Label bias

[jdaw1]

At the recent 1963 horizontal I did not vote the NN63 in my top three of the weekend.

At and after the event it was suggested that those who did might have had ‟label bias”. At the event it was suggested that those who didn’t might have been trying to avoid the appearance of label bias.

Having pondered these comments, I have realised that there are at least three types of label bias.

1 (‟The fool”). Label says expensive or reputable or both. Therefore think highly of wine, and vote accordingly.

2 (‟Narcissus”). Label says expensive or reputable or both. Therefore claim not to think highly of wine, in order to show that one doesn’t have label bias (even though actually showing only that one doesn’t have type-1 label bias).

3 (‟Relativity”). Assume comparing two wines. One has a label that says expensive or reputable or both. The other doesn’t. The one is marginally the better wine, but under-performing reputation. The other marginally worse, but greatly out-performing reputation or expectation. Vote for the other really voting by ‘relative to reputation’, rather than absolute merit.

I don’t think I have type 1, nor type 2. Dissent welcomed. But I’m not sure whether I have type 3.

Are there other types of label bias? And does anybody wish to own up to any of these sins?

[DRT]

My name is Derek, and I have Type 3 label bias.

[g-man]

How about those who finding the label attractive thinks it merits more points.

[djewesbury]

I believe that one person can exhibit all three biases, depending on the wines, the company and the social circumstances.

A famous French sociologist wrote about 'cultural capital'. There are those who have it ('it' being the advantages conferred by educational and social privileges and opportunities, and so on) and who are very keen to demonstrate that they have it; they are probably typical Type-1. Those who are anxious that they should somehow disavow or renounce it (a silly thing to try, like pretending you weren't born with two feet) are probably Type-2: excessively compensating 'egalitarians' who know enough to know that they want to prove, loudly, what nonsense it all is.

But what is Type-3? A mealy-mouthed compromise. A failure to move beyond either previous position..

So what, precisely, is the unbiased position? Faced with a bottle of NN63 and the knowledge to understand what it is, how can one fail to fall into one of the three traps outlined?

The only answer is silver foil.

[DRT]

The only answer is silver foil.

Correct.

[jdaw1]

The only answer is silver foil.

Tin foil would surely suffice.

But even that has a big problem. It is not blind to the person who wrapped it. One can make things blind to everybody, or at least less well sighted. But that entails an intricate choreography of people being in and out of rooms, of decanters, and layers of labelling. It is more difficult, more complicated, and less robust. And even more so at a multi-session tasting we have never maintained blindness past the end of a session.

So the answer is to be able to afford silver foil and a butler and two under-butlers, to arrange everything for us.

[djewesbury]

The only answer is silver foil.

Tin foil would surely suffice.

So the answer is to be able to afford silver foil and a butler and two under-butlers, to arrange everything for us.

The butlers were implicit in my choice of silver foil. Sorry, I thought that was clear.

[LGTrotter]

I was thinking of liminality, in the sense of passing through some ritual as helping to describe the types of port drinker, rather than the more static ‘cultural capital’ idea. Even those with cultural capital need to be inculcated into the rituals of port. In my small experience those with the most ‘cultural capital’ have been the worst offenders of bias 1.
Bias 2 is less clear to me, perhaps this is reserved for those in the liminal state, ie those who have not yet passed through the ritual to an initiated state and stand upon the threshold.
I take a more prosaic view of bias 3; value for money, Nacional 63 may indeed be a very good wine, but I haven’t bought one yet because hell has not frozen over, the idea of spending over £4000 on a bottle of port is inexplicable to me. Having bought a goodly stash of Warre ’80 at very reasonable prices I think that bias 3 is explained by the smug feeling that comes from spending wisely. Financial capital that is, rather than cultural capital.

[djewesbury]

I agree and disagree. Bias 1, as you describe, is what I described. But I believe that those who are able to invert that position and display Bias 2 are already 'over the threshold'; they just prefer to show their 'individuality' by refusing to concur with those acclaiming the self-evident goodness of the Thing being examined (eg, NN63).

Bias 3: yes. I agree with this because to me, there is a clear, if loose, 'value chain' in terms of most port, whereby exchange-value can, in theory, be pegged to use-value. SW77 is acknowledged good port and commands a certain price but nothing above that because people know that for the extra they can buy something even 'better'. When we move out of the foothills of the value chain, however, exchange-value becomes linked to 'monopoly': that is, to exclusivity and, crucially, the perception of exclusivity. I can understand a port being 'better' than another and that translating into an exchange-value that is agreed (by all in the market) at £50, £100 or £200 more than that other. That relative superiority of the port goes beyond this sort of rationality though, once you deal with NN63 or similar (of course, nothing could possibly be similar to NN63 but there you are). I just don't understand how a port can really be thousands of pounds 'better'; and I don't really believe it can be.

[djewesbury]

Further thought, on reflection and after conversation. Bias 1 is also based on accepting a contract, breach of which would adversely affect oneself. X accepts the price of NN63, because he knows you have to pay for Things as rare and special as this. It would be pretty silly to admit that it wasn't worth it after all.

I think Bias 3 is not really a bias; it is rational - the assertion that something doesn't transcend the usual comparison and evaluation just because it is rare and special.

[LGTrotter]

I suspect you disagree because that is the kind of man you are; and rather marvelous it is too. I particularly like the backing up of the 'cultural capital' of Bourdieu with the phenomenology of Husserl in thinking about the 'thing'.

I find 'the market force' stuff much more humdrum albeit there is an obvious relationship here. However there are us 'bias 3' type buyers who try to look for the anomolies in this geometric relationship between quality, supply and demand, they are out there.

I too should like to disagree (cos I'm a bit like that too) with some of what you have said but I have to go and pick apples.

[djewesbury]

Orleans Reinette?

[LGTrotter]

Ashmeads kernel and rosemary russet. The Orleans karked it last year, boo.

[djewesbury]

Ashmeads kernel and rosemary russet. The Orleans karked it last year, boo.

Shame. Plant a new one. Ashmeads kernel is a good one though. Mind you that could be my Type-1 bias!

[LGTrotter]

Further thought, on reflection and after conversation.

I am agog to learn where you found someone to converse with on liminality, Bourdieu and the bias of port labels. You have rich gravy indeed in which to stew.

I was also thinking that never having tried a Nacional 63 I am spared the terrible knowledge of how great a wine it is. Surely this is some kind of bias, the refusal to accept knowledge from which there is no going back.

How about those who finding the label attractive thinks it merits more points.

There are wine labels which I find attractive, less so with port. I do find the sight of a Grahams label comforting though.

[djewesbury]

Oh yes very rich gravy indeed here.

A further thought. The NN82 sounds like quite an average port by any standards, based on reading every tasting note I can find (not that scores mean much but it tends to score below 90). And yet because it is NN, the price tag will still be around £300. That must be poor value. It would be interesting to taste and rank this; perhaps just as interesting (intellectually at least) as tasting a really 'phenomenal' NN like the 63.

[djewesbury]

I am sure it's not a refusal to accept knowledge on your part; if NN63 were in front of you I'm convinced you would drink it.

[LGTrotter]

I am sure it's not a refusal to accept knowledge on your part; if NN63 were in front of you I'm convinced you would drink it.

Busted.

Consider it more of a thought experiment.

[AW77]

I would like to add a new dimension to the "Label bias" discussion: vintage.
You talked about producers, but besides the producer the label states the vintage year. I admit that vintage does not matter in a horizontal tasting, but it might matter when judging a wine in a vertical or deciding in a wine shop which bottle to buy. For example: Should I buy the Bomfim 1996 or the 2001? People might use overall ratings of vintage years as a guide for deciding this question. But a generalization of a vintage year might be as biased as is the case with producers. The SQVP of a lesser year might be as good or even better than that of the better year, just because the producer might have blundered in the better year (and not in the lesser) or the weather conditions in the lesser year were not that unfavourable for this single quinta.

[DRT]

Bias 1 is also based on accepting a contract, breach of which would adversely affect oneself. X accepts the price of NN63, because he knows you have to pay for Things as rare and special as this. It would be pretty silly to admit that it wasn't worth it after all.

I think this is a different type of bias to that described by JDAW in his question. The scenario you describe can only apply to the person who bought the bottle, whereas the description (I believe) is aimed at the entire group who have been offered the chance to taste an expensive and/or legendary wine.

In the specific circumstances that prompted this discussion I am entirely confident that the purchaser would have no qualms about admitting that any port was not up to scratch regardless of how much he had paid for it. I know that because I have witnessed him doing so many times.

My comment about suspected label bias in relation to the NN63 was intended to apply across the group assembled that evening, not necessarily including the host. I have been lucky enough to taste this and other legendary ports a number of times in large groups and the reactions of individuals is always fascinating.

[LGTrotter]

A further thought. The NN82 sounds like quite an average port by any standards, based on reading every tasting note I can find (not that scores mean much but it tends to score below 90). And yet because it is NN, the price tag will still be around £300. That must be poor value. It would be interesting to taste and rank this; perhaps just as interesting (intellectually at least) as tasting a really 'phenomenal' NN like the 63.

A further thoought to your further thought; or at least nearly the same thought put differently, I notice a label bias among some against Cockburn, a wine that usually scores poorly but I love it for its intellectual or possibly unique interpretation of port. And the only Nacional I have come close to buying was the 82.

[djewesbury]

I think this is a different type of bias to that described by JDAW in his question. The scenario you describe can only apply to the person who bought the bottle, whereas the description (I believe) is aimed at the entire group who have been offered the chance to taste an expensive and/or legendary wine.

Yes and no. There is also the factor of not wanting to speak out and be the one who says "ew stinky...", for fear of losing standing. But I am generalising and not talking about the group that was gathered at this event...

In the specific circumstances that prompted this discussion I am entirely confident that the purchaser would have no qualms about admitting that any port was not up to scratch regardless of how much he had paid for it. I know that because I have witnessed him doing so many times.

Quite.

[LGTrotter]

I would like to add a new dimension to the "Label bias" discussion: vintage.
You talked about producers, but besides the producer the label states the vintage year. I admit that vintage does not matter in a horizontal tasting, but it might matter when judging a wine in a vertical or deciding in a wine shop which bottle to buy. For example: Should I buy the Bomfim 1996 or the 2001? People might use overall ratings of vintage years as a guide for deciding this question. But a generalization of a vintage year might be as biased as is the case with producers. The SQVP of a lesser year might be as good or even better than that of the better year, just because the producer might have blundered in the better year (and not in the lesser) or the weather conditions in the lesser year were not that unfavourable for this single quinta.

This is the space that label bias allows the wise buyer to sneak into thus;

there are us 'bias 3' type buyers who try to look for the anomolies in this geometric relationship between quality, supply and demand, they are out there.

And I'd go with the 96.

[RAYC]

My gut feeling is that expectation (not just the the "Type 3" reputation bias mentioned above, but also based on prior personal experience) has a lot to do with this.

Aside from label bias, the phenomenon of clusters at one end of the table finding favour with a port that is disliked (or simply unrated) at the other end of the table also happens too frequently for me to believe this is not a factor (this happens even when ports are served in foil).

But - boringly - things can be purely a matter of taste as well. Take the Dolamore 63 - three voted it as their first choice for the Saturday afternoon tasting, whereas eight had it outside their top three. I thought it was a good port, but personally couldn't understand it making the top three. Yet it is not a candidate inviting particular label bias, nor was it in the company of too many other ports inviting particular label bias (Taylor, and perhaps the RO ports excepted). From my memory of the seating plan, those who voted for it were relatively spread out amongst those that didn't. Beyond "some liked it and some loved it", is a theory really needed to explain this?

[jdaw1]

Upon reflection, bias cannot always be prevented by blinding, whether tin-blinding (complicatedly self-blinded) or silver-blinding (by a team of butlers).

I think that NN63 is meaty. Even if you disagree, hypothesise that the fancy schmancy bottle has a distinctive flavour.

Now what if the distinctive bottle is good, but not the best? Even in a fully blind tasting a moderately knowledgeable drinker could still exhibit type-1 label bias.

[RAYC]

Even if you disagree, hypothesise that the fancy schmancy bottle has a distinctive flavour.

or colour....

So - for a properly impartial tasting - tin foil, under-butlers, black glasses, and a silence rule are required. In the line-up can only be ports that you have not previously tried. Sounds fun.

[jdaw1]

So - for a properly impartial tasting - tin foil, under-butlers, black glasses, and a silence rule are required. In the line-up can only be ports that you have not previously tried. Sounds fun.

Not only not previously tasted, also about which the tasters have no knowledge of their reputation. Have never reading a tasting note. And no ports from houses (in a horizontal) or vintages (in a vertical) with a distinctive signature.

Our policy seems to be that, most importantly, tastings must be fun. We are not paid to work there, typically we pay (bottles or money) to attend. Tastings must be fun. Indeed, industry attendees seem to enjoy them for much the same reason. In so far as possible without compromising the pleasure of attending, we can try to be scientific. But the pleasure is more important than the science. I like that general policy, and encourage its continuation.

But that doesn’t prevent our discussing possible unscientific biases.

[LGTrotter]

is a theory really needed to explain this?

Well no,not really.

But that would be the end of the thread. So I suppose that even if it is pointless there would be some statistical analysis which could be done to show that indeed there was a credible theory to explain both the clustering of opinion in a room and the apparently random spread of it on other occasions.

The broad thrust of RAYC's post is spot on, and is incidently one of my main issues with tastings blind or otherwise; it's all pretty subjective.

And how did fun manage to creep into our deliberations?

[djewesbury]

This is an excellent thread. Very much good sense has been written and I'd venture that we're generally in agreement about the proper balance between fun, informality and attempted objectivity (the aim of which is that some of what we write here might be considered informative, or even, occasionally, edifying).

I will give the butler and both under-butlers the day off for our next meeting (just to see if that affects the outcome).

[LGTrotter]

I will give the butler and both under-butlers the day off for our next meeting (just to see if that affects the outcome).

I heard they were both on zero hours contracts anyway.

[djewesbury]

I will give the butler and both under-butlers the day off for our next meeting (just to see if that affects the outcome).

I heard they were both on zero hours contracts anyway.

Yes, and given to complaining about it too; which somewhat mars one's evening.

[jdaw1]

Should I buy the Bomfim 1996 or the 2001? People might use overall ratings of vintage years as a guide for deciding this question.

To me this seems confused. Two wines, A and B, of similar cost. Which should I buy? If I have no experience myself, I’ll seek a recommendation, which might come in the form of an average score, or a score from somebody I trust. Doing so is entirely efficient and sensible.

But if I taste both, which do I prefer? If my answer uses others’ recommendations and scores, that is label bias. Which is quite different to the previous ¶.

[DRT]

Agreed. I don't think AW77s example is the same as what has been termed label bias in this thread.

I think what AW77 is referencing is a persons tendency to purchase wines based on the reputation of a particular shipper or vintage, probably having never tasted the wine. I think we probably all do that and it is in my view completely logical to do so.

What is being debated here is a persons tendency or otherwise to claim to like or dislike a wine even though they have tasted it, and then consciously or subconsciously ignored the tasting experience and rated it according to its reputation or that of the shipper or vintage. To do so is illogical, but completely understandable.

[jdaw1]

I was thinking of liminality, in the sense of passing through some ritual as helping to describe the types of port drinker

This seems to me to be the wrong model. It is one-dimensional, that dimension being time, there is a stable before, a fluid liminal moment, and then a newly-stable after. But there are people at our tastings with plenty of Rhone-type cultural capital, or Champagne-type, Madeira-type or even tawny-Port-type. They have good palates in their established specialities, but a few might be new to Port. This is, for them, not a new ritual with all sociological gubbins that might be claimed to entail, rather this is a slight variation of a ritual already performed. There is no instability; the new stability is known in advance.

Further, those at the ’63 horizontal had passed through all this three-syllable stuff. At least, if they hadn’t, then the passage is barred.

So the model seems wrong.

Note: when discussing sociological gubbins it is advised to use words like ‟ritual” and ‟perform”. It doesn’t make any of it true, but the ritual of doing so is part of the performance.

[Andy Velebil]

Even if you disagree, hypothesise that the fancy schmancy bottle has a distinctive flavour.

or colour....

So - for a properly impartial tasting - tin foil, under-butlers, black glasses, and a silence rule are required. In the line-up can only be ports that you have not previously tried. Sounds fun.

Color is something that oftens skews peoples perception of a wine, especially in an older line up where one or two are much darker than the others. People see the color and assume it has to be the better Port simply because the color is darker.

[Chris Doty]

I like this topic very much, and believe that most wine tastings should be conducted blind. Obviously, if you're tasting with the producer, or are doing a vertical, you'd make exceptions, but as a general rule, blind and silent tasting helps tasters concentrate on what they enjoy about a given wine, which is how I feel it should be.

As for my personal biases, I would say I demand more from wines with established reputations, and in the event of a close call I tend to find in favor of the more memorable/distinctive wine, even if it may not be 'textbook' or 'complete.' I sometimes use WOTN votes as ways to help reinforce my memory of specific wines, especially in very large tastings, and in instances in which my top 3-5 wines are scored very narrowly.

I look forward to seeing the impact of some double blind / silent tasting at the Doty Cup 2014! Woo woo woo

[JB vintage]

I sometimes use WOTN votes as ways to help reinforce my memory of specific wines,

What is WOTN votes?

[jdaw1]

What is WOTN votes?

Wine Of The Night.

[TLW]

I volunteer to assist in a wide range of scientific trials.

[jdaw1]

Correct facts, wrong logic. We are terrible at guessing. But that doesn’t mean that my or your judgement of ‟do I like this?” is clouded by a label.

[Alex Bridgeman]

But that doesn’t mean that my or your judgement of ‟do I like this?” is clouded by a label.

I agree, but I do wonder whether my answer to the question of "what do I think of this?" might be affected by the label. If I am sipping Croft 1945, I expect to have my socks knocked off. If my socks remain on my feet I would be disappointed by the port. Even though it was an extremely good and enjoyable port I might still be disappointed, while the same experience poured from a bottle of Real Vinicola 1983 would leave me astonished and enthusiastic.

[jdaw1]

I agree, but I do wonder whether my answer to the question of "what do I think of this?" might be affected by the label. If I am sipping Croft 1945, I expect to have my socks knocked off. If my socks remain on my feet I would be disappointed by the port. Even though it was an extremely good and enjoyable port I might still be disappointed, while the same experience poured from a bottle of Real Vinicola 1983 would leave me astonished and enthusiastic.

3 (‟Relativity”). Assume comparing two wines. One has a label that says expensive or reputable or both. The other doesn’t. The one is marginally the better wine, but under-performing reputation. The other marginally worse, but greatly out-performing reputation or expectation. Vote for the other really voting by ‘relative to reputation’, rather than absolute merit.

[djewesbury]

Correct facts, wrong logic. We are terrible at guessing. But that doesn’t mean that my or your judgement of ‟do I like this?” is clouded by a label.

To whom were you replying here?

[jdaw1]

Correct facts, wrong logic. We are terrible at guessing. But that doesn’t mean that my or your judgement of ‟do I like this?” is clouded by a label.

To whom were you replying here?

A previous post that has since been deleted. Moral: always quote the correct extract (but don’t clutter by quoting the whole thing).

[SushiNorth]

3 (‟Relativity”). Assume comparing two wines. One has a label that says expensive or reputable or both. The other doesn’t. The one is marginally the better wine, but under-performing reputation. The other marginally worse, but greatly out-performing reputation or expectation. Vote for the other really voting by ‘relative to reputation’, rather than absolute merit.

I would plead Rebellious Relativity: A wine with a label that says expensive or reputable must significantly outclass it's lesser peers when placed in the mouth, or face harsher criticism than it rightly deserves.

[LGTrotter]

but as a general rule, blind and silent tasting helps tasters concentrate on what they enjoy about a given wine, which is how I feel it should be.

Although this is true if the sole aim is the analysis of wine I think that to have that much expertise in the room and be silent is a loss. Besides this the place of wine in culture is that it should promote sympathy between people. This sympathy comes out in words and not just about the wine.

I was thinking of liminality, in the sense of passing through some ritual as helping to describe the types of port drinker

This seems to me to be the wrong model. It is one-dimensional, that dimension being time, there is a stable before, a fluid liminal moment, and then a newly-stable after. But there are people at our tastings with plenty of Rhone-type cultural capital, or Champagne-type, Madeira-type or even tawny-Port-type. They have good palates in their established specialities, but a few might be new to Port. This is, for them, not a new ritual with all sociological gubbins that might be claimed to entail, rather this is a slight variation of a ritual already performed. There is no instability; the new stability is known in advance.

I would challenge this view of liminality. It is not an event, it can be considered that people, societies and groups within them remain in these luminal states for periods of time and even go back and forth. There is even the argument that certain groups are liminal in their nature, hermits for instance. I did not intend liminality to be an exact model, but rather a way of thinking about bias. I also maintain that ritual is endemic to the drinking of port over any other wine. There was a nice essay on this sociological gubbins by Suzette Heald which I can't find a reference for now.