New Scoring Method adopted World-Wide.

Categories: Latest News


Dan Berger’s enmity has not waned toward the scoring of wines on a single-number, linear basis.

Fortunately, the real meaning of this silly affectation is declining as more and more wine buyers seek to find ever more interesting, unique, and unusual styles of wine.

As the importance of numbers on wine declines, it struck me that one of the key reasons for my belief in the utter failure of scoring to describe wine quality was due to the very reason it was a success for as long as it was. And that was: it was simple.  All anyone needed to remember was the brand name and a single numerical reference point—which meant that such other things you had to recall, such as the vineyard designation, the vintage, and other such elements, were unimportant. What was important was the 94, or 96, or 99.

In theory, it would have been rather simple to create a “better” system that could reduce a wine to a series of numbers that, if explained in full, could also be a shorthand way to describe wine quality that would be far more meaningful. So I did.

The problem here is that not only does my new scoring system look more complicated than the old template that uses single numbers, but mine requires users to actually understand a little bit about what the numbers mean. For instance, when you see a score of 92 for a red wine, which once was a pretty good score (it isn’t any more), it tells you nothing about whether the wine is one that adheres to an old world standard of structure and is based on acidity. Or is a wine that is porch-related: It has enough oak to build one, enough alcohol to fall off one, enough tannin to sand one, and enough color to paint one.

With my new system of scoring wine we have to remember more than single numbers. But what you gain in this scoring system is (a) it is a real system with real parameters (not just someone’s off-the-cuff gut reaction), and (b) it gives far more information. Let’s start with the fact that if a 100-point scale is valid, then all numbers ought to be used. So instead of the current employment of numbers that, from a practical point of view, range from 60 to 100 (thus a 40-point scale),we can fine tune things a lot finer. In a 40-point scale, the difference between an 85 and a 95 is 12.5%. In a true 100-point scale, the difference between an 85 and a 90 is only 5%. As you can see, there is a lot more fine tuning here.

Plus, wouldn’t it be fun to find a few wines that are not flawed at all—wines that are technically clean—but which are stylistically so vacuous that they warrant scores of 14 or 17? Second, we must have additional numbers that describe elements that are meaningful in terms that consumers should know if they are to get the full meaning of the new score.

So here is the legend I have created to set up this new system: The format of scores for all wines look like R; S; T; V, which are defined in this manner:

R: The Raw score that defines what the evaluator believes to be an approximation of the wine’s overall quality. Scores from 1 to 100 are to be used.

S: A Style rating based on the numbers 1 through 5, with 1 being light and delicate and 5 being a brute. (A good “score” here would be a 3.)

T: A Terroir score based on the numbers 1 through 5, with 1 being faint terroir and 5 being strong terroir component. A 5 would be appropriate for wines from single vineyard sites in assertive, distinctive regions. A score of 1 is assumed for all wines with broad appellations (such as California, France, South Africa, or South Eastern Australia).

V: A Varietal score based on the numbers 1 through 5, with 1 being inappropriate (no varietal character), and 2 so light it is clearly not a positive rating. A 4 is desired here as a way to state how varietal the wine is. (A Sauvignon Blanc that smells and tastes like Chardonnay would get a 1; a light varietal aroma is a 2.) So a classic Russian River Valley Pinot Noir that is exemplary, lighter in color, has distinctive flavors that are appropriate to the kind of wine it is would be rated 96; 3; 4; 4. A wine that is heavier and darker might still be scored well (say its first raw score would be 90), but its style is awkward for the variety and it thus has little regional character. So it would be a 90; 5; 1; 2. See how easy it is to assess the quality of a wine with this system? A Napa Cabernet with no finesse at all, low acidity, and no varietal aroma could still score as high as an 87, but its other scores would be 5, 1, and 1. Now, would you buy a wine that scored 87511 (I have left out the semicolons)? I’d sooner have an 85334; suits my style preferences a lot better.

Now, obviously we have only begun to mine the numbers ideas. How about adding additional fields for alcohol (1-5, low to high); oak (0-5 with the former reserved for unoaked wines), and a reflection of the vintage (1 to 5). Can you imagine a Pinot Gris that would be interesting to some people if it scored 89334213?

Hey, wait a minute, I think we are onto something here. How about adding fields for length of time from the vintage, use of cork or screw cap, color of the wine. I highly recommend this wine. It is a 953342554213.

Looked at another way, some might guess that this article was linked to next Monday, April Fool’s Day.

Dan Berger’s Vintage Experiences