Ninety-six sits beside one year on the chart, eighty-seven beside the year next to it, and the decision appears to make itself. Buy the 96. Avoid the 87. That instinct costs money, and more often it costs good drinking, because a single number has been asked to describe an entire region, a whole harvest, hundreds of producers and every decision they made after picking.
Charts are useful. They are just useful for a narrower job than people ask of them.
What one number is standing in for
Bordeaux is not a vineyard. It is tens of thousands of hectares across both sides of an estuary, with different soils, grape mixes and picking dates, and several thousand producers making independent decisions. A vintage chart takes all of that and returns one figure.
To see what gets lost, it helps to know what makes a vintage good or bad. Four things dominate. Flowering in June needs to be even and quick — if it is cold or wet, the crop sets unevenly and bunches ripen at different rates on the same vine. Summer needs moderate water stress: enough to slow the vine's leafy growth and push it towards ripening fruit, not so much that it shuts down. Late summer needs warmth. And September needs the rain to hold off, because rain near harvest dilutes berries and invites rot.
The thing charts flatten hardest: ripening dates
Merlot ripens roughly a week or two before Cabernet Sauvignon. That fact alone breaks the idea of a Bordeaux score.
If the rain arrives in late September, the Right Bank estates of Pomerol and Saint-Émilion, planted mostly to Merlot, may already have their fruit in the cellar. The Médoc, waiting on Cabernet, is still hanging and takes the damage. Reverse the calendar and the outcome reverses. So a year can be genuinely strong on one side of the river and mediocre on the other, and a chart that prints "Bordeaux 2014: 90" has told you nothing about which side you are standing on. Better charts split left bank from right; many do not, and people read the summary line anyway.
Soil compounds it. The deep gravel mounds of the Médoc drain fast, a gift in a wet year and a problem in a drought. The clay of Pomerol holds water and does the opposite. In any given year one is being helped and the other punished, inside the same score.
The bigger flattening: what happened after the harvest
This is what has changed most in thirty years, and charts have not caught up. A classified estate in a difficult year now has options it did not used to have. It can sort optically, berry by berry. It can pick in several passes. Above all it can declassify — sending far more of the crop into its second wine so the grand vin is made only from fruit that was genuinely good.
So in a poor year the gap between a serious estate and an ordinary one widens enormously. The spread between producers within one vintage is now larger than the spread between a good vintage and an average one at the same producer. A chart cannot express that: it has one column and the answer needs two.
Colour and sweetness break it completely
Dry white Bordeaux and Sauternes run on different requirements from the reds. Sauternes needs botrytis, and botrytis needs damp autumn mornings off the Ciron followed by dry warm afternoons. Some of the conditions that ruin a red harvest are exactly what a sweet-wine producer is hoping for. 2013 is the standing example: a genuinely difficult year for Bordeaux reds and a very good one in Sauternes. Any chart giving that year a single number is wrong in at least one direction.
What the drinking window is really worth
The "drink 2028–2045" column is a guess, often made by someone tasting from barrel eighteen months after harvest, before the final blend was fixed. Treat it as a rough ordering, not a schedule.
And note what it assumes: a cool, stable cellar. In Hong Kong that assumption fails constantly. A bottle from a highly scored year that spent three summers in a flat at 29 degrees has moved through its window far faster than the chart imagines, and will be tired when the chart says it is arriving. Provenance beats vintage. A lesser year that has been properly stored will out-drink a great year that has not,.
The advice that is wrong
Avoid off vintages. This is repeated constantly and it is bad advice for most drinkers, for one straightforward reason: a lesser year from a good estate is often the most enjoyable Bordeaux you can buy right now.
The mechanism is simple. Cooler, wetter years produce lower sugar, so lower alcohol, and less phenolic ripeness, so softer and less abundant tannin. That means the wine reaches an approachable state years earlier. It also costs less, because the market prices vintages by reputation. If your intention is to drink Bordeaux over the next five years with dinner — rather than to store it until 2045 — the year everyone is dismissive of is very often the correct purchase.
The mirror-image error is treating a celebrated vintage as the safe choice for tonight. A structured, tannic year is precisely the one that is unrewarding young. Buying the 96 and opening it this weekend gets you a wine that is closed and grippy, and you will conclude that you do not like Bordeaux.
How to use a chart properly
Use the readiness column, not the score. That is the question a chart is actually equipped to answer: is this year likely to be drinking now, or does it still need time?
Use the score to set expectations about structure and longevity, not quality. Then ignore the chart and ask two things: who made it, and where the bottle has been. Those will do more for what ends up in your glass than any number on a laminated card.
Where to start
If you are drinking Bordeaux over the next five years rather than storing it to 2045, the dismissed year from a good estate is usually the correct purchase — lower alcohol, softer tannin, ready now, and cheaper because the market prices reputation.
- Bordeaux Red — sort by producer and readiness rather than by the number on the chart
- Château Gloria 2013 — the vintage every chart tells you to avoid, from a Saint-Julien estate that made good wine anyway, around HK$270
- Château Cantemerle 2008 — a classified growth from an unfashionable year, seventeen years on and open for business