by Ben Rotter » Wed May 21, 2008 9:11 am
Given the variety of influences on tasting assessment (e.g. environment, mood, experience, temperature, prediction of development, etc) I actually think that the degree of agreement is quite amazing. If you are using other people's TNs as a quality assessment of the wine in question then there's definitely value to be gained from the text, particularly when you understand how the author's taste relates to your own. I think any reading of someone else's TN should always be read with reference to both our own preferences and that of the taster's (in as far as is possible).
I think the value of a numerical score is when we compare scores given by the same author that compare wines of the same styles/appellations/same-producer-but-different-vintages, since a 93 over a 90 then has more meaning. Parker seems to have one of the more consistent palates among the critics, but I agree that even then the consistency of his accuracy probably has a resolution of +/- few points (BTW, if you want to be analytical and pedantic about it, then statistically analysing Parker's scores would probably give more insight on features like this than anything else available. For example, his scores exhibit significant compression in the 85-92 range, so a wine scoring <85 or >92 has significantly more "meaning" in terms of it's score (the wine being damned or exhalted) than wines scoring within the 85-92 range).
Surely the best use of scores involves underpinning the score information with our own interpretations, using what we know about the author's tastes, the methodology of the scoring, etc to better inform us about what that score really means. If we don't do that then aren't we reduced to "score whores"? As David pointed out, different people have different approaches. Doesn't that just make the world a richer place? These days, when I use others' TNs and scores to inform my buying decisions, I always try and use multiple sources. I think my buying decisions are all the better for it.