Is it a Myth? Decanting Wine

Conventional wisdom has it that wine breathes. Not only does it breathe, but this aspiration improves the wine. This has led to sales of decanters and any number of devices created to mimic or accelerate this breathing process. It is said that some, if not many or all wines, improve with exposure to oxygen, one of the components of air.

Time to test! It would be insightful and fun to do experiments involving a pure oxygen environment, but that is less practical than this simple procedure anyone can follow.

The quick answer is that on this night, with these tasters, and this wine,
there was no great variation between the wines.

Six identical wines were chosen. In this case a 2011 Domaine Bousquet Reserva Cabernet Sauvignon from the Uco Valley in Mendoza, Argentina, purchased at the winery. The wines were chosen for being more or less typical medium level young red wines with moderate oak and a few years of bottle aging.

The wines were blended together carefully, to reduce splashing - although they definitely were aerated to some degree during the process. The blending was done in order to remove bottle variation as a variable.

Four of the wines were immediately and gently, poured into nondescript bottles. Two of the bottles were closed, and two were left open. Two bottles worth of wine remained in open buckets at room temperature. After one hour all of the remaining wine was bottled and capped. All six wines were now in bottles closed with used screw tops.

Being Argentina in the middle of summer, room temperature at the start of the process was 33C/91F and the altitude was 1000 meters.

The bottles were marked with an indication as to their status: cap on, cap off or left in bucket. The bottles were then bagged, and mixed by two people independently, before being numbered.

The wines were then tasted double blind by a mixed group of professionals and consumers (primarily the spouses of the professionals). The group meets and tastes together regularly. I was out of the room, to reduce subconscious bias, but close enough to enforce a no talking rule.

The tasters were only told that the wines had been moved into nondescript bottles, and were given no other information. They were tasked with assessing the quality of the wines.

The tasters were given 15 minutes to taste the six wines, and then asked to rate the wines in order of preference. This was done by placing slips of numbered paper, one for each wine, into corresponding containers (you put your favorite of the wines into the 1st place container, your least favorite into the 6th place, etc).

The scores were recorded. The specific results are in the appendix at the end of the article.

The quick answer is that on this night, with these tasters, and this wine, there was no great variation between the wines.

While there may have been a trend that better statisticians than I may find, there were no obvious trends, and there was no dramatic difference (if any difference at all) between any of the samples.

Notably, the wines did not score in pairs. That is, the 2 wines left open to the air in the buckets did not score the same, nor did the other pairs. This may suggest that any difference between the wines may have been slight, or imagined - creating false positive indications of differences is common in blind wine tastings.

There were any number of problems with this experiment. The combining of the wines definitely exposed them to air prior to being re-bottled. The heat and even the altitude, may have made a difference. The fact that it was a single group that tastes together may be problematic. The experiment did not take into account placement bias, in that the same wine was poured into the same numbered glass for each participant. And finally, but not necessarily the only other problem, the use of this particular wine may have made a difference.

Future experiments, and I urge you all to conduct them, could alleviate some of these issues by having a larger number of tasters, chosen at random. A more controlled way of homogenizing the wines would be very useful. Various wines, old, young, tannic and or those with noticeable sulfur aromas, should be tested. The environment should be controlled more completely, and a procedure that allowed the wines to be presented in a different order for each taster may also be called for.

It was far from the perfect experiment, but it did what it was designed to do. It is an easy to replicate test that removes many biases and adds important controls.

That this particular time the test showed no dramatic difference between the wines should be a call for others to run their own tests. It doesn't prove that decanting and or aerating a wine is ineffective, rather it demonstrates a procedure for testing the hypothesis, and showing that much more testing is called for.

As an anecdotal aside, I have run this test with many groups over the years, and have always had similar results. This has affected my ability to believe in the power of decanting, but there should only be one way for you to decide if this experiment has any bearing on your beliefs or habits. Test, test, test! Question everyone, and everything. Especially me.

Appendix