ENGLISH writer Neal Martin, hired by the world's most influential wine critic, American Robert Parker, as his ‘critic-at-large', visited a well-regarded South Island winery a few months ago and awarded its 2007 pinot noir a rating of 91 out of 100. In the Parker world, the score indicated "an outstanding wine of exceptional complexity and character". Later, in a press release, the winemaker declared he was delighted with the rating, "particularly for what was only a barrel sample".
A barrel sample? The whole idea of critics tasting raw, unfinished wines, then reviewing and rating them, has recently and rightly been under fierce attack in connection with the Bordeaux en primeur system. En primeur is a French wine trade term for wine which is sold as a ‘future' - before it is bottled. Every spring, the great estates of Bordeaux offer barrel samples of the previous year's harvest. Members of the international wine trade and wine media descend on the the world's most important red wine region to assess the samples and the wines are offered for sale - two years before they are released - to wine brokers in Bordeaux, who then sell them to wine merchants around the world, who in turn offer them to their customers on a cash up front, delivery later basis.
English wine writer Stephen Brook last year slated the en primeur system as a con trick, perpetrated on a gullible public by wine merchants, chateaux and journalists. The author of Bordeaux: People, Power and Politics, Brook ridiculed the way wines that haven't even been blended are tasted, reviewed and sold. "It's like Karl Lagerfeld presenting a sketch to clients and saying, ‘I haven't decided where to put the buttons, and may change the colour, but you get the idea. Now please give me your money'."
English wine authority Jancis Robinson strongly supported Brook: "Wouldn't it be nice if we didn't have to play this game?" So did Howard Goldberg, wine columnist for the New York Times: "Why should wine journalists and merchants, who lack knowledge of behind-the-scenes cellar tinkering, trust that tank and barrel samples are the real, final thing?"
A top French winemaker also recently attacked en primeur, dubbing it "a media creation". Jean-Claude Berrouet, who worked at Chateau Petrus - the merlot for millionaires - from 1964 to 2007, argues wines should be judged over time, rather than in "a moment of madness". It is equally hard to understand why critics in the US, Australia and New Zealand score wines out of 100, when the implied degree of precision is fantasy.
In his Wine Advocate newsletter, Parker allocates scores ranging from 50 (for "the most repugnant of wines") to 100 (for "the most glorious"). Many retailers, especially in the US, use Parker's scores to sell wine and his judgments exert a powerful commercial influence.
Wine drinkers find scores out of 100 easy to understand, which no doubt explains their popularity. But for a numerical measure of a wine's quality to have any real value, it would have to be reproducible - and scores out of 100 are not. Wine assessment is far too subjective and inconsistent, and wine itself too variable, for this to be possible.
According to Professor Dennis Lindley, a top British statistician, the evidence suggests that "when experienced tasters work on a 0 to 20 scale, they can, about half the time, be out [from each other] by as much as two and, not infrequently, by as much as four". On the 100-point scale, a variation of just two out of 20 translates to the difference between a score of 81 (which can make a wine hard to sell) and 91 (which can make it hard to find).
Parker prefers his 50-100 point scale to the 20-point scale used in Australian and New Zealand competitions and magazine tastings. Here, in competitions, the scores are not made public, but are instead converted into one of four results - no award, or a bronze, silver or gold medal. The critical advantage of the medal and star rating systems commonly employed in Australia and New Zealand (including, of course, Winestate) is that they cover a much wider band of scores and are therefore far more likely to be reproducible.
Hugh Johnson has suggested a radical alternative to the 100-point system, which you can use at home. The minimum score is one sniff, with a step up to one sip. Two sips indicate faint interest. One glass means tolerance, even general approval. Four glasses means the wine tickles your fancy, and two bottles means it's irresistible.