The Ratings Game
By Alexis Beltrami

Should wines be rated? Keats warned that science would empty the poetry out of nature, or "unweave a rainbow." Many people in the wine world feel, similarly, that it is deplorable to reduce wines--which ideally are unique expressions of grape, place and winemaker--to numerical scores. Nevertheless, the business of evaluating wines, usually on 100-point scales, is booming. Wine Spectator, The Wine News, Wine Enthusiast, Robert Parker's Wine Advocate, Steve Tanzer's International Wine Cellar, et cetera--there's no shortage of confident quantifiers out there, whose ratings are lapped up by a public thirsty for guidance.

There are many good reasons to object to the ratings game. First of all, there are no absolute standards of quality in wine; taste is subjective. While one taster may be bowled over by richness, intensity and exotic flavors, another may be seduced by complexity, depth and elegance. Is one standard superior to another? The professional tasters are seldom up-front about their biases, which we can only infer over time.

Then there's the familiar apples-and-oranges problem. Is it fair, or even meaningful, to weigh Beaujolais and Barolo on the same scale? The two red wines are radically different in nature and intended to fulfill different needs. And yet, because the dominant assumptions that underlie wine scoring favor the Barolo model, no Beaujolais, no matter how sexy and fabulous and perfect, will ever score 100 points. (If someone out there can show me a Beaujolais that has scored 95 or higher, anywhere, please let me know.)

Perhaps the most annoying feature of the 100-point scale is its illusion of precise and scientific measurement, which causes unwary consumers to place undue faith in the numbers. Never mind the fact that equally authoritative reviewers (sometimes tasting for the same publication) routinely differ in their appraisals by four to eight points-- a huge spread, given that the 100-point scale is actually 50 points, and, in practice, more like an inflated 20-point scale, in which anything under 80 isn't worth drinking (like the children of Lake Wobegon, all wines apparently are above average).

The most important objection of all, however, may be that numerical ratings drain the poetry out of wine. Hard, cold numbers reduce something ineffable into a mere commodity, subject to scientific quality-control standards. Great wines, and even good wines, can be like works of art, expressing individualistic personalities, changing over time, speaking differently according to the moods and experiences of particular tasters. Would we dream of scoring paintings? Perhaps Botticelli's Venus rates a 98, but Rembrandt's Polish Rider only a 92?

Now, having said all this, I must declare unequivocally: wine ratings are helpful, useful, even necessary tools. They are, on balance, a good thing. Why?

Most wine retailers, in my experience, cannot be trusted. Not that they are intentionally dishonest (although I've encountered some suspect ones), but they simply are not disinterested parties. In contrast to independent blind tasters, retailers have all kinds of good business reasons for pushing one wine over another. Of course, any good retailer wants to earn his customers' trust, and will try to recommend a wine that he thinks you will like. But since most wines in the marketplace today are good (but not above average!), that's a fairly low-risk proposition. The problem is that the recommended wine is inevitably either more expensive, or slightly less good, or both, than what you came in looking for.

There's also the mystique factor. Certain wine regions, producers, and even grape varieties have inflated reputations, which result in outrageous prices that are seldom justified. Numerical wine ratings can do a splendid job of cutting through the hype, prejudice and snobbery. Not to mention marketing: ratings consistently--and embarrassingly--demonstrate that many wineries' "reserve" or prestige bottlings are no better than, or marginally better than, their much cheaper basic releases. In practice, reality-check ratings do not seem to have had much of a braking effect on escalating prices, but at least we now can see when the emperor has no clothes. We can choose to let other fools part with their money.

On a more positive note, high ratings can bring welcome attention to previously obscure or undervalued wines and regions. The high scores recently earned by New Zealand Sauvignon Blancs, Argentine Malbecs, and southern Italian wines will, I'm confident, translate into a significant impact in the marketplace-- just as the high scores awarded to Tuscan and Piedmontese reds beginning in the late 80s helped elevate those wines to the status of Bordeaux and California Cabernets.

Finally, the success of ratings-based wine publications, like that of the Zagat restaurant surveys, suggests that they are answering a public need. The wine world is vast and rapidly changing, and it is understandable that busy people who wish to enjoy wine will seek impartial guides to get them through the wilderness, even if the guides are not perfect.

The wine publications need to do a better job, though, of contextualizing their scores. And the 100-point scale, with its misleading implicit claims to accuracy and objectivity, should go. In its place, I recommend a letter-grade scale: immediately understandable by anyone who has been to school, it allows for broad discriminations while avoiding the pretenses and fetishism of the numbers.