Blockquote What they found: The group does a statistical analysis of every group of wines it tastes, and has found, with one notable exception, that they cannot “determine dependable differences in wine quality,” per Malkiel, even when they’re drinking such spectacular wine as Mouton Rothschild or Romanée-Conti. The one outlier came in a tasting where a 1995 Dom Perignon was clearly preferred to “a $2 bottle of Russian sparkling wine bought at a Moscow airport.”
I’m going to say I know of tasters who I would bet on to get it right more often than not.
I will go further and say that many out there can peg the region a wine comes from and the grapes in it more often than not (>65% accuracy?).
Lowering the odds a bit, some will even nail specific producers and fewer still the vintage.
Everyone make similar claims. In the end, the only thing that will truly settle this is take about a hundred berserkers, lock them in a large well-sourced cellar featuring at least 10k bottles covering the major old and new world regions over the last 125 years, and wait until we either reject or fail to reject the null hypothesis.
Sounds like after 40 years of tastings they realized that getting together with friends and having a great time together was the point all along. When I saw the math equations in the article I almost gave up, glad I read to the end.
Looking at the summaries of this group’s tastings, the conclusion is not overly surprising. The vast majority of the themes are (a) of like in kind wines, e.g. by variety or region, and (b) generally are of a very good to excellent qualitative level.
Above a certain qualitative threshold, it’s generally going to come down to taster preference and the state of each bottle. Which is to say, the difference between a $100 and a $200 bottle, both from great producers in a good or better region, is almost always going to be personal preference. Same principle when comparing multiple wines with consensus ratings 95 to 100.
It’s quite likely that if the sampling was more heterogeneous, there would be more statistically significant spread in preference. Especially if including more basic, commodity wines as ringers against those produced from high quality terroir by quality focused producers.
It’s an interesting paper indeed - and their ‘statistical analysis’ is commendable . . .
But we have no idea about the palates of each of the tasters and differences that exist between them. In addition, because they ‘verbalize’ their scores rather than write them down to be tabulated, there is always the risk of ‘bias’ and folks changing their positions on wines at that time.