huevos con vino

Due to some recent increases in spam, I implemented two new Wordpress blog plug-ins that should make this blog a bit better.

First, I turned on an anti-spam comment filtering system.  Hopefully no real comments will be deleted, although I have manual control over all comments so it should be fine.  I have already blocked over 100 spam comments in 2 days with this new plug-in, so I’m happy.

I also added a plug-in that allows users to track comments posted to the blog.  We’ll see how that works!  I want to make people sign up for user accounts before posting comments because I want to see who is reading this blog and which blogs my readers are running so I can add new links.  I know the value of this policy is a matter of opinion, but I think this solution is a good one overall.

Finally, I’m heading back to southern California on Wednesday for a brief vacation.  We’re going back to the place where my wife and I had our honeymoon last September.  I look forward to drinking a whole bunch of good wine while I’m down there!

The following photo was taken during my 2005 Eastern Washington winery tour.  We were at Hinzerling, which is an interesting winery in that they focus on dessert wines and sherry, rather than table wines.  Their Web site is firmly rooted in the terroir of 1995 Web development (optimized for Netscape 2.0!) but I can assure you their wines are better than their design sensibilities.

Anyway, here’s the photo:

Doesn’t she look like she’s having fun?  Actually, she was a very nice hostess and we had a great time there.  I just loved this photo though because she seems to think we’re complete jerks.  We weren’t at that point in the day, but once we reached Terra Blanca I think all of us (aside from the 2 drivers) were cackling like, well, people who hadn’t done a 10-winery tasting in one day before.  Ouch.

As I’ve been blogging about wine rating systems and whatnot, I’ve also been thinking about the pointlessness of many textual wine reviews.  I mean, really, why do we need a score and a sentence or two?  Why not just a score telling us how that wine stacks up against similar wines made from the same grape varietals?  Or why bother with the score when you can just describe the wine in words?

Well, one problem with textual descriptions is the subjectivity of each reviewer.  Mike over at the Shiraz blog has an interesting post on the subject of palate uniformity and the calibration of your palate to that of a wine critic.  Actually, I guess it’s a series of posts on palate uniformity and the calibration of your palate to that of a wine critic.  These posts relate strongly to the subjectivity of personal taste.

Mike also points to the sort of study that I love (albeit a bit different from my idea of a single person tasting the same wine repeatedly): Tom Cannavan conducted a study in 2001 of a specific wine tasted and rated by many different people…and again in 2004 he conducted the same study with a different wine.

I think the idea of palate uniformity is interesting for a few reasons. First, I think one of the problems with writing a wine review is that pro reviewers feel the need to branch out and get creative.  You can only use fruit names so many times before every wine you review starts to sound the same (”raspberry,” “cherry,” and “blackberry” and their variants seem to be the most common fruit flavors).  So I think many reviewers take some artistic liberties with the things they smell and taste in a glass of wine.  I remember seeing “wet chalkboard” as a flavor description.  I could vaguely imagine a wine that smells like a wet chalkboard, but a wine that tastes like a wet chalkboard implies, in my opinion, that the reviewer has licked a wet chalkboard.  Was there chalk on the board, or was it clean when the licking occurred?  I wonder…because the answer might determine whether I want to buy that particular wine.

Second, I think it would be tricky to find a reviewer with whom my palate matches closely enough that I can trust him/her to be correct most of the time.  I don’t have the money to spend on even one or two bottles of $30-$100 wine, so if one critic I trust rates a bottle at 95 points and another trustworthy critic rates the same bottle at 87 points.  This is the problem: If you associate your palate too closely with one reviewer, you may miss a whole bunch of wines that the particular reviewer dislikes because, for example, he hates southern Australia and loves Napa.  Or something similar.  So why would I want to make some other reviewer’s biases my own?

Third, from the perspective of aesthetic philosophy, uniformity of palate is about as likely as uniformity of appreciate for, say, Jackson Pollock or Led Zeppelin.  Plus, tastes change over time, both for an individual and for a culture.  So my palate might match closely with Robert Parker today, but next week I might prefer another reviewer’s personal taste.  It’s the same with art: Ingres was big in his day until the Impressionists took over, and then he became a didactic old misanthrope.  Same with Pat Boone.  But then not everyone liked Ingres or Boone to begin with.

Finally, I read a lot of comments on blogs that declare the uniqueness of individual palates.  I’m not sure about that…I think there are many similarities between people’s palates.  The difference occurs in taste preference, not in what we taste, at least among novices.  Some people like the dark plum flavors of Torbreck’s Woodcutter Shiraz.  Some don’t.  But is it uniformity of palate if those two people both taste the same dark plum flavor?  Hard to say….

I checked out Wines: Their Sensory Evaluation (Amerine and Roessler, 1976) from the Chemistry Library (!) at the University of Washington. This book is not only a fascinating primer on how to enjoy and taste wines, it also has a tremendous set of empirical wine rating systems in Part II - Statistical Procedures. Hot damn.

A quick side note: I went into the Vino 100 store in Bellevue, WA yesterday. I was excited to see how they arrange their wines; they have a group of tasters who decide how fruity/dry in flavor and how light/full in body each wine is compared to every other wine in the store. They call this tool their Wine Barometer™ and I think it’s a little pretentious to have the ™ in there, but still it seems like a smart marketing ploy…and I like how this Wine Barometer allows the store to avoid relying on external numeric ratings.

Anyway, the Amerine and Roessler book is absolutely riveting for someone like me. I love the prose style of these two scientists: “Between the frankly yeasty and the obviously mercaptan there is a range of odors that resemble each other” (36). I had never thought of anything as “obviously mercaptan” before I read this book, but now I think I’ll suddenly find many thing to smell “obviously mercaptan.” It sounds so…serious.

I am up to page 38 right now, but I’m looking forward to the discussion of rating systems and whatnot that comes in Part II. For example, it warms my heart to see that my instinctive wish for a smaller number of possible points and an ordinal scale has been fulfilled (121). More interesting, though, is the Davis score card (123), where aroma and bouquet are granted up to 4 points out of the 20 total possible points.  Flavor only gets 2 points in this system; indeed, of the 10 scoring categories, 9 of them are worth up to 2 points or simply 1 point. Only aroma/bouquet gets more possible points, but that makes sense since much of what we think of as “flavor” is actually an olfactory experience that coincides with eating or drinking something.

Better still is the idea I had in an earlier comment about a golf-style scoring system where the best wines get the lowest scores. The Office International de la Vigne et du Vin Score Card (127) sets the perfect wine at a score of zero, with multiplying factors for characteristics such as odor intensity and taste quality. Interesting.

So, this book appears to be quite interesting but also quite dated (the most recent publication date is 1983) given the tremendous changes in the wine industry since the 1970s and early 1980s. Still, I look forward to reading their discussion of rating systems.

I am preparing a much larger article, or series of articles, on the subject of statistical and methodological validity of wine rating systems.  It’s a major undertaking…so major, in fact, that from my literature reviews it seems few people have tried to tackle this subject before.  Basically, my mind started churning on a simple problem: If many wine critics and reviewers agree that the 100-point system is flawed, and if there are several competing systems out there to quantify wine quality numerically, which system is the best?  When I say “best” I mean, which system is the most accurate and repeatable among many critics?

Here’s a list of the journal articles and books I was able to track down that appear to relate to my subject:

• Ashenfelter, Orley, and Quandt, (1999), “Analyzing a wine tasting statistically”, Chance, New Directions for Statistics and Computers, 12 (3) , 16-20
• Lindley, Dennis V.  (1993), “The analysis of experimental data: The appreciation of tea and wine”, Teaching Statistics, 15 , 22-25
• Amerine, Maynard A. and Roessler, Edward B.  (1982), “Wines: Their sensory evaluation”‘, W. H. Freeman & Co (New York)

Of these three, only the Ashenfelter and Quandt article seems relevant to my topic.  Happily, I discovered the Liquid Assets Web site (run by Ashenfelter and Quandt themselves).  At this site, the journal article listed above is provided, along with several other relevant articles.

Interestingly, Ashenfelter and Quandt seem to have focused on comparative wine tastings as their method of choice.  Additionally, they focus on using advanced statistical analyses to check the wine tasters’ results against one another; they developed a software program, in fact, that conducts these analyses and generates a final report automatically.  As they put it, ”The primary goal in the analysis of a wine tasting is to determine the extent to which the conclusions that have been drawn are likely to be reproduced on another occasion.”

All of this makes sense to me, but what I was hoping to find was a statistical analysis of wine rating systems, not blind tasting results.  Blind tastings tend to involve judging wines against one another, whereas individual wine tastings and ratings simply assign a score to each wine.  This score then becomes the widely printed and reprinted quantitative measure of that wine’s quality.  But how accurate is this sort of assessment?

That’s the question I want to answer using statistics, sound methodology, and survey design techniques.  I learned all about this stuff as an engineering masters student, so I know how to design a study, for example.  Because of this background, my interest in this topic was piqued when I read Robert Parker’s description of how he scores a wine.  He and I agree that the 20-point systems don’t make sense, and to that I would add the Wine Spectator’s 16-point system (85-100 points possible in printed reviews).  But I have some questions about Parker’s system too, questions that I will investigate in another blog entry in the next few days.  In the meantime, you can see how he ranks his wines here.