Friday, August 10, 2007

How Do You Standardize Taste?

A couple of weeks ago Tim Elliot at suggested a standardized wine rating system for bloggers. Tim recommended a 5 star system (with half stars) basically breaking down a wine into horrible to classic. You can read his proposal here.

When Tim made the recommendation, it made sense to me. Web 2.0 (and Tim's Wine 2.0) is all about displaying information in new and interesting ways, and having standards is paramount to that. Bloggers will continue to grow as a source for wine reviews, and having a way to share those reviews with little effort would be great.

Imagine if every review that comes from every blogger, every publication, everything, ended up on Snooth (with appropriate credit of course). Wouldn't it be amazing to have a portal where you could see everything anyone had ever written about a wine, condensed down to a single page of information? Scores would be aggregated (maybe weighted on trustworthiness) and presented as the world's collective opinion on a wine. Reviews would be piled high giving you a single stop to read what your trusted critic, favorite blogger, and your next door neighbor all had to say about the wine. Sure that's a little scary if you are a producer making second rate wine, but for the consumer, it is an incredible resource.

That's an amazing vision, and one that I hope Snooth (or another wine site) is able to realize. The problem is, how do you standardize a review. How do you take an opinion of value, quality, reputation, and uniqueness and boil it down to a standard that everyone can work inside? At the time Tim made his proposal to the blogosphere I commented:

The inherent problem I see with a standard scoring process, is that it leaves out the uniqueness of each reviewers system. For instance, I may want to include some sort
of a value quotient and a score. With this system I can’t do that.

When I wrote that, I hadn't even considered the possibility of unconventional reviews. Then today I saw two posts from Fermentation and Winehiker about Chateau Petrogasm. This is a novel and thought provoking new way to review a wine. Images that represent a wine, it borders on genius. I especially like this one. If a picture was ever worth a thousand words, it's this one.

I'm torn though, because I can't figure out how to quantify it. If wine reviews become standardized, then what happens to reviews like this? Where is their home? The IT industry has long argued that standards stifle innovation. Anybody whose idea doesn't fit inside the standard shrivels and dies. Would the wine community lose innovative new ideas like this if we were standardized?

I applaud Tim for his standard. Heck I more than applaud it, I use it. I think it is a bold step in the right direction for creating an invaluable resource for consumers. I just hope that someone smarter than myself can figure out a way to fit fringe creativity like Chateau Petrogasm into a standardized mold. After all, while the Internet is all about providing better, faster, bigger information, it thrives because it is full of creative people.


winecast said...

I think another aspect of my proposal that has not been discussed much is blog readers can use the same rating system. Since bloggers and readers carry on an online conversation having a normalized rating system makes sense. Particularly in the context of Wine 2.0 sites like Snooth as you point out.

Thanks for joining me rating wines with 5 stars.


Arthur said...

Integral to standardization is not agreeing on what system of communicating the overall impression of the wine (the final score/rating), but how one arrives at the score.

Stars, moons, puffs, points etc, mean nothing without context and they mean little when they represent personal opinions of qulaity.

Any system -stars, 5, 10, 20 or 100 points become more meaningful when the reader knows how the score or rating was derived.

That is one aspect of context. The other is the seemingly unavoidable one in today's wine writing - amateur and professional (WE, WA, or WS) alike: more often than not, a score represents the taster's enjoyment of the wine. This is not often rooted in the fundamentals of what makes for a well made wine or what is the best example of a wine form given region and year - what it's components and character should express.

razmaspaz said...

I think that's really the point. If we can't agree on how to arrive at a score, then a score is meaningless as a standard. You may decide that a wine is well made and so deserves 4 stars. I might decide that it tastes like mangoes (and maybe I don't like mangoes). so I give it 2 stars. There has been significant debate on rating systems, and most reviewers agree that it is the notes that matter, not the score.
As far as your comments on a wine being an example of a region/year...There is a valid argument out there that those things shouldn't matter. If a wine is free of flaws and tastes good, why shouldn't it be highly rated. I agree that there should be some component of correct terroir in a professional review, but thats because I put stock in a wine's ability to provide an intellectual pleasure as well as a sensory pleasure. There are many wine drinkers though, who don't care about this, and for them it is not going to matter that the wine tastes like a Russian River Pinot from an especially cool year.

Arthur said...


I think we are on the same page (and team) when it comes to adjusting for variations of vintage: If a year was crappy and the wines were lean, or the year was hot and the wines all lack acids (despite post-harvest manipulation), then after evaluating or assessing the wine in and of itself, there needs to be some accounting for the fact that the rest of the wines from that year/region were iffy. Sort of like grading on a curve.

As for context of terroir or vintage: I agree with you but allow me to elaborate: Joe Schmoe Winedrinker may not care for or seek out regional distinctiveness - at least not consciously. However, they may have a preference for one style of wine versus another - say a juicy, fruity cab with minimal tannins - like some hot climate, East Paso Cabs and not a more demure and possibly vegetal Cab from, say Arroy Seco in Monterey County. If we go with enjoyment-based scales, then when Joe Schmoe, above, gives 5 stars to the Paso Cab and you really like the Arroyo Seco type Cabs and hate fruit bombs, we are back at square one: discrepant ratings.

This is why I said in the last section of my comment: that personal enjoyment of a wine should be removed from rating. We can still use scores or stars, but they should mean something concrete and rather absolute instead of the degree to which the reviewer enjoyed the wine and how likely they would be to recommend it.

I agree that the descriptors are key. But that is a can of worms in itself. I have seen wines described as: having “luscious aromas of cabernet grape”…Given the range of styles and climates and the resulting variations in expressions of aromatics in a particular grape variety, this is not an informative description. Then, there are reviews attributing aromas to a variety that are atypical to say the least, and possibly even undesirable. The author then goes on to attribute those to the expected characteristic of the wine, saying: “that is just what you would expect from a _____”. That’s like having a duck waddle in front of you, and saying: “Look at those stunning antlers! That is just what I expect to see on a mature dog.”

I think it's great how more and more, younger and younger people are getting into wine and the ensuing discussion. However, what I think that those who want to write about wine and those who want to benefit from written communication about wine should educate themselves about wine. Particularly, we should learn to correctly identify aromas as what they are. It’s more than consistently and reproducibly calling a duck a duck (that will allow us to communicate clearly and understand better the taster’s experience of the wine and help define in our minds what we like in a wine of a given variety or origin). It’s knowing that sharp, melon-like or acetone-like aromas might indicate Volatile Acidity - a flaw in wine production, and thus a poorly made wine. This may also point to the possibility that another variety was blended into the wine to fluff it up. Yes, some people like wines with flaws – whether they can detect and recognize them or not – but if we want standardized rating (which inherently includes recommendations) we should all have solid familiarity with the essential aromatic characteristics of the component grapes and those contributed by cooperage and aging. Then, we will make better recommendations or at least write better summaries of the wine we taste – neither based on our enjoyment of the wine, but on the merits of the wine itself.

Unfortunately, some things are not easy and some things are not simple. It may be easier (and cause less conflict) to say: “everyone’s perceptions of what’s in a wine are accurate, true and valid”. However, that would require a few rearrangements of the universe. Yes, there are arguable nuances in a wine, any wine. However, there are core characteristics that each grape (or production method, meticulous or sloppy) brings to a wine (with some variation due to weather, farming, region, barrel, methodology etc) that cannot be argued – just as there are a set of distinct flaws that are attributable to compounds which are invariable.

I know this may take the magic out of wine for some, I know it may spoil the mystery and shatter images of the winemaker as artist-alchemist. And it sure flies in the face of individualist notions of uniqueness that many (erroneously) extrapolate into the realm of human sensation, perception and neurophysiology. The logic of “I smell/taste A, B and C and you taste/smell X, Y, Z in the same wine” (which leans way too heavily on notions of supertasters and sometimes misinterprets what the research in that field really tells us) seems to disregard the fundamentals of chemistry and human physiology and suggest somehow that the same wine changes its components and properties for each taster.

Despite all the differing opinions about sensory perception in the blogosphere, these things can be learned like a language. Although some want to believe otherwise, the single molecule responsible for [insert aroma/flavor here] always triggers the same smell/taste receptors in all homo sapiens – much the way that typing “dir” (in DOS) makes your computer do one and only one thing and the same way that flipping the light switch won’t make the toilet flush.

One final thought: If we actually do employ an absolute standard based system of rating (one that requires accurate identification of aromas, flavors and textures) it inevitably will make the process of assessing and rating wines more complicated. I still think that this can be translated into a simple 5 star scale. Just not one indicating personal enjoyment.

razmaspaz said...

There is a lot there to disect. I'm out of town and out of touch the next 5 days. I'll take some time to really read through that and respond when I get back.

Arthur said...

Sounds good.

Feel free to contact me via the contact option on my website.

razmaspaz said...

So liek I said, there is a lot there. First off, I think it is important to draw a distinction between a criticism and a review. I think when you make comments like, "that personal enjoyment of a wine should be removed from rating. We can still use scores or stars, but they should mean something concrete and rather absolute instead of the degree to which the reviewer enjoyed the wine and how likely they would be to recommend it.", you are asking for a criticism, not a review. I totally agree that a criticism needs to analyze the wine for flaws, regional character, and varietal correctness. That is totally fair.

When you review a wine, that is something different. A review should indicate a recommendation of the wine. Do you like it? What does it taste like? Is it a good value? Those are important components of a review.

Then there is the contingent of "reviews" like chateau petrogasm. If you are calling for a standard review system, are you saying those are invalid or somehow improper reivews?

Arthur said...

I think the words “review” and “critique” mean the same thing ( and An explicit statement of endorsement or recommendation is not inherent in either of those. Neither is it in "criticism", which also gets us to the same notions: (
So two (three) terms are really interchangeable and it would derail the crux of the matter at hand to get mired in nuances of semantics. Ultimately, the content of the review/critique/criticism serves as an endorsement or admonition – whether you say that you recommend something (and how enthusiastically) or not.
In clarifying that comment about removing personal preference from wine evaluation, I propose to you that one can be objective about wine description:
Think of a food that many others enjoy but you just hate. You could learn to assess it on its own merits and convey to someone what the food is like (using objective terms) and if the particular example before you is a quality one or not.
Another scenario: you know a girl who just rubs you the wrong way, but you are able to acknowledge that she is not some evil ogre. Rather, she is a normal, healthy person but you and she just don’t mesh. However, she has a lot of the qualities that would make your friend very happy in a relationship. As a matter of fact, you really think this is the right girl for your friend (about whom you care a lot and want the best for). Putting your prejudice for the girl aside in order to convey to your friend what she is like requires the same mindset as rating or reviewing a wine objectively. You could also do that without interjecting (or divulging to the listener) your preferences.
Again, a recommendation may be stated but it should not be based on your enjoyment of the wine.
This is where the skill in being a critic or reviewer of anything comes in. I don’t have to like Puccini to recommend a local production of Turandot or to know if it is a good one. I do have to know something about opera and what the challenges of a Puccini opera are and what elements make a production of Turandot a success. This example goes on if you substitute Rock ‘n Roll or Shakespeare, Adam Sandler or whatever in place of Puccini.
As Karen MacNeil says: " You may not like reading Shakespeare but agree that Shakespeare was a great writer nonetheless".
I think that chateau petrogasm has an interesting angle. It may work for some as buying only wines with scores over 90 points from PR does. It doesn’t tell me a thing about the wine, though. “But a picture tells a thousand words”, no doubt many would argue. I say: “Yes, a thousand *different* words to a thousand *different* people. I want to hear a description of the components of the wine. Then I will make my decision if I want to try the wine. Please see paragraph 4 of my Sept 6th post. For what I need to see and what I feel a good review should have. Accurate and true descriptors (despite the popular “we-all-perceive-different-things-in-wine” philosophy) are consistent and can be learned, see paragraph 5.

Arthur said...


In paragraph 5 of the Sept 6 post I say: "…but if we want standardized rating (which inherently includes recommendations) we should..."

I said "inherently includes recommendations" for two reasons: because, at the end of the day, we want these written descriptions to convey to the reader whether the wine is worth buying and drinking AND because when a reader sees a group of descriptors that appeal to them, the content of the write-up becomes an endorsement of recommendation and it becomes important to the writer’s credibility to make accurate observations about the wine at hand.