Caymus vs Pichon Lalande

Nice graphed comparison in today’s Eric Levine email between Caymus and Pichon Lalande CT scores over time.

I think it says something about the wines as he says, but also something about their drinkers.

Yeah, that was my reaction too. I do think seeing the fairly clear inflection point(s) for a particular wine over vintages is quite interesting.

Cool graphs, thanks for sharing.
Are you able to share the sample sizes for both?

The sample size is all vintages of those wine and all scores of all CT users for those wines.

Contrary to Al, I think its highly interesting and useful. Especially as it doesnt have to stop there: With all these data, CT can tell us which vintages are drinking great today or not, not only by where the scores are today or how they have changed/improved over the last months, but also by comparing them to where comparable vintages usual are at that age. How many conversations on this board or in general are about which vintages or wines to open or not to open right now. CT can deliver much better answers than all uf us here (or in your wine group) which always base our advice on an incomplete set of just a few wines we tasted recently instead of the entirety of the wines of a specific vintage/region/cru level out there.

That’s not true at all unless you believe that your taste should correspond with the average drinker’s, and you believe that the average drinker has a preference corresponding to yours when it comes to maturity. Some people like wines young, some older, and a collection of random opinions won’t tell me whether a vintage is drinking well today or not.

GregT - I think there is a nuance to that. Sure your exact grade might differ from someone else’s (or the collective score) but it seems like the trend behavior would seem harder to buck.

To me its like making the observation that professional athletes tend to have deteriorating performance after their peak season. Sure a Hall of Famer might have better performance at a given age than a run of the mill player, but both are subject to the same laws of gravity, and will have a dropoff in time.

It seems like those charts just illustrate that Caymus ‘builds’ their wines to drink well when when young, while Pichon Lalande - for perhaps a variety of factors - tends to taste better after some age.

Neat chart!

Gives a reason for more users to start putting numerical scores in notes.

Note that even that 12.5 year “peak” is not statistically significantly higher than the release date level, to the point I’m surprised that confirmation bias alone isn’t producing a bigger delta. That even people who self-select as Caymus drinkers don’t find any improvement from cellaring it past release is an interesting data point and probably replicable across many famous/expensive CA wines.

Eric can do that, not me. He didn’t say in the email that I recall.

To me its like making the observation that professional athletes tend to have deteriorating performance after their peak season. Sure a Hall of Famer might have better performance at a given age than a run of the mill player, but both are subject to the same laws of gravity, and will have a dropoff in time.

Someone forgot to tell Tom Brady! Must be the avocado ice-cream.

Yep. That’s exactly why I said I think it also tells us something about the drinkers of both wines.

Wonder how many data points they actually have for each wine… so many pitfalls when you do data analysis. Even more so on subjective scores.

+1

Not sure if there’s a general answer to this. It reinforces the need for words as well as numbers. Everyone knows their taste and probably interprets dat their own way.

Similarly with drinking windows. I think a more thorough way is to split up the overall window:
Primary / shut down if applicable / secondary / tertiary / senile

But maybe a vector in 5-space is too complicated?

12427 tasting notes for Beaucastel: https://www.cellartracker.com/list.asp?Table=NotesVintage&iUserOverride=0&Wine=Ch�teau+de+Beaucastel+Ch�teauneuf-du-Pape
8,813 tasting notes for Pichon Lalande: https://www.cellartracker.com/list.asp?iUserOverride=0&Pivot1=Vintage&Wine=Ch�teau+Pichon+Longueville+Comtesse+de+Lalande&Table=PivotNote
9,610 tasting notes for Caymus Cab Sauv: https://www.cellartracker.com/list.asp?iUserOverride=0&Pivot1=Vintage&Wine=Caymus+Cabernet+Sauvignon&Table=PivotNote

We have more than NINE MILLION notes overall, all written by the community.

I think that’s what CT does by default with Riesling on the “ready to drink” list, though it’s more of a 3-space not 5.

So Pichon-Lalande 2020 has average score of 96+
That has to be biased by future potential not current drinkability.
The data need to be grouped by duration which is what I assume the graphs did.

Somehow this talk reminds me of something John Paul of Cameron wines did many moons ago. He wrote a parody of Parker called the Art Advocate :

Guernica…harsh, angular…viewed nine times…59 pts
Renoir…rich, lush…etc…94 pts

Then there were two Russians who polled people on what they liked in art. The winners were:
1/blue sky
2/mountains
3/lake
4/naked women

So then they painted something with naked women cavorting in a lake on a sunny day with mountains in the background.
That reminds me of the company here that profiled everything Parker liked and helped you reverse engineer the process. You too could score 100.

To get to the point, let’s remember that Caymus changed styles around 2010. In the '80s and’90s their wines were tannic but now they have gotten smoother and perhaps sweeter.
Of course, all of Bordeaux has moved towards smoother/drinkable sooner.

I have to admit, Pichon Lalande and Caymus are two wineries I would never group together.

96+ for 2020 Pichon has to be scores from people who have tasted barrel samples, as i would assume the wine will not be bottled for a while.

On the contrary I think this year has shown that he’s not immortal and he’s dropped off pretty obviously. But in any event he certainly isn’t still improving… Just like Caymus.

Thanks Eric. Thats actually more than i expected per wine. And i don’t hope you took it as any form of critic. I am just a programmer working close with my companies data science team on a daily basis. So these type of analysis is of real interest to me

There’s also a bias related to who’s buying and drinking various wines. With some culty wines, for example. nearly all the notes are from big time fanboys. You can get a 96 pt average for what’s actually an outlier style, where a democratic vote might put it at 88. A bias along the lines of your comment could be that the self-selected typical Caymus drinker buy them for how they show young. They only know young wines, so tasting an older one they downgrade it because the familiar youth isn’t there. Contrast that to the self-selected PLL buyers, who (presumably) expect the wines to not show their best until they mature to their peak.