I used to score everything. Every bottle got a number, recorded in a spreadsheet, and I had a system I was quite proud of. I stopped several years ago and I don't intend to go back.
The reasons weren't primarily philosophical. They were practical, and they emerged from noticing what the act of scoring was doing to how I tasted.
The anchoring problem
The first thing I noticed was that my score for a wine was heavily influenced by what I'd expected before opening it.
A bottle I'd paid a lot for, or that came recommended, started at a high number in my head, and my tasting became a process of confirming it. A cheap bottle started low. The tasting was doing much less work than I'd assumed.
Once I saw that, I couldn't unsee it. And the more I read about expectation effects in perception, the clearer it became that this isn't a personal failing — it's how everybody works, including professional critics, and blind tasting is the only real defence.
The commitment problem
The second issue was more subtle. Writing down a number commits you to it in a way that a description doesn't.
Once I'd assigned 92 to a wine, my subsequent glasses were interpreted through that judgement. If the wine opened up and improved over an hour — which they frequently do — I found myself reluctant to revise, because the number was already recorded.
Prose doesn't do this to the same degree. You can write "closed and a bit austere at first, opened considerably after an hour" and both observations coexist. A single number forces a resolution that the experience didn't actually have.
The false precision problem
The one that finally decided it. I ran an informal test on myself: tasting the same wine blind on two different evenings within a week and comparing my scores.
The variation was substantial. Not enormous, but consistently three or four points, occasionally more. Which means every score I'd ever recorded had error bars wider than the differences I was using them to express.
This is entirely consistent with published research on tasting consistency, which finds that individual judges show considerable variation when re-tasting the same wines. I just hadn't expected to be quite so ordinary about it.
If my measurement error is four points, then reporting integers is claiming a precision I demonstrably don't have. That felt like a straightforward accuracy problem rather than a matter of taste.
What I do instead
Three categories and a note.
The categories are roughly: would buy again, glad I drank it, wouldn't repeat. That's the actual decision-relevant information and it maps onto how I behave.
The note covers the structural facts — sweetness, acidity, tannin, body, oak — plus a sentence on style and a sentence on what I'd eat with it. Plus the price, which is the context that makes the rest meaningful.
What I've found is that this is more useful to me a year later than the numbers ever were. Reading "92" in a spreadsheet tells me nothing about whether I want to buy it again. Reading "high acid, no oak, needs food, excellent with the roast chicken" tells me everything.
The objection
The obvious counter is that scores are useful for consumers who don't have the vocabulary or the time, and that's fair.
Somebody in a supermarket with three minutes and no wine knowledge is genuinely helped by a number, and a system that requires everybody to read a paragraph is a system that mostly won't be read. I don't dismiss this.
My response is that a coarser system would serve that purpose equally well without the false precision. Three or five bands, clearly defined, would help the browsing consumer just as much and would be honest about the resolution of the underlying judgement.
The reason we have a hundred points instead of five is not that the extra resolution is real. It's that it enables the price-threshold effects that benefit the people selling wine.
What changed in my drinking
The practical consequence I didn't anticipate: I enjoy wine more.
Scoring turns every bottle into an assessment task. There's a part of your attention permanently occupied with judgement rather than experience, and it doesn't switch off. Removing it meant I could just drink the thing.
I also became much less anxious about whether a bottle was good enough for an occasion, which had been a real constraint. Without a number attached, a wine is just something you're drinking, and it's either enjoyable or it isn't.
What I'd suggest
If you keep tasting notes, try dropping the number for six months and see whether you miss it. My prediction is that you'll find the descriptive part was carrying all the useful information and the score was carrying your ego.
And if you rely on published scores, one adjustment: read the note underneath. The prose usually tells you the style, and style determines whether you'll like it far more reliably than a number ever will.