TN: Two Rinaldi Baroli and Pisca Port with Oliver

[Brézème]

originally posted by David from Switzerland:

That's a misunderstanding, and I do not remember saying anything the like: there should be greater potential for greatness in general for modern-styled CdP than tradionally styled Piedmontese Nebbiolo. I'm confused. The potential for getting a category (a piece of work) right is per se 100% (or 100 points on a scale). A perfect cheese cake is no less or more perfect than a perfect snail... I'm sorry, maybe I fail to see what you're getting at.

Greetings from Switzerland, David.

Let's see :

originally posted by David from Switzerland:

Clos Saint Jean Ch“teauneuf-du-Pape Deus Ex Machina 2007
From 60- to over 100-year-old vines, a blend of 60% Grenache aged in tank, and 40% Mourvèdre aged in small oak barrels (in part new, in part one-year-old). Believe it or not, yours truly ordered this (one of the most fairly priced wines on the list, if an excuse were needed, and: “curiosity killed the cat”), even though I found the standard 2007 CdP perplexing. Was happy to find I liked it much better than the 2004 and 2005. Virtually opaque purple-ruby-black. Roast lamb (or beef). Roasted herbs (for once hard to tell if Provençal or other). Oriental-spiced, big and superripe jammy blood pudding fruit. The old-vininess shows to some extent. Not the ultimate in tannin depth, with touches of coffee, marzipan and malt. Soft and a bit blood-orangey acidity. Quite long on the finish. Faint viscosity. Meat-juicier with airing. Balanced in the modern sense of the term, by which I mean, a quantitative there is no telling if a wine like this can achieve anything like “harmony” with bottle age (Nick and I strongly doubt it, and would recommend drinking this while it gives so much up-front, mindless pleasure). Happen to find consultant Pilippe Cambie’s style weirdly reminiscent of a Stéphane Derenoncourt to Michel Rolland styled Bordeaux, in this particular case a bit like the CdP version of Canon La Gaffelière. Be that as it may, the wine has its undeniable qualities. Glad to have had a chance to taste it in this context (everyone liked it for what it is, no less, no more), leisurely and from fine stemware. Rating: ~94(-?)

and

Giuseppe Rinaldi Barolo Brunate-Le Coste 2006
Thanks to Oliver. About 75% from Brunate and 25% from Le Coste. Medium ruby. Rather closed on the nose. Dried rose petal, ethereal, pretty plum, herbs such as dried oregano, soft brown spice, faint tar. Faintly dry tannin only, nicely favourful, on the whole, the 2006 is still quite traditionally-styled. Lovely depth. More Burgundian than the 1996. Balanced, long. Beautiful wine, and promising, this should evolve well in bottle, and quite a long time. Wines like this are becoming rare. Fair QPR. Rating: 93+/94

Am I wrong understanding that if Pascal Morel can achieve more "tannin depth"
the wine would deserve an certain upgrade in terms of point?

I don't have the same feeling through your TN about the Rinaldi.

I might be wrong, but the general feeling I get from reading your TNs is that Beppe Rinaldi Brunate-Le Coste is some kind of achievement of the traditional style from Barolo, and that Deus Ex Machina is a raw piece of grenache that could achieve something better if getting more "Harmony" (though you're not sure it could).

Now if Chateauneuf is cheese cake, and Barolo is snail, I understand that comparing your pts is useless and even stupid.

But is modern chateauneuf also part of cheese cake family or do I have to treat it separatly?
If so, then I may compare your points within the Chateauneuf boys working with Cambie, but not with the traditional styled producers of the same place.
Tough work...

 

[David from Switzerland]

The problem is, it's partly the style that keeps a Cambie wine from being better, as you say. The question is not, do I like it, but is it good at what it is. That's why the tasting note is more important than the note, IMHO. There I express my personal preference. The rating is merely an acknowledgement of the product's quality. Note I don't like eating snails, even if we've got ones in the garden that I think are beautiful animals. Even so, I'm sure that if you served me a hundred, I be able to group them according to how good they are - being forced to eat a hundred would also imply I'd be fractionally more qualified to judge them on a qualitative scale. Having said that, having tasted something like ten to thirty thousand wines in my life doesn't mean I'm an expert, but it does make me feel as if I have more of a qualitative grasp on wine than snails. Now, to come back to your point, qualitatively speaking, I find it debatable whether these Rinaldis are closer to their ideal, i.e. a 1989 Collina Rionda or 1978 Monfortino, than that Deus Ex Machina is to e.g. a 1989 Célestins. Stylistically, of course, the Rinaldi is much closer to its peers, and the Deus a far cry from any Bonneau, be it good or bad.

Greetings from Switzerland, David.
_________________

J'ai g“ché vingt ans de mes plus belles années au billard. Si c'était à refaire, je recommencerais.“ Roger Conti

 

[Rahsaan]

originally posted by David from Switzerland:

originally posted by Jeff Grossman:

originally posted by David from Switzerland:
What it really means to me is something like "got the gist, on to something new".

Have you ever had a bottle so changeable, so elusive, so complicated that you felt you needed a second bottle to understand it?

Great question. It's tempting to think such a wine might exist. But no, I haven't.

Someone as careful as you doesn't find new dimensions of understanding by drinking multiple bottles of the same wine? (Not necessarily in the same evening).

It seems to me that three bottles minimum are useful for beginning to get a handle on a wine. The first two inevitably show different aspects and the third may give a sense of which direction is more reliable. But don't we all know how elusive this wine knowledge thing is?

That said, often I am fine only buying one bottle of a particular wine. Either because it doesn't inspire me or I just don't have enough time. But there's always more knowledge out there.

 

[Ian Fitzsimmons]

originally posted by Rahsaan:
... It seems to me that three bottles minimum are useful for beginning to get a handle on a wine....

I would have said at least two, at leisure, over a couple of weeks (or more), with different foods and time to reflect. Three would be even better, naturally.

 

[Ian Fitzsimmons]

originally posted by SFJoe:
It's a math problem. I wish the VLM were more supportive, since it's more his line, but really a linear scale doesn't work. You can project your vectors all on a single line and get to a scalar outcome, but you loose a huge amount of your information in the process. The same scale can indeed be used to evaluate a minivan and a Porsche, but the minivan buyer probably values the cup holders more than the Porsche buyer. And the speed through a turn may matter more to a Porsche buyer.

Plotting Muscadet and Chateauneuf on the same line leads to silliness, since it turns out that it's not really the same line, it's just that some of us pretend it is for a bit. As with minivans and Porsches, the wines have different utilities that should be recognized with different vectors.

/math

J

This is an excellent digest, if I understand you right, of the main point in the New Yorker article you cited recently. If I'm getting David's subsequent explanation right, he seeks to normalize his assessments of wines in different categories by mapping them, in effect, onto a meta-scale of potential: how nearly that particular wine approaches its own potential, which is represented by a best-in-class wine for the category.

This approach is a sensible one for ordering one's own thoughts about wine, and reminds me a bit of Matt Kramer's discussion of wine quality in Making Sense of Wine. But for a second person to use a rating like this intelligently, they must, first, understand the rater's approach, and then also have a lot of additional information about the categories and definitions of potential the rater adopts in each category. The normalized ranking alone - as you and Gladwell point out - is so stripped of detail that it is really pointless, appearances to the contrary notwithstanding.

And then you have the issue of observer bias, which Gladwell brings up at the end of his article, and is not addressed by the number of wine (or snail) samples taken by a single observer, especially where measures of quality aren't quantitatively operationalized in an explicit way.

I've always taken numerical wine rankings as an idiomatic expression of personal preference, which can be useful if you know something about the preferences of the person doing the ranking. Suckling's become a butt of some humor because of his "I'm X points on [such and such wine]" formula, but I believe his is a more honest expression of point ratings than allusions to undefined but objective-sounding quality frameworks like "perfect, classic, excellent, good," etc.

BTW, did you see the NYer article a couple of months ago about the pattern of apparent experiment result amendments with repetition? I forget what the author named this pattern, but the focus, in the end, was on observer error.

 

[Jonathan Loesberg]

This is getting to have the appearance of one of those Parker threads on points. Points with regard to wine, regardless of what the critics who bestow them might thing, aren't numerals and don't signify numbers. They are rather more like a painting of a numeral. They look exactly like numerals but nevertheless they are different. They are a rough and ready evaluative language. If Parker and/or David think that when they give a wine a 94, they are comparing it against some scale that would compare it relative to some other wine, even from the same region, as opposed to vastly different ones, that's because they are deluded about what they are doing. 94 means something akin to "this wine tastes pretty darned good," and should be taken that way. If he says that about both a traditional wine and a laboratory concoction all he's saying is "this wine tastes pretty darned good" about two different wines, after all.

As far as Deus ex Machina, I would say David's note was pretty clear about what was going on in the wine. I have a hard time imagining thinking the 07 would be better than the 04 or 05 if one found those wines too much of a muchness, but I've only had small tastes of the 05 and have never tasted the 07, so I can't say more about that. I doubt I would ever be able to identify these wines as CdP in a line-up (which may be my inability)and so they don't interest me much. But it's easy to imagine a set of preferences by which one might identify them as "pretty darned good."

 

[VLM]

originally posted by SFJoe:
It's a math problem. I wish the VLM were more supportive, since it's more his line, but really a linear scale doesn't work. You can project your vectors all on a single line and get to a scalar outcome, but you loose a huge amount of your information in the process. The same scale can indeed be used to evaluate a minivan and a Porsche, but the minivan buyer probably values the cup holders more than the Porsche buyer. And the speed through a turn may matter more to a Porsche buyer.

Plotting Muscadet and Chateauneuf on the same line leads to silliness, since it turns out that it's not really the same line, it's just that some of us pretend it is for a bit. As with minivans and Porsches, the wines have different utilities that should be recognized with different vectors.

/math

J

A multidimensional problem can be reduced to fewer dimensions or a scalar value (it may be called a priciple component score or a factor score, among other methodologies). What this does is find a solution through the parameter space that explains the most variance. Assumptions are made about the distribution of the scores along this plane and a scalar becomes the score. It can be a very useful way to look at high dimensional problems, but you lose granularity.

I actually mind number much less than many folks here. In fact, I'm more interested in them than "descriptors" which I don't really care about. A number would is fine with me. It expresses, pretty well, their qualitative judgment about the wine in question. This is a valuable data point.

 

[SFJoe]

originally posted by Ian Fitzsimmons:


This is an excellent digest, if I understand you right, of the main point in the New Yorker article you cited recently.

Thanks. It's also the argument I've been having with the vlm for years, but the math, or the analytic geometry if you prefer, is the same.

I've always taken numerical wine rankings as an idiomatic expression of personal preference...

That's probably the best that can be done with them, as in, "I like this one a little more than that one..."

The hitch is that David and some other assigners of points explicitly disavow this approach. In the thread I link to above, David gives a bunch of points to a wine he found undrinkable, so I think your proposal cannot be an account of David's behavior and intended usage.

 

[Jonathan Loesberg]

originally posted by SFJoe:

originally posted by Ian Fitzsimmons:


This is an excellent digest, if I understand you right, of the main point in the New Yorker article you cited recently.

Thanks. It's also the argument I've been having with the vlm for years, but the math, or the analytic geometry if you prefer, is the same.

I've always taken numerical wine rankings as an idiomatic expression of personal preference...

That's probably the best that can be done with them, as in, "I like this one a little more than that one..."

The hitch is that David and some other assigners of points explicitly disavow this approach. In the thread I link to above, David gives a bunch of points to a wine he found undrinkable, so I think your proposal cannot be an account of David's behavior and intended usage.

I'm aware, of course, that David and some others disavow my intepretation. I think they are wrong about themselves, though. David's analogy of grading papers doesn't quite work because teachers generally have a set of criteria for papers (clarity of thesis, logic of organization, clarity and correctness of language)that can be separated out and subjected to community judgment. Even here, of course, there is subjective difference in estimations. But giving points to a wine one found vile and undrinkable (as opposed to a style one didn't prefer) would be akin to lauding a paper for its style of organization when one found it to be a random hodge-podge of ideas. I don't think David does it, even if he thinks he does.

I'm not sure of your reference to the prior thread. I tracked down one of your links to a review of a modern Italian concoction. He has reservations about it, but he certainly doesn't think it undrinkable and he does use adjectives that praise taste.

 

[SFJoe]

originally posted by Jonathan Loesberg:


I'm not sure of your reference to the prior thread.

I had this in mind: "Remo was politely tight-lipped, whereas I could not bring myself to finish the second glass I was poured against my will."

I may have overstated. But personally I don't usually have the spontaneous experience of "being 90 pts on this" when I would also refuse a 2nd glass of the wine. But maybe that's why I don't go there.

 

[Ian Fitzsimmons]

originally posted by VLM:
...
A multidimensional problem can be reduced to fewer dimensions or a scalar value (it may be called a priciple component score or a factor score, among other methodologies). What this does is find a solution through the parameter space that explains the most variance. Assumptions are made about the distribution of the scores along this plane and a scalar becomes the score. It can be a very useful way to look at high dimensional problems, but you lose granularity.

...

Thanks for this refresher. I started to remember this technique after I posted, but managed to confuse it with multi-variate analysis, which, of course, is something completely different.

originally posted by SFJoe:

I've always taken numerical wine rankings as an idiomatic expression of personal preference...

That's probably the best that can be done with them, as in, "I like this one a little more than that one..."

The hitch is that David and some other assigners of points explicitly disavow this approach. In the thread I link to above, David gives a bunch of points to a wine he found undrinkable, so I think your proposal cannot be an account of David's behavior and intended usage.

Yes, right, of course. It's this aspect that made me think of Kramer, who talks (iirc) about connoisseurship as the ability to discuss the overall quality of a product knowledgeably and fairly, in a way (on a scale) that's distinct from one's personal preferences.

I guess this is where the NYer article's point comes in: without a detailed knowledge of the ranking protocol's component quality dimensions, their respective scales, and the weighting assigned to them in the final ranking, the results are idiosyncratic, and second persons using them may easily be mislead.

 

[SFJoe]

originally posted by VLM:

A multidimensional problem can be reduced to fewer dimensions or a scalar value (it may be called a priciple component score or a factor score, among other methodologies). What this does is find a solution through the parameter space that explains the most variance. Assumptions are made about the distribution of the scores along this plane and a scalar becomes the score. It can be a very useful way to look at high dimensional problems, but you lose granularity.

In your rigorous, fair and nonjudgmental statistical way, you seek the vector that explains the most variance. But many of these other systems of evaluation assign arbitrary weight to particular axes or components and will find a different set and order of scalar values.

 

[Florida Jim]

A wine can suck.
In can be delicious.
Maybe even, monumental.

Can it be premptory, navigable, spiteful or courteous?
Of course, these are non sequiturs; inappropriate language by which to describe wine.
That is what numbers are.

That said, I do agree with David that the note is the most important part of the post - from that, at least, I am able to extrapolate some information that may be of assistance to me.
The numbers are meaningless to me - I simply assume they have meaning to him.

Best, Jim

 

[Jay Miller]

originally posted by Florida Jim:

The numbers are meaningless to me - I simply assume they have meaning to him.

Best, Jim

True, though after the explanations they have some meaning for me.

For a while when I was jotting down letter grades I'd attach an addendum of exactly how I was using them.

 

[Florida Jim]

originally posted by Jay Miller:

originally posted by Florida Jim:

The numbers are meaningless to me - I simply assume they have meaning to him.

Best, Jim

True, though after the explanations they have some meaning for me.

My point, exactly.
By themselves, they communicate nothing.
Only through explanation is there any communication achieved. So I am pleased that David writes a note before he attaches a score. His note is articulate, the score is not.
Best, Jim

 

[VLM]

originally posted by Florida Jim:

originally posted by Jay Miller:

originally posted by Florida Jim:

The numbers are meaningless to me - I simply assume they have meaning to him.

Best, Jim

True, though after the explanations they have some meaning for me.

My point, exactly.
By themselves, they communicate nothing.
Only through explanation is there any communication achieved.
Best, Jim

That's not true. By themselves they communicate some sort of ranking. Gather enough of these and you converge on real information.

My entire life is based on this principle.

 

[Florida Jim]

originally posted by VLM:

originally posted by Florida Jim:

originally posted by Jay Miller:

originally posted by Florida Jim:

The numbers are meaningless to me - I simply assume they have meaning to him.

Best, Jim

True, though after the explanations they have some meaning for me.

My point, exactly.
By themselves, they communicate nothing.
Only through explanation is there any communication achieved.
Best, Jim

That's not true. By themselves they communicate some sort of ranking. Gather enough of these and you converge on real information.

My entire life is based on this principle.

A ranking that is specific to David, not you.
And what has been communicated by that? Only what he likes (or in this case, some variant on that theme).
And I doubt seriously that your life is based on this principal. Your occupation/career, perhaps, but I know you enough to know that rankings are not where you live.
Best, Jim

 

[VLM]

originally posted by Florida Jim:

originally posted by VLM:

originally posted by Florida Jim:

originally posted by Jay Miller:

originally posted by Florida Jim:

The numbers are meaningless to me - I simply assume they have meaning to him.

Best, Jim

True, though after the explanations they have some meaning for me.

My point, exactly.
By themselves, they communicate nothing.
Only through explanation is there any communication achieved.
Best, Jim

That's not true. By themselves they communicate some sort of ranking. Gather enough of these and you converge on real information.

My entire life is based on this principle.

A ranking that is specific to David, not you.
And what has been communicated by that? Only what he likes (or in this case, some variant on that theme).
And I doubt seriously that your life is based on this principal. Your occupation/career, perhaps, but I know you enough to know that rankings are not where you live.
Best, Jim

I meant the convergence of real information based on large sample sizes.

If I had scores on the same wines from everyone on this bored on the same wines, I could model that data and come up with some real information about this universe of wine drinkers and how "objectively" David's scores represent that.

If you can't measure it, it doesn't exist. ;-)

 

[Rahsaan]

originally posted by VLM:
My entire life is based on this principle.

So the stakes are high.

 

[Ian Fitzsimmons]

That assessment would depend on your valuation of Nathan's life, preferably expressed as a score on a 100-point scale :)