wa2.gif (4241 bytes)

abut9.gif (3095 bytes)

abut12.gif (3207 bytes)
abut10.gif (3636 bytes)

abut11.gif (4039 bytes)

Scoring wines: does it measure up?

By Jamie Goode

There are few subjects as controversial in the world of wine as the practice of assigning numerical scores to wines. While debate about this practice is heated in Europe, it is raging even more fiercely in the USA, where these scores, when assigned by influential critics, have a huge impact in the marketplace. Americans seem to love statistics. Take golf as an example. In contrast to their European counterparts, American golf fanatics will avidly keep track of a golf star's statistics, such as their average drive, their GIR (greens in regulation) or their puts per round. They do a similar thing with baseball and American football. Measuring things facilitates objective comparison -- it lets you work out what is 'the best'. So it is understandable that they should want to do the same with wine. Perhaps the most famous scoring system is the 100 point scale devised by US lawyer-turned-wine-guru Robert Parker. This scale starts at 50, but only relatively poor wines score less than 70. You can be sure that any wine he scores in the mid-90s or above will be in such demand by his devotees that it will sell out fast, and the subsequent vintage will probably sell at an increased price! Similar 100-point scales have also been adopted by the influential Australian writer James Halliday and the US magazine 'The wine spectator'. But there are also other scoring systems that have their advocates, such as the UC Davis 20 point scale (and its derivatives, one of which is very popular in Australia). This scale breaks down the characteristics of a wine into categories such as colour, nose and palate, in an analytical and pseudoscientific fashion.

Fundamental objections to numerical ratings of wines
Personally, I have several problems with the practice of wine scoring:

It doesn't have a basis in reality
Scoring wines has its basis in the flawed assumption that wines have attributes that can usefully be summarized by means of a numerical rating. While there are some aspects to a wine that can be measured such as its residual sugar, optical density or titratable acidity these figures tell us relatively little about what the wine is actually like. What a wine score is really made up of is an agglomeration of value judgements by the taster. Returning to the golf analogy I began with, the statistics connected with golf are useful and provide an objective assessment of the performance of each professional golfer. They are based on measurements that have a good degree of precision; measurements which would be almost exactly the same when measured by a dozen independent measurers. Thus they are based in reality and allow us to say who the 'best' golfer is. In contrast, value-judgement-based wine scores have their basis on quality judgements by taster and will likely differ between even experienced tasters. In some ways scoring wines is like wandering up and down the National Gallery giving marks to all the different paintings: a rather pointless and silly task.

People see scores as an attribute of a wine
Part of the problem with scores is that people see them as an attribute of the wine. For example, it is common to hear people talk about a ninety-something-point Parker wine, as if that particular bottle owned that particular score, in some sort of invariant manner. What a 95-point Parker score really means is that on one particular day, Robert Parker thought that bottle of that wine deserved a rating of 95 points. To you, with your palate, in the particular environment you happen to be drinking the wine in, given your context, things could be very different.

It is ugly
Some of the objections to scoring wines are quite emotive, but in their own way they are still valid. Many people see the practice as ugly or undesirable in an aesthetic sense, and I'm inclined to agree. When it comes to matters of personal taste or pleasure, a score is a quite inappropriate way of assessing things. There is something cold, mathematical and calculating about it. Most people would think it quite odd if I were to measure the beauty of a sunset, the joy of a holiday or the depth of my friendships on a 100-point scale. In the same way, I appreciate wine in an aesthetic and artistic sense, and find it increasingly foreign to try to reduce this to a number.

It ignores palate differences
There's good evidence for discrete differences between the palates of individuals (a topic I cover at length in another feature). If this is not taken into account, then it makes a nonsense of the idea of wines 'having' scores. Besides these discrete palate differences which are genetic in origin, there are also less defined but equally pronounced differences in preferences. For example, while we may all agree that Chateau Latour is a 'better' wine than Gallo Turning leaf Cabernet, would we all agree that it is 'better' than Chave Hermitage, or Penfolds Grange, or Chateau d'Yquem? The concept of the 'best' wine is highly dubious here.

Is it ever justified or useful?
Despite my strong personal objections to scoring wines, occasions do exist where I think it can be justified, and even useful. First I'll add a qualifier, though: the only permissible basis for scoring wines is as a shorthand way of saying how much you like a particular wine, under the conditions you tasted it in. That is, if I give a wine a score of four stars out of five, this means that I really liked it -- and liked it more than a wine I scored three stars out of five, for example. In the same way, when you see a wine with a Parker rating of 90, you can bet that it's one he really liked, and, depending on how well your palate correlates with his, you may like it too. And I do find that Parker ratings have some utility, because the scores are all assigned by him personally and he has a pretty consistent palate. But I have little regard for scores that are awarded by committees or tasting panels.

In summary
There are some severe objections to the numerical scoring of wines: they lack a basis in reality, they ignore individual palate differences, and the practice is fundamentally an ugly one. Despite this, there are limited circumstances where a score can be useful: most particularly it can rapidly communicate how much a taster 'liked' a wine in a qualitative fashion. Overall, though, I think numerical wine ratings are unhelpful.

(see also: Nick Alabaster: our regular columnist's article on 'Rating wines', and My rating system explained.)

Back to top