai 0.673 1.000 0.588 edgar 0.115 0.588 1.000 fabrizio -0.370 -0.285 0.067 claude -0.673 -0.794 -0.721 herbert -0.345 -0.673 -0.588 helmut 0.406 0.794 0.648 annette -0.418 -0.176 0.055 juergen 0.576 0.394 -0.139 karl 0.455 0.697 0.503 bernd 0.055 0.297 0.539 fabrizio claude herbert wolfgang -0.370 -0.673 -0.345 kai -0.285 -0.794 -0.673 edgar 0.067 -0.721 -0.588 fabrizio 1.000 0.103 0.139 claude 0.103 1.000 0.806 herbert 0.139 0.806 1.000 helmut -0.200 -0.806 -0.855 annette -0.152 0.127 -0.200 juergen -0.697 -0.236 -0.212 karl 0.188 -0.624 -0.515 bernd 0.188 -0.236 0.115 helmut annette juergen wolfgang 0.406 -0.418 0.576 kai 0.794 -0.176 0.394 edgar 0.648 0.055 -0.139 fabrizio -0.200 -0.152 -0.697 claude -0.806 0.127 -0.236 herbert -0.855 -0.200 -0.212 helmut 1.000 -0.042 0.261 annette -0.042 1.000 0.127 juergen 0.261 0.127 1.000 karl 0.600 -0.576 0.091 bernd 0.103 -0.576 -0.297 karl bernd wolfgang 0.455 0.055 kai 0.697 0.297 edgar 0.503 0.539 fabrizio 0.188 0.188 claude -0.624 -0.236 herbert -0.515 0.115 helmut 0.600 0.103 annette -0.576 -0.576 juergen 0.091 -0.297 karl 1.000 0.552 bernd 0.552 1.000 Pairwise correlations in descending order 0.806 claude and herbert Significantly positive 0.794 kai and helmut Significantly positive 0.697 kai and karl Significantly positive 0.673 wolfgang and kai Significantly positive 0.648 edgar and helmut Significantly positive 0.600 helmut and karl Significantly positive 0.588 kai and edgar Significantly positive 0.576 wolfgang and juergen Significantly positive 0.552 karl and bernd Not significant 0.539 edgar and bernd Not significant 0.503 edgar and karl Not significant 0.455 wolfgang and karl Not significant 0.406 wolfgang and helmut Not significant 0.394 kai and juergen Not significant 0.297 kai and bernd Not significant 0.261 helmut and juergen Not significant 0.188 fabrizio and karl Not significant 0.188 fabrizio and bernd Not significant 0.139 fabrizio and herbert Not significant 0.127 annette and juergen Not significant 0.127 claude and annette Not significant 0.115 wolfgang and edgar Not significant 0.115 herbert and bernd Not significant 0.103 fabrizio and claude Not significant 0.103 helmut and bernd Not significant 0.091 juergen and Karl Not significant 0.067 edgar and fabrizio Not significant 0.055 wolfgang and bernd Not significant 0.055 edgar and annette Not significant -0.042 helmut and annette Not significant -0.139 edgar and juergen Not significant -0.152 fabrizio and annette Not significant -0.176 kai and annette Not significant -0.200 fabrizio and helmut Not significant -0.200 herbert and annette Not significant -0.212 herbert and juergen Not significant -0.236 claude and bernd Not significant -0.236 claude and juergen Not significant -0.285 kai and fabrizio Not significant -0.297 juergen and bernd Not significant -0.345 wolfgang and herbert Not significant -0.370 wolfgang and fabrizio Not significant -0.418 wolfgang and annette Not significant -0.515 herbert and karl Not significant -0.576 annette and bernd Significantly negative -0.576 annette and karl Significantly negative -0.588 edgar and herbert Significantly negative -0.624 claude and karl Significantly negative -0.673 wolfgang and claude Significantly negative -0.673 kai and herbert Significantly negative -0.697 fabrizio and juergen Significantly negative -0.721 edgar and claude Significantly negative -0.794 kai and claude Significantly negative -0.806 claude and helmut Significantly negative -0.855 herbert and helmut Significantly negative Comments:
Bernd: The tasting was outstanding and hardly comparable to any other tasting I joined, because of its fantastic wines. Fixing the ranking was very difficult, but the Barolo Brunate from Roberto Voerzio had a wonderful taste, full of fruit, perfectly handled with wood, and with enough acid, which all together gives this wine power and character. I feel sorry for ranking the Giacosa so low, but it was very moody and the taste differed from good to strange, but not really bad.
Wolfgang: It was a wonderful selection of the piemont cru-wines, for me only two wines were really disappointing: Bruno Rocca’s Rabaja (problems with sulfur !!) and Borgogno’s Brunate (why must a wine smell like curd???).
Karl: All wines were wonderful and worth to spend a night with, the level of the single wines were very similar and of course very high. All bottles were opened 7 hours before the tasting and I guess it would have been easier to taste and to distinguish them if they would have had more time to breathe. I share the assessment of Bruno Rocca completely, although it reminds me of burned hair (not mine). I guess that it is not due to a sulphur problem but to very hard/strange toasted barriques. But - incredibly - some people like it (Annette, Claude, Herbert). As for the Roberto Voerzio: In my opinion there is no doubt that it was the best balanced wine, with lots of fruit and a very high density. I think also that the Voerzio was one of the sweetest wines of the tasting.
Helmut: Fine selection of wines from the Piedmont. However, from my point of view Voerzio and Saffirio should be ranked clearly above the average of the rest. They are perfectly balanced, elegant and their taste reminds me of Burgundy wines. Gaja and particularly Borgogno were disappointing. Rregarding the price-quality ratio we should have in mind that Rinaldi&Filgli presents the best performance.