Airing madeira - an experiment

[Steve Shapiro]

The question frequently arises about the proper airing of madeira. I'm convinced that I've diminished the wondrous flavors of the wines on occasion, by inadequate air exposure prior to drinking. In the past I've suggested that for every decade the wine has been in the bottle, the wine should be aired for one day prior to pouring. With time, I evolved to airing for a week per decade in the bottle. The only authority I've come across mentioning the issue was Noel Cossart in his 1984 book. He suggests a few hours of air time for young wines and overnight airing for older wines, presumably those that have been in bottle for a number of years. My son David and I cooked up the following experiment to try to resolve the issue. I took 5 identical bottles of Miles 1863 Malmsey solera, bottled in the 1980s, bought as part of a full case in the early 1990s, and stored upright together in my temperature and humidity controlled cellar. All bottles were decanted, the bottles cleaned, and then redecanted immediately back into the original bottle. The bottles were then left unstoppered from the time of decanting to the time of the tasting on January 29, 2010. The various bottles were decanted 6 weeks, 2.5 weeks, 3 days, 1 day, and 2 hours prior to the tasting. The wine decanted 2 hours before the tasting was passed one time thru a Vinturi wine aerator. The wines were bagged and served blind to 23 tasters who were told only the the wines were malmseys with soleras started during the U. S. Civil War. My son, a wine professional and myself knew the purpose of the tasting. The order of service was random and unknown to all. The following are my tasting notes, in the order of wine service:

All wines were dark amber/brown in color with a medium green edge. The scores are mine. The group scores reflect fisrt place/ second place/ last.

#1. (Open 3 days). Nose of molasses, a touch of cloves, and a little coffee. In the mouth molasses, a little oriental spice, excellent acidity. Full, rich finish. 18, improving to 18.5 over 2 hours in the glass. 7/4/0.

#2. (Open 6 weeks). Nose sweet, but molasses less obvious; picked up a little pecan; much less aggressive nose than other wines. Medium body with molasses, but seems thinner than #1; acidity fine. Slightly shorter but still lovely finish. 17+ early and late. 3/7/0.

#3. (Open 2.5 weeks). Full molasses nose plus toffee, nuts and a little cloves; alcohol shows thru a bit. Excellent balance in mouth with mixed spices, moderate molasses, fine acidity. Full, rich finish. 18 early, 17.5 late. 5/4/1.

#4. (Open 2 hours, passed thru Vinturi aerator once). Overwhelming nose of fuel oil plus alcohol, overwhelming the orange peel and molasses. Again molasses predominates in mouth, with a little B and B, but the petrol smell persists enough to be distracting; excellent acidity. Rich long finish. 16 early and late. 0/1/19. SRS note: Based on previous experience with the wine, I detected no obvious effect of the Vinturi aeration.

#5. (Open 1 day). Initial nose had enough petrol to be off-putting, but much of this blew off with time, resulting in a nose almost identical to #1. Again similar to #1, but seemed a little less full and rich; excellent acidity, long finish. 17 early, 18 late. 3/8/0.

If one can extrapolate from the results of this tasting, I'd say that airing the decanted wine for about one day per decade in the bottle looks like a good guess. It allowed enough time for the reduction/ bottle funk to blow off and for the flavors to develop. Longer airing initially added to the complexity, but eventually resulted in decrease in the intensity on the nose and in the mouth. At any point in the airing process, stoppering the bottle stabilizes the wine which the remains little changed for years.

I'm interested in the experience of others.

[Eric Ifune]

Facinating stuff.
I usually recork my bottles which I'm saving with good results. I don't have side by side comparisons like yours, however; so its all antedotal.

[Alan Gardner]

I loveit! Exactly the type of experiment I love to do. And this is typical of the things we try in our group (except my cellar isn't as deep in any individual wine).

After about 15 years of 'similar' experiments (not necessarily with madeira) we found this methodology 'mostly reliable'. There were (unfortunately) a couple of side effects, which ONLY affected the first and last served wines. The initial wine seemed to have a bias for finishing top. The final wine served had a bias towards finishing last. I discussed this with a friend who had been a member of the San Francisco based 'Vintners Society' and they had also found the smae bias.
Knowing this we are VERY careful in assigning serving order, and personally I now taste pseudo-randomly, (i.e. not first served through last served) and ALWAYS do a 'dummy first taste', of whichever poured wine has the highest level in the glass, which is never scored in any way.

Having said all that, it seems clear that the 'Vinturi' sample, most recently opened, was clearly the laggard (I've never ever used the Vinturi pourer).
Our 'formal experiments' haven't involved Madeira, but we did test 3 different wines over a 24-hour period (Bordeaux, Barolo, Calif Cab) and in ALL 3 situations the 'uncork and pour' sample beat out the decanted sample and the day-old sample. (OK there was a tie for first in one bottle). Since then we have AVOIDED decanting at all (except when sediment).
This, of course, is totally the opposite of 'conventional wisdom'. And is, of course, the reason we do blind tastings wherever possible.

We have tried a comparison of 'early bottled vs later bottled' of the same Madeira (on 3 occasions) and in ALL situations we preferred the 'early bottled' i.e. for the same source wine, we prefer 'less barrel, more bottle' to 'more barrel, less bottle'. However, with only 3 samples (and different wine each time) it's hard to draw a generally applicable conclusion.

But here's to more experimentation!
Thanks for the notes.

[Peter Reutter]

Great experiment!
Speaking from my Madeira experience (not from any detailed experiments) I would say, that your rule of thumb "a day per decade" sounds reasonable. Most older Madeiras I have had needed two or three days to reach their best. Also many of these stayed at their peak for weeks or even months. Only very few lost intensity after two or three months, but it can happen, even with nitro! So bottom line: a few days (!) of decanting are allways beneficial, a couple of weeks or even months of decanting might pose some risk to older Madeiras.
And I really wonder if the results would have been different with a VINTAGE Madeira... To bad my cellar is not that deeply stocked with some ancient 18xx vintage Madeira to repeat this great experiment...
Best
Peter
PS:
Regarding the recorking, I use half bottles with almost no airspace left at all with very good long-term storing results. If there isnt enough wine to reduce airspace to minimum, I take sterile glass balls to bring the level up to the max.

[Steve Shapiro]

My thanks to the respondants. I'll incorporate the half-bottle idea in the future.

[Julian D. A. Wiseman]

Sounds like a good experiment.

I failed to understand the x/y/z scores: please could I have an explanation in shorter words?

[Peter W. Meek]

Number of votes for 1st/2nd/last place, from 23 tasters, who were voting on 5 bottles of the same wine, decanted with 5 different lead times before the tasting.

Could have been all five places, but the first two plus the last give the most useful information with the least clutter.

[Julian D. A. Wiseman]

Thank you. Let’s give points as follows:

• First = +2 points;
• Second = +1 points;
• Third = 0 points;
• Fourth= 0 points;
• Fifth = –2 points.

Then we have scores as follows:

• 3D: +18;
• 1D: +14;
• 6W: +13;
• 2½W: +12;
• 2H: –37.

Though the winner was the 3-day decant, the overall impression is really to decant Madeira for at least a day. More helps, sometimes, a bit.

Further, the absence of a clear ordering (if 3 days is best, why is 6 weeks better than 2½ weeks?) suggests that bottle variation might explain some of the observed votes.

[Steve Shapiro]

My monthly couples wine tasting group is a mix of talented amateurs, interested amateurs, and just plain folks who enjoy good wine and /or good conversation. About half fall into the first group. We frequently add an area professional to the mix, as we did for this madeira tasting. When I started the group 35 years ago many of the members had little experience with fine wine. To add to the mix, everyone has different preferences. Most months we have a few wines while people congregate, followed by six wines tasted blind, all with a known theme. To keep the scoring simple, the rules have been as follows: "You can vote as many times as you like for the wines you liked best, next best, and last." On occasions it's difficult to chose a last place wine, or to choose between two wines as a favorite. Since wine evolves in the glass, occasionally members vote as follows: "I liked it best early, but it faded over the two hours and was fourth later." And so it goes. The simple scoring system has served us well, and helps preserve the point of the evening -- to have fun. Cheese and bread accompany the bagged wines, a dessert wine and dessert follow. I hope the above helps answers questions about our scoring system.

[Julian D. A. Wiseman]

I was not at all criticising the scoring system, and even less criticising the idea of enjoying oneself. Just reporting what I thought that your data said.

[Steve Shapiro]

I didn't take it as criticism. I agree with your interpretation. Have fun.