Fantastic Freakonomics post on Wine Descriptors

http://freakonomics.blogs.nytimes.com/2009/07/06/do-taste-and-smell-adjectives-signal-value-or-do-they-create-it/?awesm=Q5g&scp=5&sq=wine&st=cse&utm_campaign=BottleBuzz&utm_content=site-basic&utm_medium=awe.sm-twitter&utm_source=direct-awe.sm" onclick="window.open(this.href);return false;

“Two things have to be true before wine ratings can become useful for the average wine drinker. Since there are many wine writers, and there is a substantial overlap in the wines they write about (particularly Bordeaux wines), it is important that there be substantial agreement among them. And secondly, what they write must actually convey information; that is to say, it must be free of bullsh*t. Regrettably, wine evaluations fail on both counts.”

It’s quite true. With the improvement in communication - instant, now - we see how the previous models fail (and Otto, see that I didn’t originally use the word ‘fail’ in this case). Agreement would be incredibly difficult, and while it does occur on some wines, within reason, it does not on most.

I wonder if there should be an average score, that it would be the standard, rather than one particular ‘expert’s’ opinion/rating.

Okay, so I am biased against this “Freakonomics” thing. The book was an inferior, dumbed-down retread of Steve Landsburg’s Armchair Economist and David Friedman’s Hidden Order but somehow became a bestseller because… I have no idea, actually. Certainly nothing in there, or in the NYT column capitalizing on the brand, is “freaky” as opposed to just plain-old regular economics.

So, on to Goldstein’s column. The headline is, Do Taste and Smell Adjectives Signal Value, or Do They Create It? The studies he discusses in the column supports the conclusion that they do signal value. Nothing in the column deals with whether they create it, which would have been a somewhat interesting topic of study because (i) the CW is that points have a big price impact, but the price impact of any text whatsoever in a tasting note is almost non-existent, and (ii) it would probably be very difficult to study the price impact of text in a tasting note because you would need to isolate, and control for, the price impact of the accompanying point-score (and all of the other factors that affect price).

The “On Wine Bullshit” article the story links to and which Roberto block quotes is a pretty good read though…

Keith, the book was a best seller because, A) it was well written and funny and B) it applied plain old economics to subjects not normally treated by economists (for popular audiences anyway) and the results were at least provocative (Roe v Wade is responsible for drop in crime rates, not $$$$ spent on law enforcement and the fact that most drug dealers are, in fact, poor) if not shocking (the tale of the Harvard professor and the Unibomber and their backgrounds).

Get the Landsburg and Friedman books - they do the same thing but much better… and they did it first!

Friedman - “Machinery of Freedom” played a huge role in how I think. I was on a Libertarian/Techno mailing list with David a long time ago.

That said, the Freakanomics stuff applied real-world numbers to interesting subsets as well as topics not normally given economic analysis. Overlapping with the other two, but different.

What the analysis doesn’t cover is that there are obvious families of flavors. For example, when I see chocolate/mocha/coffee/tobacco I have a pretty objective idea of where the flavor is. The difference in descriptor is a matter of experience. Someone for whom chocolate is Hershey’s will taste more coffee/tobacco when hit with the same kind of notes. On the other hand, someone who regularly eats dark high-cacao chocolate will go more for the chocolate/mocha descriptors. More than overlap of wine-tasting, I would like to see sets of critics taste the same sets of items used for descriptors.

A.

The average CT score, which is now used for some wines on WineBid, may be moving in this direction.

A friend (who’s also in my wine group) published a book last year on randomness, called “The Drunkard’s Walk”, in which he devoted a chapter to wine ratings and their relative worthlessness:
http://books.google.com/books?id=UJxRLCq9l3IC&pg=PA131&lpg=PA131&dq=drunkard’s+walk+wine+ratings&source=bl&ots=5RcmDwIy1W&sig=FFnD9aFc3UM84LJajttA3edg_jQ&hl=en&ei=lOdSSpGoGsa5jAegocWcCQ&sa=X&oi=book_result&ct=result&resnum=1" onclick="window.open(this.href);return false;

The wine section starts on the bottom of page 130, but you’ll need to buy the book to see how it comes out…