It has proven to me as a better guide for where a bottling is in its lifespan than any of the other metrics; “drinking window” “ready to drink” etc.
Yet it is much less prominent and harder to use. If there is a way to sort or filter your cellar based on this statistic, I can’t find it.
I’m sure the CT power users are all over this. But maybe some others will benefit from shining light on the feature. And maybe the collective wisdom can expose some ways to use the stat more effectively.
Aside from the significant data integrity issues noted by Tom and David, I doubt that my personal consumption preferences match those of the “average” CellarTracker user any better than my drinking window preferences. But it’s worth a look as another data point. What percentage has become your green light for opening bottles, and at what percentage do you start worrying about a wine being over the hill?
If you think that the lack of updating their consumption is a pattern among CT users I don’t think it should matter that much for the predictive use of that statistic. An estimator - even if biased - can be useful if its consistent.
I think people are best off figuring out where they like their wines on the maturity curve. I’m biased toward whites/sparkly on the young side, Rhones at age 10+, BDX in the 10-20 age bands depending on the estate.
You guys are funny. There are errors in the other direction as well (bottles sold, given away, dropped, or otherwise consumed by means other than drinking). You can draw inferences with certainty and make important decisions using sampled data sets with errors and bias. There’s math for it.
The thing that makes this stat more reliable than the alternatives in my view is that it reflects actions rather than beliefs, has more data points (ie. more people enter bottles than enter drinking windows), and the data points are more independent of one another (eg people’s drinking window estimates are often copied from critics rather than a personal belief).
A stronger criticism of the % consumed statistic is that other people may be following a “conventional wisdom” that does not align with your own personal tastes. I can understand that. But since I am often buying wines in small quantities and in styles and from regions and grapes that I have little familiarity with, the conventional wisdom is often better than no wisdom at all.
More than anything, I would say that when I check this statistic. I am looking to see if it shows under 30% consumed. When this is the case, absent better info, I will wait.
Just to reiterate what others are saying, for a large portion of users CT is a collection tool and not a consumption tool. So I think the consumption data for higher end wines will be skewed.
For the more obscure wines you speak of this might actually work. And I’m glad you pointed it out as I often drink less ‘popular’ wines myself. But these can also run into some distortion in that unless there is some sufficient number of users buying and logging their consumption of these wines in CT you might merely be looking at one or two drinkers preferences.
I do like that it reflects actions. Another issue is how many are cellaring more than they ever end up drinking. When the numbers are large, once you find the right percentage for your palate, the errors don’t matter.
Just went through a bunch of Bordeaux I’m familiar with and the 30% threshold works for many of them, at least for my palate. There are exceptions. I don’t think I’d prioritize it over advice from someone I know who’s recently tasted the wine. But it does look like a better metric than drinking window or ready to drink. Good tip!
I poked around on the CT forums for more info on this. It looks like people have been asking for a way to use this data more effectively (sort, export, filter) for years and the answer is something like “no you can’t do it currently / yeah we understand it would be nice / it’s not easy to do with our current data model and not a high priority”