Expensive means good, right? Don’t rely on common beliefs: test everything! I did test this when I lived in France many years ago.
Every autumn, French supermarkets turn into “wine fairs”: huge areas dedicated to wine, stacks of boxes, nice catalogues, special prices, and a sommelier opening bottles and giving advice. Wine is sold in boxes of 6 or separately by bottle.
At one of these wine fairs I bought 80 different bottles. Some landed in my shopping cart on sommelier’s recommendation, but most of the bottles were chosen by me independently, at random, stupidly. All wine was French, all red, all dry, from different years and regions of France. Prices ranged from €2.50 to €60 per bottle. My car was full of wine.
I thought I should be expecting something like this:
I put those 80 bottles in my wine cellar at the right temperature, horizontally, comme il faut. I moved the cashier’s cheque into a safe place without reading, I forgot about the cheque for a while, and the experiment began. Lehaim!
I am not a connoisseur at all: a name or a vintage on the label tells me nothing, I couldn’t say if the wine being tested was expensive or cheap — this ignorance guaranteed the purity of my experiment.
My experiment lasted almost a year, not one night: I am not an alcoholic, I am an experimenter and a failed scientist (maybe in a next life). After finishing a bottle, without the cashier’s cheque, on a separate sheet pinned to the wall, I was assigning my appreciation to every bottle without knowing its price: e.g., “Château Babique 2006 Côtes du Rhône = -1.2”.
No stars! Stars are not numeric. A 1/5 star arithmetically gives you 20%, that’s wrong. It’s like a ‘2’ mark in a Soviet school: one thinks 2/5 = 40%, but in reality it’s 0.
My appreciation scale was mathematically correct with 5 meanings: “2: Excellent”, “1: Good”, “0: Fair”, “-1: Not really”, “-2: Shit”. Decimals were allowed.
When I finished this multi-months alco-tour, I retrieved the old yellowish hardly readable cashier’s cheque and made a spreadsheet putting together my appreciation and the price in EUR. I’ve got 80 observations. Easy enough, then: an XY-scatter plot, a linear regression (least-squares method), the correlation formula and the coefficient of determination 𝑹². Basic statistics I learnt at 16.
Hypothesis:
𝒙 (Price) is the independent variable; 𝒚 (Quality or Appreciation) is the dependent variable, i.e.: 𝒚 = 𝒇 (𝒙)
Then, 𝒄𝒐𝒓𝒓 (𝒙, 𝒚) = ? 𝑹² = ?
The result shocked me. It looked like this ↓
Conclusion: there was no relation at all between the price and the quality. A wine at €2.50 could be “Excellent”, and a wine at €60 could be “Shit”. And vice versa. No relation, no correlation.
Before I started the experiment, I had mixed feelings. Social pressure was pushing me to accept the common belief Expensive ⇒ Good. My deep intuition was quietly telling me that there was no link, that 𝒙⇏𝒚, but the social pressure is strong, isn’t it? Ideologemes, again… My blind experiment proved statistically and nearly scientifically: there was no relation between the price and the quality.
I was very happy to have proven my initial deep intuition.
Unexpectedly, another correlation was accidentally detected: the further south the wine was from, the better it tasted. Languedoc beat Bourgogne in all observations.
Conclusion:South>North.And another surprising finding: the age of the wine made no effect on my tongue.
Conclusion:Old≠Good.
A good friend of mine, a scientist from CERN and a wine connoisseur, later criticised the methodological flaws in my pseudo-scientific analysis: too narrow sample, ageing of wine was not respected, the 𝒙 price scale was not distributed equally, my non-double-blind experiment could be biased by the label, my subjective “taste detector” was not enough calibrated because of lack of oenological experience, etc. My friend’s criticism was entirely justified, I fully accept it: it’s not a science. But this correlation was zero, goddamn! Not 0.1, not 0.05, but zero!
It was an excellent experience anyway, despite the flaws. It taught me a lot. Since then, I know for sure: go for the thing you like, whatever the price, whatever they say, whatever the rating, whatever the shit. Go for your preference and listen to your heart!
Thank you, Mother Mathematics, thank you Father Common Sense, I love you both!