Aging experiment no. 1

Eight months ago I brought some wines into the house to see if the higher temperatures there had a discernible effect on aging. Yesterday for Thanksgiving we had our first experiment with the 2014 Rhys Pinot Noir Alpine Vineyard. We compared the bottle from inside the house with the bottle from the cellar using the usual protocol. Nine guests participated, four pairs sharing the same set of glasses. Each person in a pair was allowed to guess independently.  One person correctly chose the bottle that was different from the others but misidentified the source; nobody else correctly identified the bottle. I would have correctly identified both the bottle and the source if I had stuck with my original impression, but I was talked out of it by my partner. Probably the cross talk between pairs had an effect, so it might be cleaner to have everybody with their own set in the future.

Here is a graph of the temperature difference between the two locations over the last 18 months. The red line is the inside room, the green line the cellar (blue is humidity inside).

Screen Shot 2018-11-23 at 3.58.41 PM.png

We’ll do this again next year!

Discrimination experiment no. 3

Sally, Housten and I tasted an Australian cabernet for this experiment. Usual protocol: half the bottle in a decanter for 3 hours, half under argon gas in the bottle, each person presented with two glasses in random order. I explained that we were looking for the wine that had fewer primary and tertiary aromas and more fruit aromas, and that that wine should be the decanted one. All three of us got it right, and all three of us thought the difference was marked. There was some confusion initially about the instructions: Housten asked if the wine from the bottle should have a more complex aroma (I said yes) and Sally initially got the wines the wrong way around, but her description fit the correct identification. That is, she said the fruitier one was from the bottle.

This and the previous experiment stimulate a couple of thoughts. First, in the previous experiment we were drinking out of glasses that were not ideal for detecting aromas; they were too small so it was hard to swirl and release the aromas into the glass. Second, the instructions are clearly confusing. So for future experiments I should probably stick to my standard wine glasses. And I think I’ll dispense with the long complicated description and just ask two questions: which wine is fruitier and which wine has a more complex aroma? Then correlate the answers with the source.

Here is my tasting note on the wine.

Speaking of wine glasses, maybe that’s another experiment. Riedel has 14 variety-specific glasses in the Vinum series. I chose the Riesling/Zinfandel glass as my all-purpose glass on the recommendation of Making Sense of Wine, by Matt Kramer. It has the right size and shape to swirl and sniff without being absurdly bloated. I’ve always assumed that the classification of glass shapes was a bit like the proliferation of tartans in the 19th century; a fiction designed to sell more glasses (or wool). But I do happen to have some absurdly bloated chardonnay glasses that were sent me by mistake once. Not sure exactly how to manage a blind experiment here; get someone else to hold the glass?

I’m heading off to Italy for a month on Friday, and I don’t think it’s likely they will allow Coravin capsules on the plane, so this blog might become a little sporadic. I suppose I could run experiments comparing a decanter with an open bottle.

Discrimination experiment no. 2

Interesting dinner party with a group of friends. I brought this cru bourgeois Bordeaux, half of which had been decanted 3 hours earlier. We followed the protocol described in the previous discrimination experiment. I described the difference we were looking for in terms of primary versus secondary and tertiary aromas. Of the 7 participants, only one correctly identified the source of the wines. (It was not me.) Now, people’s reasons for their choices did not always follow the description I had given at the beginning; many described one wine as having more complex aromas than the other (that was identified in these cases as the wine from the bottle); some made reference to the acidity. So, maybe there was a flaw in the instructions. Me, I found it very hard to distinguish between the two. None of the people in the group was a self-professed conoisseur.

So, good addition to the data set. I think we need a larger N size.

Discrimination experiment no. 1

This is a new series of experiments following the protocol outlined in my last post. It involves just two glasses, and the test to see if the presence of secondary and tertiary aromas discriminates between the bottle and decanter. For the first experiment in this series I used an inexpensive Bordeaux. Here are my notes, written before the reveal:

On the left, fine spicy cedar notes and good fruit. On the right, a slightly funky mushroom overtone. So the one on the right is the bottle.

I was right. And tasting after the reveal, I thought the tannins on the decanted palate were mollified by the soft fruit. Perhaps this is behind the notes you sometimes read saying that the tannins softened with air. I’ve never believed it possible that the tannin composition changes chemically over such a short time, but possibly the revealed fruit aromas create a perception of softening.

I found the difference quite decisive. In previous breathing experiments the difference often seemed very subtle. Maybe formulating a hypothesis sharpens the mind’s attention in a way not present when you are simply trying to decide if there is any difference at all.

Here is my tasting note on the wine.

Analysis of breathing experiment statistics

After 10 breathing experiments I decided to run some numbers on the results. The raw data is here. I don’t know how a statistician would handle inferences from this data, but I can calculate probabilities for a binomial distribution. The fundamental question was whether there was a detectable difference between wine that had been allowed to sit in a decanter for a few hours (typically 3–6) and wine from the same bottle that had sat under cork and argon gas (or from a different bottle of the same wine in earlier experiments). Subjects were presented with three glasses and had to pick the one that was not like the others, following this protocol. Here are the statistics for the whole group, and for me alone. The last column is the probability of doing as well or better by chance alone (assuming a 1/3 chance of a correct guess in each trial).

Group No. of trials No. correct
Everyone 30 16 0.02
Me 10 7 0.02

I am sufficiently convinced by this that there is a difference detectable by humans. But feel free to check the data and disagree in the comments.

I also ran the conditional probabilities on correctly identifying the source of the wine given that you had correctly picked the odd one out.

Group No. of trials No. correct Probability
Everyone 16 10 0.22
Me 7 5 0.23

This is less convincing, although of course the sample size is smaller. But it suggests to me the next series of experiments. Having established that decanting makes a small but detectable difference, the next question is: what is the difference? Reading over my notes, I think it is reasonable to make a hypothesis in terms of primary, secondary, and tertiary aromas. There’s a nice description these at Wine Folly. Primary aromas are the aromas coming from the grape itself (fruit, flowers, herbs). Secondary aromas come from the fermentation process (yeast, cream, lees, mushroom, and Brettanomyces aromas such as barnyard, game, and bacon) and tertiary aromas come from the aging process, particularly from aging in oak barrels (vanilla, nuts, baking spices, cedar, smoke, tobacco, leaves). We will adopt the terminology from Wine Folly: “aroma” means the primary aromas, and “bouquet” means the secondary and tertiary aromas.

The hypothesis we will work with in this series of experiments is:

Wine from a freshly opened bottle has a stronger bouquet, which tends to mask the aroma. The aroma is more prominent in wine from the decanter.

I’ll use a new experimental protocol. The question is no longer whether there is a difference, but whether a particular description of the difference is accurate and usable by humans. I’ll present subjects with two glasses blind, one from the bottle and one from the decanter, ask them to read the description of aromas above, and then ask them to apply the hypothesis to identify which is which.

I will also collect data on preference, and for that reason I will organize the experiments by varietal, starting with Bordeaux varietals and blends. Then I will move onto pinot noir. The conventional wisdom is that breathing is good for the former and not for the latter. The breathing experiments so far somewhat support this.

Breathing experiment no. 10

This time I chose the 2014 Matthiasson Napa Valley Cabernet Sauvignon. Matthiasson makes wonderful wines that blend the vividness of California with the structure and balance of France. (Jason, photo of label below.) But I digress. Usual protocol. My notes:

First impression: first one is more floral and fruitier, second two more masked with a cedary spicy note, so dbb. No discernible difference on the palate. In this case the masking note is quite nice. At one point I almost wanted to reverse the classification of the first and third bottles, for bbd. But I started to get confused, so went with my first impression, even though it has failed me the last couple of times.

Failed me again: the correct configuration was my second guess, bbd. I seem to be having trouble second guessing whether I should second guess. Still, either way, the thing I thought I was detecting—the presence or absence of a masking note—is consistent with previous breathing experiments with cabernet and Bordeaux blends. So I think there’s something in that hypothesis, and maybe I should try to design an experiment to test it explicitly. Perhaps a very long decant to emphasize the difference.

John Gilman’s often talks about a top note of bla bla bla. I think I’m beginning to understand that, and used the phrase in my tasting note for this wine.


Consistency of tasting notes across time

Not so much an experiment this morning as a collection of raw data. I was struck by the consistency of this sequence of three of my tasting notes over a three-year period, so I decided to examine other sequences. I have too many sequences of three to handle without some serious data-mining, but here is a list of all the sequences of four or more, excluding sequences that included an un-rated or flawed wine or a note written from memory.

2000 Il Marroneto Brunello di Montalcino

2008 Domaine de Terrebrune Bandol

2012 Jamsheed Syrah Beechworth

2012 Jauma Syrah Scyras Clarendon Vineyard

2013 Denis Jamain Reuilly

2014 Jean-Louis Dutraive (Domaine de la Grand’Cour) Fleurie Chapelle des Bois

2004 Château Bernateau

2010 Château Haut-La Pereyre

1998 Pertinace (Cantina Vignaioli) Barbaresco Vigneto Castellizzano

It’s a mixed bag. I don’t really have time to go through them all (have to get on a plane this morning), nor do I have the tools to analyze them in a rigorous way. The 1998 Barbaresco shows wide variation, which is reflected in the other notes, and it seems reasonable to suppose that it was undergoing change. Others are very consistent, for example the 2014 Fleurie. I have to run now, but will examine these further and perhaps add some thoughts in the comments section. I should note that I never look at previous notes on a wine before writing a new one, and I don’t remember much from tasting to tasting, so these are largely independent notes.



Breathing experiment no. 9: short decant

Following on my last Bordeaux experiment I tried a short one-hour decant of a classically styled California cabernet. Usual protocol. My notes (as always, written before the reveal):

First impression, more fruit on no. 1, more stemmy notes on the others. This  difference fades almost immediately. I began to think no. 3 was more like no. 1 after a while, but will go with first impressions again. Based on previous observations, no. 1 would be the decanter.

Well, if I had changed my mind on no. 3 I would have gotten it right; my configuration was dbd. The note on no. 1 is consistent with observations in previous Cabernet or Bordeaux experiments, that the fruit comes out a little more after a decant, or rather that the stemmy mask drops away. But the difference was very slight; it’s hard to be confident that the guesses were anything other than guesses here.

What is the difference between Côte de Beaune and Côte de Nuits?

The Côte d’Or is a ridge of limestone in Burgundy, divided into the Côte de Nuits in the north and the Côte de Beaune in the south. It’s where the most famous red burgundies come from, with the famousest (and most expensive) coming from the Côte de Nuits. Here is a typical description of the difference between the two:

The top reds from the Côte de Nuits . . . often have greater intensity and a firmer structure than red wines from the Côte de Beaune . . . . By contrast, the top Côte de Beaune reds are frequently softer and sometimes more lush. In general, reds from all over the Côte d’Or are prized for their soaring, earthy flavors, often laced with minerals, exotic spices, licorice, or truffles.

—Karen MacNeil, The Wine Bible

In the last five years I have written 1030 tasting notes on CellarTracker, 41 on red wines from the Côte de Beaune and 34 on red wines from the Côte de Nuits (as of this writing; there will probably be more by the time this post is published). So I thought I’d do a text analysis to see if my notes reflect the difference described above. I used this online utility to create a word frequency list for each set of tasting notes. From that I created a list of descriptors. This included most words that refer to the aroma, flavor, or position on the palate. Some words were problematic: for example, the word “tannin” might be modified by “strong” or “weak,” but that information is lost in the frequency count. So I eliminated words related to strength of aroma, tannin, acid, fruit, or finish. It will take a later analysis to detect differences there. I also consolidated some terms—for example forest floor, stems, stemmy, heath, brambles, bracken, leaves, undergrowth—where I thought the distinction was likely to be noise.

Finally, I computed a CDB index for each descriptor: the frequency of a descriptor per tasting note for Côte de Beaune divided by the sum of the frequencies for Côte de Beaune and Côte de Nuits. Descriptors in the first section in the table below, with CDB index less than a third, are more than twice as likely to occur in a Côte de Nuits tasting note than in a Côte de Beaune, and it’s the other way around for descriptors in the third section. The descriptors in the middle section apply pretty well equally to all wines in the Côte d’Or.

So, to take a bit of poetic license with the table, red wines from the Côte d’Or in general are complex, with aromas of berry, violet, smoke, and undergrowth, and a nice spread of fruit on the palate with a focused core and crackling acids (well, “spread” and “focused core” are contradictory, so really it’s either/or there). Red wines from the Côte de Nuits are austere, elegant, and balanced, with shy flavors of strawberry, raspberry, blackcurrant, black cherry, earth, mocha, and thyme, whereas red wines from the Côte de Beaune are rich and soft with penetrating jammy fruit and crusty aromas of mushroom, leather, and sweet spice.

How much of this do I believe? Well, taking the flavor descriptors with a grain of salt, I think my tasting notes bear out the distinction noted by Karen MacNeil between the firm structure of  Côte de Nuits and the soft lushness of Côte de Beaune. Beyond that, there is a savory/sweet dichotomy to the spices, perhaps some greater complexity in Côte de Nuits, perhaps some greater fruit concentration in Côte de Beaune. I suppose the next step is to do a blind tasting to see if these distinctions correctly guide identification.


Breathing experiment no 8: a win for decanting

I thought I was pretty well done with breathing experiments, but I realized there were some gaps, such as classified growth Bordeaux. So I pulled this bottle from my cellar and ran the usual protocol, just Amy and me. Amy guessed the odd wine out correctly, and I did not. My notes:

Funky note on 2, fruit clearer on 1, dbd

Which means I thought glass 2 was from the bottle, and glasses 1 and 3 were from the decanter. I was right about glasses 1 and 2, but wrong about glass 3, which was also from the bottle. I had in fact hesitated over glass 3, but went with my first impression, which failed me this time.

But notice the distinction between glasses 1 and 2. I found the decanted glass had better fruit aromas, whereas in the glass from the bottle the fruit was obscured by a funky note. This has happened twice before, once for a burgundy and once for a cru bourgeois Bordeaux. In all three cases the fruit aromas were more prominent after decanting. In this case I also detected a better fruit flavor on the palate, which I would attribute to retronasal sensing of the better aroma.

In the case of the burgundy, I quite liked the initial savory note; it’s something I appreciate in a red burgundy and I wouldn’t want to lose it to air.

For the Bordeaux, on the other hand, I preferred the decanted wine in both cases. This may call for more experiments with Bordeaux, in particular finding out what the shortest effective decanting period is. In this case the wine was in the decanter for 5 hours before the experiment. And I should extend the experiment to new world Bordeaux blends as well.

As with the other experiments, I did not sense any difference in the acids or tannins on the palate, and, as always, the differences were slight, not the sort of thing you would detect without focused attention.