Wednesday, August 12, 2009

A nose for nonsense

Wine ratings should be viewed with extreme suspicion. This is just one of the many useful lessons to be gleaned from The Drunkard's Walk, How Randomness Rules Our Lives. Leonard Mlodinow's purpose is to explain to non-mathemeticians why things we accept as "true" in modern life are frequently wrong.

Wine drinkers love numerical ratings and awards. But how can it happen that the 1999 Mitchelton Blackwood Park Reisling earns five stars and is named Best Wine of the Year by The Penguin Good Australian Wine Guide, while the same plonk was deemed "worst vintage produced in a decade" by the competing On Wine's Australian Wine Annual.

Mlodinow explains that subjective factors (influence of other judges, influence of price, the complexity of taste perception, expectations, etc.) can produce a wide range of ratings across a panel of judges. When subjected to an objective test in 1990, 1/3 of the time "experts" were unable even to identify which wine of three sampled was not identical to the other two!

In a 2008 study, in which a bottle labelled $10 and another labelled $90 were both filled with the same wine, the $90 bottle scored higher.

Surprise, surprise.

No comments:

Post a Comment