The American Association of Wine Economists (what a job) met in Reims last month. Robin Goldstein, of the New York Times Freakonomics blog, picked up some intriguing research presented to the learned audience.
Coco Krumme (what a name) has been investigating (with a Bayesian filter and all) the relationship between the price of wines and the taste and smell adjectives used to describe them. Her findings will ring true with anyone acquainted with wine labels and reviews – a wine that is “elegant, smoky and suggestive of tobacco” is likely to be more expensive than one described as “lively, pleasing and fruity”:
| Words predicting expensive wines | Words predicting cheap wines |
| Authenticity or exclusivity, full flavour: old, elegant, cuvee |
Accessibility, lightness: pleasing, refreshing, value, enjoy |
| Dark: intense, supple, velvety, smoky |
Light: bright, light, fresh, tropical, pink |
| Specific: tobacco, chocolate |
General: fruity, good, clean, tasty, juicy |
| vintage | harvest |
| Pair with: steak, shellfish |
Pair with: chicken |
More interesting is the question of causality:
Causality, of course, is unclear: is it that more expensive (elite, specific, dark) flavors command higher prices, or that expensive wines are described with a verbiage discrete from that used for cheap or “value” wines?
And of more interest still to the ordinary wine buyer might be the question of whether expensive words on cheaper wine are indicative of quality at a bargain.
