3 ms·

A chilled red wine tastes and smells completely different from a non-chilled red wine. Go ahead and try it yourself. Also, wine tasting is absolutely a narrow

A chilled red wine tastes and smells completely different from a non-chilled red wine. Go ahead and try it yourself. Also, wine tasting is absolutely a narrow classifier trained on limited data (in this case, tasters are not trained to classify red wine that is chilled). It's not surprising that wine tasters have overfit their classifier to the types and conditions of wine that they have been trained and tested on.

caconym_ 8y ago
> A chilled red wine tastes and smells completely different from a non-chilled red wine. Go ahead and try it yourself I believe this, but it's not what I said. I also assume a room temperature white smells and tastes completely different from a chilled one. If there is nobody in the world who can, with significant accuracy, sort a set of reds and whites held at the same temperature into a set of reds and a set of whites, then it is reasonable to say that a lot of the conventional wisdom around wine tasting is probably wrong. If wine tasters are too far up their own butts, use regular people instead, or people with more generally refined palates (e.g. chefs). Train somebody specifically for this task, even. At this point it's a hypothetical exercise. Claiming e.g. "only trained wine tasters are qualified to make this distinction, and serving wines at 'incorrect' temperatures makes it impossible for trained wine tasters to tell the difference because they've 'overfit their classifiers'" is absolute hogwash.
- cycrutchfield 8y ago
  The problem with using regular people or chefs is that most cannot distinguish a merlot from a cabernet, so why bother asking if they can distinguish a chilled red from a white? Basically what this boils down to is making inferences from these experiments is meaningless unless we actually have some people train to distinguish between chilled red and white wines, and then evaluate whether they are able to do so reliably in a double-blind setting. But I imagine the results will end up being "yes" and that's not really going to drive clicks because it doesn't agree with the anti-wine snobbery circlejerk.

caconym_ 8y ago

> A chilled red wine tastes and smells completely different from a non-chilled red wine. Go ahead and try it yourself I believe this, but it's not what I said. I also assume a room temperature white smells and tastes completely different from a chilled one. If there is nobody in the world who can, with significant accuracy, sort a set of reds and whites held at the same temperature into a set of reds and a set of whites, then it is reasonable to say that a lot of the conventional wisdom around wine tasting is probably wrong. If wine tasters are too far up their own butts, use regular people instead, or people with more generally refined palates (e.g. chefs). Train somebody specifically for this task, even. At this point it's a hypothetical exercise. Claiming e.g. "only trained wine tasters are qualified to make this distinction, and serving wines at 'incorrect' temperatures makes it impossible for trained wine tasters to tell the difference because they've 'overfit their classifiers'" is absolute hogwash.

cycrutchfield 8y ago
The problem with using regular people or chefs is that most cannot distinguish a merlot from a cabernet, so why bother asking if they can distinguish a chilled red from a white? Basically what this boils down to is making inferences from these experiments is meaningless unless we actually have some people train to distinguish between chilled red and white wines, and then evaluate whether they are able to do so reliably in a double-blind setting. But I imagine the results will end up being "yes" and that's not really going to drive clicks because it doesn't agree with the anti-wine snobbery circlejerk.