Where the price isn't the wine
Which wine regions give you more for your money, and which charge you for the name? I compared 585 appellations like for like.
This site started with a question I kept asking at tastings of famous wines: how much of what I am paying for is in the glass, and how much is on the label? Some of those bottles were remarkable. Many were simply very good wines at extraordinary prices. So I went looking for the opposite: the places where the price isn't the wine.
How I measured value
A fair comparison has to be like for like. Pinot Noir is dearer than Malbec, old wines cost more than young ones, and some critics score more generously than others. So each wine is compared only with wines of the same grape, the same vintage, of a similar age and reviewed by the same critic. Then for every appellation I ask two questions:
- The points bonus. At the same price, do its wines score higher or lower than comparable wines?
- The price premium. At the same score, do you pay more or less for them?
An appellation only counts as clear value if it wins on both, and as clearly overpriced if it loses on both, by more than chance would explain. Each one needs at least 25 reviews from at least 5 producers, so a single star estate cannot carry a whole region.
Every red appellation, by price and score
Across: how much more or less you pay than for wines of the same score. Up: how many more points its wines earn than wines at the same price. Top left is more wine for less money; bottom right is paying for the name.
The best value, by style
These are the appellations that win on both counts. Extra points is how many more points their wines earn than wines at the same price; price vs same score is how much less you pay than for wines that score the same. The typical US shelf price is there to give a sense of level.
Reds
42 appellations with clear value
Show all 42Show the top 10 only
Whites
15 appellations with clear value
Show all 15Show the top 10 only
Rosé
3 appellations with clear value
Sparkling
1 appellation with clear value
The red list is dominated by places that have never been fashionable: Madiran, Cahors and Gaillac in the south-west of France, Chinon on the Loire, Gattinara in northern Piedmont, Portugal's Dão. For the same score you would pay noticeably more elsewhere: in Madiran's case, around 60% more. Places close to my heart are on it too: Bandol, Chianti Classico, and Vino Nobile di Montepulciano, from the Tuscan hill town where I first fell for wine.
Is it just the grape?
A fair objection: Tannat is grown in few places, mainly south-west France and Uruguay, so comparing Madiran only with other Tannat could flatter it. In the reviews, 41 of the 155 Tannat wines are from Madiran, and most of the rest are Uruguayan wines that score lower. So I ran everything again without the grape comparison, measuring each red against every red at its price.
Madiran stays first of all 324 red appellations, with +2.7 points and −34% on price. 29 of the 42 clear-value reds keep their place. The ones that drop out, marked “value within its grape” in the lists, are good value only inside an expensive grape: Gattinara and California's Pinot Noir regions are cheap for Nebbiolo and Pinot, not for red wine in general. They are the right swap if you want that grape, not if you simply want the most wine for your money.
Paying for the name
At the other end are the names everyone knows. Most of them are not bad wines; their scores are roughly what wines at a similar level achieve. The difference is the price.
How much more you pay for the same score
Famous appellations, compared with wines of the same grape, vintage and score. The figure on the right is how many extra points their wines actually earn.
Pomerol, Vosne-Romanée, Meursault, Puligny-Montrachet and Condrieu all cost roughly twice what wines with the same score cost, and earn no extra points for it. There is one honest exception: Saint-Julien is +27% dearer for its score, but its wines also score +1.7 points higher than wines at the same price. That is a premium that buys something.
What to buy instead
A good swap has to be the same kind of wine, not just a cheaper one. Where the data allows, these keep the same grape and a similar style; where they cannot, I say how the wine differs.
Same grape. Gattinara, from northern Piedmont, is a little lighter and more perfumed; Langhe Nebbiolo is younger and earlier-drinking. Gattinara scores higher than other Nebbiolo at its price, and costs far less than Barolo for the same score.
Same grape. Brunello is the one Tuscan name in the data that clearly charges for the label. For a wine closer to Brunello's weight and ageing, look at the Riserva and Gran Selezione bottlings of Chianti Classico.
The same rich, barrel-fermented style without the Côte de Beaune premium. Chablis is fairly priced too, but it is a different wine: leaner, steelier and usually unoaked, so not a swap for Meursault.
Not the same wine: Cabernet Franc from Chinon is lighter and more herbal, Tannat from Madiran firmer and darker. But both give a serious, age-worthy red with structure, and Madiran is the best-value red appellation in the data.
Bandol is built on Mourvèdre rather than Grenache, so it is firmer and more savoury, but it shares the southern warmth. It scores higher than wines at its price, and costs less than wines of the same score.
No Riesling region is a clear bargain, but the Rheingau charges a premium for no extra points, while its neighbours to the south are fairly priced and score a little higher. Riesling runs from bone dry to sweet, so match the label: Trocken or Grosses Gewächs for dry, Kabinett or Spätlese for off-dry.
Pinot Noir is the hardest case. In this data, every famous Burgundy village charges a large premium, and the value is mostly in California's cool coastal valleys, which can be hard to find in the UK. The most natural Burgundian swaps for Meursault, the lesser Côte de Beaune villages such as Saint-Aubin, Saint-Romain and Auxey-Duresses, do not have enough reviews in this data to judge. And my own favourite discovery for that style is not in it at all: the Toques et Clochers from Limoux, barrel-fermented with new oak and locally nicknamed “the Petit Meursault”, at under £20.
What it doesn't show
- US prices and one publication. The prices are US retail at the time of review, and the scores come from Wine Enthusiast critics, mostly for vintages from the mid-2000s to the mid-2010s. UK prices, duty and availability differ, so read this as relative value between regions rather than as a shopping list.
- Appellations are averages. A clear-value region still has dull producers, and an overpriced one still has bargains. This tells you where the odds are on your side.
- Scores are not everything. Rarity, age-worthiness and the pleasure of a famous name are real, and some people will happily pay for them. This only measures what a critic found in the glass.
Conclusion
The label and the price tell you surprisingly little. Across nearly six hundred appellations, the most famous names consistently charge for their reputation, and the best value sits in places with no reputation to charge for. If you want wines worth buying again, start with the regions that have to earn your attention: Madiran, Bandol, Gattinara, Chinon, Chianti Classico. Then, once you have found a producer you trust, buy them again.
Reviews and prices: Wine Enthusiast tasting notes from the public “Wine Reviews” dataset (winemag-data-130k), wines priced $5 to $500. Method: for each style, two regressions, log price on a spline of score and score on a spline of log price, each controlling for grape, vintage, age and critic; appellation effects are the average residuals, shrunk towards zero for small samples, with 90% intervals. “Clear value” means both intervals exclude zero in the right direction. Sweet and fortified wines are excluded.