Put two scoresheets side by side on the same desk. On the left, a tasting card with a row of small numbered boxes from one to twenty, half-point marks penciled in between, the kind Jancis Robinson and Clive Coates still use for the wines they evaluate every week. On the right, a 100-point page running from fifty to one hundred, the format Robert Parker codified in 1978 when he printed the first issue of his Baltimore-Washington newsletter on a borrowed mailing list. Both sheets are scoring the same flight. Both arrive at numbers that get reported in the trade press, quoted on shelf-talkers, used to set the opening price of a Bordeaux release.
Only one of them claims a hundred points of resolution. And only one of them is telling the truth about how many points it actually has.
The Bottom Half of the Scale Does Not Exist
The 100-point scale, the structure Parker popularized in conjunction with his friend Victor Morgenroth and introduced in The Wine Advocate, does not run from zero to a hundred. It runs from fifty. Wikipedia’s account is direct: “51 rather than 100 different ratings are possible.” Half the apparent precision is mathematical fiction. Nothing scores below fifty because nothing is meant to.
Like this? The TERROIR Letter arrives every Thursday.
What is meant to be there narrows further the moment the sheet leaves the critic’s desk. Parker’s own quality bands set the terrain: 80–89 is “above average to very good,” 90–95 is “outstanding,” 96–100 is “extraordinary.” Wines below 80 are functionally untradeable; the trade press will not run the reviews, retailers will not write the shelf-talkers, importers will not place the orders. The commercially meaningful range, the band that moves money, is 90 to 100. Ten points are doing the work the scale appears to assign to a hundred.
This is the compression problem at the heart of wine criticism. A 91 and a 93 are not “two points apart” the way two degrees of Celsius are two degrees apart. They sit on adjacent steps of a ten-step ladder, and the ladder does most of the work in the buyer’s head.
How the System Was Built
Parker designed the structure to do something specific. In the mid-1970s the American consumer wine press ran on either narrative reviews or the British 20-point sheet that Robinson and Coates still defend. Parker’s stated reason for moving to 100 points, the version Wikipedia carries, was to counter what he believed to be confusing or inflated ratings by other wine writers, many of whom he accused of a conflict of interest. The 100-point scale, he argued, would look familiar to Americans who had spent twelve years in a school system that graded exams the same way. A 92 reads as an A-minus. A 17 out of 20 does not.
The familiarity was the point. So was the apparent precision: a single integer reads as decisive, easier to cite, easier to print on a shelf-talker than a paragraph of comparative prose. Within a decade the 100-point format was the default of American wine criticism. Wine Spectator, Wine Enthusiast, and most of Parker’s competitors adopted it. The 20-point sheet held on at Decanter, at JancisRobinson.com, and inside the wine-trade education system. Almost nowhere else.
What a Single Point Is Worth
If the scale’s resolution were honest, a one-point gap would carry one percent of the difference between an undrinkable and a perfect wine. In the actual market, a one-point gap can carry the difference between a sale and a clearance. The figure most often quoted in the trade, attributed to a Bordeaux shipper cited in Elin McCoy’s The Emperor of Wine, runs as follows: “the difference between a score of 85 and 95 was 6 to 7 million Euro.” The same passage notes that a bottle rated 100 can multiply its price fourfold.
Château Quinault, a Saint-Émilion property that had struggled to move bottles at 100 francs, saw its 1998 rise to 125 francs in half a day after Parker awarded it a 92. Twenty-five percent of the price moved on a two-point shift, in a single afternoon, on the basis of one critic’s number. The wine in the bottle did not change. The number on the page did.
That asymmetry between rating-points and revenue-points is the structural feature buyers rarely see. An 89 and a 91 from neighboring parcels, same vintage, comparable winemaking, are functionally identical wines. The market treats them as occupying different price tiers. The scale’s compressed resolution magnifies decisions that fall well inside the margin of error any honest critic would admit to.
The Critique From Inside the Trade
Robinson has spent much of her career pointing at the same compression. Wikipedia summarizes her position alongside Coates’s: numerical rating systems are questionable given the subjectivity of wine tasting and the variance a wine’s age and the circumstances of tasting can cause. The 20-point sheet is not a claim to superior accuracy. It is a refusal to claim more accuracy than the act of tasting can deliver. Twenty points with half-point increments gives a critic roughly forty meaningful steps, approximately what a careful taster can defend on the evidence of a single afternoon’s flight. The 100-point scale offers fifty-one and then collapses, in practice, to ten.
The deeper critique is about what the apparent precision does to the reader. A wine described in prose as “structured, slightly closed, will benefit from another four years in bottle” can be argued with, qualified, revisited. A wine scored 93 is fixed. The number passes from notebook to shelf to auction catalog to spreadsheet, and the qualifications drop off at every step. By the time the figure reaches the consumer it has shed every hedge that justified it.
Is the Scale’s Power Now Waning?
The question moving through the trade in recent years is whether the format still does what it did in the 1990s. Critic-driven retail purchasing, the channel where the shelf-talker quoting a Wine Advocate score did the heaviest lifting, has lost ground to allocation lists, direct-from-domaine club sales, and sommelier-driven on-premise programs that source by relationship rather than rating. Younger buyers rely less on aggregate scores and more on producer biography, importer reputation, and the wine-bar pour they trusted.
The institutional position remains substantial. Bordeaux en primeur prices still move on barrel scores. Auction houses still print critic numbers in catalog copy. Wine Spectator‘s annual Top 100 still generates the year’s most-searched wines. But the share of wine sold by score has narrowed. The grip on high-end Bordeaux and California Cabernet remains firm; the grip on the rest of the wine world is loosening toward formats the scale was never built to measure. Allocation, sommelier curation, and producer-direct discovery run on a different mechanism: trust in a small human filter, not confidence in a number.
Set the two scoresheets back on the desk. The 20-point card carries forty meaningful steps and admits it. The 100-point page carries fifty-one and behaves as if it carried a hundred. Both critics are doing the same work. Only one of them is asking the reader to believe a number that has, for the past forty-seven years, been carrying more weight than the math underneath it can bear.
The TERROIR Letter
The story behind every bottle.
One feature. One dispatch. One bottle worth opening. Every Thursday — free.