beer

by

Vodka Illusions

Bill Deys recently wrote about a Business Week article stating that, in a blind taste test, all vodkas taste pretty much the same.

It was an informal test with a writer and a few friends. Without statistical analysis, it’s impossible to tell if the friends were guessing at an above-chance level or not (there was one correct guess about vodka brand during the trial, but who knows if it was based on taste or just a lucky guess). Still, the theory behind it makes sense; vodka is basically alcohol and water, without any oak barrels or extra ingredients being added, so differences would have to be subtle if they exist. And if people who claim to be able to distinguish one brand from another obviously can’t do so at all even in an informal test, differences can’t be as major as we’ve been lead to believe.

The implication here is that all vodkas are the same. Is that really true, though? I don’t think so. I’d argue that the appeal of a drink is about more than just the electrical signals going from our tongues and noses to our brains. It’s also about atmosphere, expectations about taste, preparation rituals, discussion of the drink with other people, etc. These factors are eliminated from a blind taste test, but present in real life. A blind test may reveal that vodkas are the same in the absence of knowledge about what brand is being drunk (drinken? drunken?), which is interesting information, but doesn’t exactly map onto real-life drinking situations.

In real life, the subjective experience of a drink is different depending on the brand. For some people, buying a $100 bottle of vodka, putting it in the freezer, garnishing it and mixing it with just the right amount of ice (or not) is more enjoyable than doing the same with a $20 bottle. Furthermore, it probably actually tastes better to them. It may be an “illusion” in the sense that the difference in taste is not purely based on receptors in the tongue and nose; but does it really matter if good taste signals are originating in the tongue or in the drinker’s own biased brain? No; a better taste is a better taste.


The problem, though, is if people knew that all vodkas were physically identical, they might have a harder time deceiving themselves into believing that “better” brands actually taste better. I guess that’s the difference between actual physical differences in taste and illusory differences; illusions can disappear as soon as one becomes aware of them. It’d be hard to enjoy a $100 bottle of vodka knowing that the stuff inside is the same as the stuff in the $20 bottle.

Luckily I’m not so into vodka after several pukey experiences with it, and I doubt the same lack of brand differences applies to more complex drinks like rum, scotch, wine, and beer. “Still”, a lot of the differences are probably all in our heads, and there is nothing wrong with that.

Here is a dog made of beer labels:

(from here.)

This is the 2nd post in an unintentional series of posts about the link between alcohol and psychology. See the 1st: Beer and Statistics.

Advertisement
by

Beer : Statistics :: Peas : Carrots

I once got slightly intoxicated while “studying” the night before a major exam in statistics. Normally this would not be something to be proud of, but the fact is, despite the morning headache, my mind was clear of distractions and all that wonderful statistical knowledge flowed onto the paper just as smoothly as the beer flowed into my belly the night before. I aced the exam and secured my future in psychology. (This is a story that Nick likes to tell whenever someone mentions exams and drinking in the same sentence).

It turns out there is a very good reason that beer and statistics go together like birds of a feather. The study of statistics has been linked with beer since its early history. Anyone with basic stats knowledge has heard of Student’s t-distribution, often used to tell if two groups are different from each other on some measure. Student was the pen name of William Sealy Gosset, a statistician working in Dublin. The dude chummed with some of the more familiar names in stats, like Pearson and Fisher.

The thing is, Gosset didn’t give a crap about discovering the inner workings of the mind by poking and prodding samples of unsuspecting humans. No, Gosset just wanted to use mathematics to brew tasty beer. He worked for the Guinness brewery, applying statistical knowledge to growing and brewing barley. Guinness wanted to protect this powerful secret knowledge from competitors, so Gosset was forced to publish under a fake name, and apparently more math-creative than naming-creative, chose the name “Student.”

So that’s how Student’s t-distribution was born. And that’s why having a few pints of Guinness before a major stats exam should be encouraged. Even if it results in failure – and very well might – mention to the prof that it was a tribute to the long and fascinating history of beer and statistics. That’s gotta be worth a few bonus marks.

…..

P. S. I hope you noticed the subtle normal curve in the picture of the Guinness up there. That took some serious Photoshop skills you know.