Bottle-to-bottle variation

[Duane Bowman]

Recently Jeff Leve described an experience with an '83 Margaux as: Tasted for the third time in two weeks. Each bottle has been different .....

Which lead me to a topic that I've been thinking about for some time now but never heard discussed. How wines bottled side by side and stored side by side can have very different flavor and aroma characteristics - a phenomenon that winemakers experience because we taste the same wines many times, but one most others may not - or may just shrug off as differences in how they were feeling that day or their day's changed "palate." But I don't believe those are what's at play. Now, I've never seen science/research on this topic and some of my more mystical winemaker brothers attribute the differences to phases of the moon or a host of different wooo-wooo factors, but I'm not quite that mystical. I just know it exists.

So why would anyone care? The question of bottle-to-bottle variation in wines really is about how wine reviewers or critics account for it in their ratings. How does one give a wine a 93 while the next reviewer gives it an 87? We usually say it's just the difference in the reviewer's palates, but is it really? Are both reviewers really tasting the SAME wine? Will the consumer be tasting the SAME wine as either of those reviewers if/when they purchase it? Is there a better way to review bottle variable wines than giving some simplistic score? How do we give "reliable guidance" to consumers when we're speaking of a product that varies in the very things consumers buy it for? Do we pretend bottle-to-bottle variation just doesn't exist? Should reviewers say how many separate times they've tasted a particular wine to arrive at their review? I'd love to hear thoughts from the wine community.

[Jeff Leve]

Recently Jeff Leve described an experience with an '83 Margaux as: Tasted for the third time in two weeks. Each bottle has been different .....

Which lead me to a topic that I've been thinking about for some time now but never heard discussed. How wines bottled side by side and stored side by side can have very different flavor and aroma characteristics

Hi Duane

Thanks for coming over and posting. You asked a bunch of good questions.

Variances are more likely to take place in older wines. Young wines should be relatively stable and offer the same experience. Mature wines are a different story. With older wines, two possibilities arise, cork and storage. Each cork allows a different level of seal. Bottle conditions from the same case can vary with time. I've opened older cases and some bottles have higher fill levels than other bottles. This is common with wines over 20 years old. Some corks offer a tight seal that doesn't allow much air to enter the bottle. Other corks from the same case don't fit as tightly. If air comes in, wine needs to leave to make room. The bottles that allowed more air to enter into the bottle caused the wine to evaporate at a faster rate. The ullage or fill level will change a bottles characteristics.

Storage definitely effects the condition of the wine, which make for a different tasting experience. Older bottles with high fill levels will show a wine that tastes younger and fresher. Wines with low fills will taste completely different.


The question of bottle-to-bottle variation in wines really is about how wine reviewers or critics account for it in their ratings. How does one give a wine a 93 while the next reviewer gives it an 87?

We've covered bottle variation. All tasters, professional and amateur alike can taste the same wine and come away with different impressions. We all have different preferences. For example, some tasters might find the scent of truffles in a wine off putting. Truffle is a common scent found in older Bordeaux. I love it! But for people who do not like that in a wine, they will score it lower. Experienced tasters will have individual ideas on what a wine should taste like. If a wine does not offer the sensations they think a wine should have, professional writers could downgrade the wine. Take 2009 Cos d'Estournel for example. For me, it is one of the top Bordeaux wines of the vintage. I posted 3-4 reviews on the wine already. I think it is a candidate to score 100 Pts when it's in bottle. A few people. professionals and friends of mine did not agree. They did not like wine. It's all about personal taste and expectations.


Are both reviewers really tasting the SAME wine? Will the consumer be tasting the SAME wine as either of those reviewers if/when they purchase it?

For young wines, yes. For older wines, probably not.

Is there a better way to review bottle variable wines than giving some simplistic score?

If you are reading a score and that's all that is offered in the review, throw that away. You need to read a detailed description on what the wine offers in taste and feel. It's the tasting note that counts. Regardless of a wine earning a high or low score, if you do not like the sound of the wine based on the descriptors being used, you are not going to like it. Perforce, if the adjectives being used are a turn on, you are prone to liking the wine.


How do we give "reliable guidance" to consumers when we're speaking of a product that varies in the very things consumers buy it for? Do we pretend bottle-to-bottle variation just doesn't exist?

Bottle variation should not come into play much with young wines. A review says as much about a wine as it does the writer. As for the quality of the wine writer, look at their body of work. Do you agree or disagree with the critics previous reviews more often than not. Track records matter.

I am not a professional wine critic. But I taste a lot of wine and write about it when I can. if you want an idea of what I like or not, there are almost 2,500 wine reviews on my site that you can read to determine the styles of wines I like or not.

Should reviewers say how many separate times they've tasted a particular wine to arrive at their review?

Why not? However, most notes are based on tasting the wine once or twice for each dated review. I hope this helped and that I answered your very good questions.

I look forward to seeing you over here more often.

[mustafa camlica]

Old wines are extremly volatile. Even in the same case. It is a kind of Russian R.

[Tony Velebil]

I think bottle to bottle variation, at least as it relates to professional/paid critics and consumers, is largely due to taste preferences UNLESS:

1. the wine was recently bottled - say less than 6 months early. Recently bottled wine tends to go through many ups and downs in that first six months post bottling.

2. the wine is slightly tainted but at levels below threshold.

Other than that, all things being equal it's taste/style preference. For instance, i don't think James Laube particularly likes Rhone blend wines (he may like Syrah all by itself but not blended) and those wines typically don't score well with him. - Anecdotal evidence only.

For older wines, assuming they were delivered from the winery in ideal conditions and stored perfectly, whatever perfectly is, I think it comes down again to taste and corks.

Corks are not perfect closures. You look at a bag with 500+ corks in it and there are no two identical corks. They allow air at varying rates which could definitely affect how the wine "performs" when opened some 10, 15, 20 years down the road.

I also think importers didn't pay as close attention to shipping details 20+ years ago. My guess was much less refrigerated reefers and delivery trucks.

Heck, I saw a delivery truck delivering inexpensive wines to a retailer in Calistoga on a 100+ degree day and from what I could tell no air conditioning.

[Tim Oregon]

Bottle variation is bound to occur, because you will have variations in all the different vats/tanks from which the wine have been bottled. I just wonder how much quality control is being done in wineries. Not having been to a winery to see how wine is made I just know from working with large samples in a science lab, you will get intra and inter variations within any sample size greater than 1, this is at the start of the process, then when it gets bottled, external factors will influence the quality of the wine in the bottle once it reaches the consumer and is consumed.

[Tony Velebil]

Tim - many smaller US wineries will blend everything into 1 or 2 tanks prior to bottling so variation from that is limited. Much more a factor in Chateauneuf where some producers bottle by barrel for various importers. I don't know Bdx practices or other regions. But the wineries I worked at here in Nor Cal, we blended everything into tanks and the let the wine settle for up to a week to make sure it was settled and knitted together.

[Christer Byklum]

On the 09 Cos d'Estournel, I'm one of those not agreing with you Jeff. To me it taste and behaves like an overripe, over extracted, and way too alcoholic wine to be a Bordeaux. That wine could be made everywhere on the planet. Feels like it's on steroids. Not a high marker from me. A very nice bottle of wine, yes indeed, but an awfull example of a great Saint Estephe......

[Tim Oregon]

Hi Tony

Even one or two tank, the wine will be different (to what degree/limits are the tolerances OK). If you are talking multiple barrels the difference between barrels would be interesting to know and if there is quality issues maybe the wine makers can try to bring the wine into a range.
Anyway here is a question, When taster's taste the wine the first time say at EP (Bordeaux), does the wine come from the Finest Barrel chosen by the Wine Maker or is it randomly chosen from the many barrels in the Cellar. Just Curious about the process !!!!

[Tony Velebil]

If it is like what I have seen in Nor Cal...they choose the best barrels. You want to put your best possible foot forward as they say. Oh, it should still be close to what the final blend will be, but there is definitely a focus on THE BEST barrel(s).

[Tim Oregon]

Thanks good to know, because if that is the practice in the wine world, then what the consumer will ultimately get is not the 93 point wine awarded by the taster, but something a tad inferior in my opinion.

[Tony Velebil]

To be fair, a really high barrel score is still an indicator that the wineries have some really good material to blend with and the wine shouldn't be hugely different once in bottle - maybe 2-3 pts but if more than that, either something happened at bottling or their was an effort to deceive or the finished bottle tasted may be tainted in some way (TCA, brett, etc).

To be really fair, sometimes even though they pull best barrels at the time, they haven't gone through full blending trials yet. So they go with what is showing best in the winery at the time. By the time blending trials happen maybe the blended is as good or better.

Long story short, I wouldn't say necessarily that a wine will be a lesser quality than the barrel score. It should be close to what the barrel sample scores/tastes like. If you see wild swings well then...

[Christer Byklum]

That is why I always take barrel samples as a indication, and I also rate them on a score like 93-95, as one does not quite know. To taste the bottled wine is a much more secure way of finding out what it really is.