Willat Effect Experiments With Tea

The Willat Effect is the hedonic change caused by side-by-side comparison of similar things. Your hedonic response to the things compared (e.g., two or more dark chocolates) expands in both directions. The “better” things become more pleasant and the “worse” things become less pleasant. In my experience, it’s a big change, easy to notice.

I discovered the Willat Effect when my friend Carl Willat offered me five different limoncellos side by side. Knowing that he likes it, his friends had given them to him. Perhaps three were homemade, two store-bought. I’d had plenty of limoncello before that, but always one version at a time. Within seconds of tasting the five versions side by side, I came to like two of them (with more complex flavors) more than the rest. One or two of them I started to dislike. When you put two similar things next to each other, of course you see their differences more clearly. What’s impressive is the hedonic change.

The Willat Effect supports my ideas about human evolution because it pushes people toward connoisseurship. (I predict it won’t occur with animals.) The fact that repeating elements are found in so many decorating schemes and patterns meant to be pretty (e.g., wallpapers, textile patterns, rugs, choreography) suggests that we get pleasure from putting similar things side by side — the very state that produces the Willat Effect. According to my theory of human evolution, connoisseurship evolved because it created demand for hard-to-make goods, which helped the most skilled artisans make a living. Carl’s limoncello tasting made me a mini-connoisseur of limoncello. I started buying it much more often and  bought more expensive brands, thus helping the best limoncello makers make a living. Connoisseurs turn surplus into innovation by giving the most skilled artisans more time and freedom to innovate.

Does the Willat Effect have practical value? Could it improve my life? Recently I decided to see if it could make me a green tea connoisseur. Ever since I discovered the Shangri-La Diet (calories without smell), I’d been drinking tea (smell without calories) almost daily but I was no connoisseur. Nor had I done many side-by-side comparisons. At home, I had always made one cup at a time.

In Beijing, where I am now, I can easily buy many green teas. I got three identical tea pots (SAMA SAG-08) and three cheap green teas. I drink tea every morning. Instead of brewing one pot, I started making two or three pots at the same time and comparing the results. I compared different teas and the same tea brewed different lengths of time (Carl’s idea).

I’ve been doing this about two weeks. The results so far:

1. The cheapest tea became undrinkable. I decided to never buy it again and not to drink the  rest of my purchase. I will use it for kombucha. Two of the three teas cost about twice the cheapest one. After a few side by side comparisons I liked the more expensive ones considerably more than the cheaper one. The two more expensive ones cost about the same but, weirdly, I liked the one that cost (slightly) more a little better than the one that cost less. (Tea is sold in bulk with no packaging or branding so the price I pay is closely related to what the grower was paid. The buyers taste it and decide what it’s worth.)

2. I decided to infuse the tea leaves only once. (Usual practice is to infuse green tea two or more times.)  The quality of later infusions was too low, I decided. Before this, I had found second and later infusions had been acceptable.

The Willat Effect is working, in other words. After a decade of drinking tea, my practices suddenly changed.  I will buy different teas and brew them differently. I will spend a lot more per cup since (a) each cup will require fresh tea, (b) I won’t buy the cheapest tea, and (c) I have become far more interested in green tea, partly because each cup tastes better, partly because I am curious if more expensive varieties taste better. When I bought the three varieties I have now I didn’t bother to learn their names; I identified them by price. In the future I will learn the names.

To get the Willat Effect, the things being compared must be quite similar. For example, comparing green tea with black tea does nothing. I have learned a methodological lesson: That tea is a great medium for studying this not only because it’s cheap but also because you can easily get similar tasting teas by brewing the same tea different lengths of time. I haven’t yet tried different water temperatures but that too might work.

I have done similar things before. I bought several versions of orange marmalade, did side-by-side tastings, and indeed became an orange marmalade connoisseur. After that I bought only expensive versions. After a few side-by-side comparisons of cheese that included expensive cheeses, I stopped buying cheap cheese. You could say I am still an orange marmalade and cheese connoisseur but this has no effect on my current life. Because I avoid sugar, I don’t eat orange marmalade. Because of all the butter I eat, I rarely eat cheese. My budding green tea connoisseurship, however, is making a  difference because I drink tea every day.

My posts about human evolution.

43 Replies to “Willat Effect Experiments With Tea”

  1. I get two take-aways from this.

    1) If I want to consume more of something that I don’t really enjoy, I should do side-by-side taste tests of different versions to find the better tasting version and develop a taste for it.

    2) If I enjoy an inexpensive product, I should avoid side-by-side taste tests so that I do not develop an aversion to the inexpensive version.

    1. David, yeah, that’s correct. I used to enjoy green tea less than black tea. Now I like it more. The bigger point is that here is a faucet of pleasure and here is how to open it. Something that produced a tiny amount of pleasure per day or whatever can now produce pleasure at a much greater rate. Sure, it will cost you but it will feel worth it and you will feel good about the whole thing.

  2. To me, the interesting question is Scrimshaw’s. Is one really better off deliberately engaging the Willat effect? You say you are happier but…

    I think we’d all agree that it would be a bad idea to engage the Willat effect for everything (no one but a millionaire can afford to be a connoiseur in their tea, their milk, their bread, their computer and computer chair, their house, their car, their book etc etc.), which implies that there’s some point at which the Willat effect would make one *worse* off. Where is that?

  3. As an aside, I’m pretty sure the Willat effect exists in at least some animals. My parent’s dog is certainly picky when it comes to dog food (though I suppose that could be a matter of the different types of food being categorically different), and some fruit trees will be picked bare by birds and other critters while others in the same orchard go relatively unscathed.

  4. I think this is a pretty interesting effect. I’ve experienced it for myself at coffee and wine tastings.

    I wonder how much it is related to context as opposed to an absolute scale of what is ‘better’. When you set up a side by side comparison there are a few implications:

    1. A grading is implied.

    2. If you aren’t alone, you risk social standing by not being able to identify the ‘best’.

    3. Costs are already sunk.

    It seems to me that these are the circumstances of a game that would naturally lead one to seek specific criteria for choosing the ‘best’. Perhaps other games are possible.

    Here are couple of examples from my own experience:

    I regard myself as something of a coffee connoisseur. In the past I’ve spent more than $100 on a single pound of coffee, and pre-ordered specific auction lots prior to harvest. I’ve tried almost all methods of coffee brewing.

    Occasionally I buy a coffee by mail order that sounds good by the description, but turns out to be a long way from my tastes – to the point of unpleasantness. When that happens, I have learned to start tweaking the brewing method, temperature, contact time, amount of coffee, grind, etc. to try to get the best out of the specific coffee – it regard it as a challenge. Sometimes I will experiment with blending. I’ve found that this approach leads me to suspend judgement and focus much more on the distinct qualities of the ‘bad’ coffee. Through the process, the bad coffee turns into a coffee that I understand, and can like because it has a place in my mental map. It can become something I would choose again for the contrast it provides to the supposedly better coffees.

    The point here is that by playing a different game, I am able to extend my preferences along more than one dimension of good/bad.

    Similarly with wine – I’ve had the opportunity to compare some pretty expensive wines on occasion, but even one bottle of the really good stuff would be out of my price range. So for the wine I drink at home, I think of experiences of drinking table wine at lunch time in the South of France.

  5. (sorry – tapped post prematurely)

    …and so I have tried to find the ‘best’ cheap french table wine to fit the context I am imagining – which for me turns out to be a $6 côtes du rhone. More expensive wines have different connotations for me – and although I like them, I don’t particularly want to drink them every day.

    Once again, my point is that by playing a different game, I am able to derive a great deal of pleasure using different criteria to grade what I consume.

    So my question is – how much of the perceived quality is built into the structure of the game and the context in which you play it, and how much is inherent in the product?

  6. Seth, I find the Willat effect fascinating, and I feel like I can confirm it from personal experience with beer. I do have a follow up question with Seth — what about the so-called experienced quality affect with price? http://media.caltech.edu/press_releases/13091 Wouldn’t you want to blind yourself to the cost of items to make sure this wasn’t having an effect on percieved quality?

  7. As for others concern about cost, keep in mind the additional utility/pleasure experienced the more often you use the product (basically unless you use something very infrequently you should see payback quickly.) If you are buying a lot of something, perhaps you should test it out to see if you like something else better, because you may be spending more money on something you don’t like as much…

  8. You say the Willat Effect isn’t helpful in the choice between, say, green and black tea. How does one know when two items are near enough to make the side-by-side comparison useful. I am thinking specifically of various beers. As you know, there are lots of beer “styles.” Would I want to use Willat within a style or would it also apply across styles? Or, all red wines, or only among cabernet sauvignons? (I assume no value in comparing a chardonnay to merlot. )

  9. I predict there is a normal distribution of people susceptible to the Willat Effect. In other words, a few people in a group could participate in tastings day after day and still not be able to discern a difference. Another small group of people are born with this capability; many times they become great cooks (whether in the restaurant business or simply renowned among family and friends). These people can read a cookbook and start wincing at ridiculous combinations, because they have stored flavors in memory. That’s why they can instantly tell you if a particular item they’re tasting right now is best in class to anything they’ve tasted before. Most people are in the middle and can be influenced by the Willat Effect, yet usually aren’t, unless they get caught up in a social event with somebody who has enthusiastically put together a tasting.

  10. I would second that the cheaper offerings can be identified as better.

    I have done several blind taste tests of Vodkas, and have consistently found that humble Gordon’s Vodka tastes cleaner than top flight brands like Grey Goose, and Ketel One.

    How is this possible? First, I guess the power of marketing. Second, apparently the Vodka industry is in large part supplied by bulk distillers that provide pure vodka to liquor brands. They all put their spin on the final taste, but since the starter material is basically tasteless, it’s more feasible for a low end product to carry through the tastelessness.

  11. Seth,

    Here’s the flaw I see in your plan.

    It has been established for quite some time that, in a rigorous scientific taste test, a person will choose the wine in the glass that they are *told* is the most expensive – even when the wine, in reality, came out of the exact same bottle.

    So, it doesn’t surprise me that you self-selected the most expensive item as being the best.

    In order for you to apply the same rigor and validate the Willat Effect, you need to have someone else purchase three kinds of tea on at least two separate occasions:

    – One occasion where they do not tell you which tea is which. You must select the best tea, and then it will be revealed which tea is at which price point.

    – A separate occasion where they tell you which one is which, but lie to you about it…giving you the cheap tea, and telling you that it’s actually the expensive tea.

    The Willat Effect is only valid if the taste differences among the teas is sufficient for you to pick out the best one consistently, and more importantly *blindly*.

  12. Saw the link to this on Boing-Boing (prepare for the deluge) – and I find this all very fascinating (certainly more so then the work I should be doing).

    Two comments from me:

    1. Hedonic? Ouch – even recognizing the root, I had to look that one up to be sure.

    2. I work in snow-ski development, and we exploit this effect when we test product on-snow, to help testers differentiate between a group of very similar models. 5 skis in a group, 5 skiers – we each take ONE run on a pair of skis and then switch to a new pair, eventually looping back taking two runs on the pair of skis that were skied 1st. Any good skier can adjust their technique to accommodate most skis, but by switching every run, you get a good “1st impression” of the ski. And even intermediate skiers that are convinced they won’t be able to feel differences are surprised that they DO feel significant differences between similar skis when tested in this way. And with on-snow testing, this process has the added benefit of making the test go quickly – useful for battling the effects of muscle fatigue & changing snow conditions. (BTW – ideally, these are blind tests with all skis sporting the same graphic).

  13. Fwiw, whenever I’ve done blind tests of wine (more than once, less than five times and typically with 3-4 different wines) I’ve almost always done a reverse ranking by price. Roughly, I’ve ranked the sweeter, less complex wines as better tasting than the more tart, complex wines.

    I don’t really enjoy wine, spirits or beer all that much though.

  14. I used to sell televisions and this effect is staggering. In the electronics store- side by side you can see the difference. but once you get the television home nobody notices those differences. Maybe not the cheapest TV, but the $5K version vs the $2500 doesn’t hardly make a difference (once you’re not in the store anymore)

    If you haven’t watched this TED lecture, much of it relates to this exact effect you’re talking about here.


    a couple things,
    #1 Never disclose the prices of what you’re comparing for accurate results.
    #2 Just pretending something is worth more will convince you it’s better

  15. Maybe I missed something but I really think this would have to be done on a blind testing basis. Or are you saying that the Willat effect really is just people having a tendency to try to be snobbish when they think they can get away with it, calling it intellectual curiosity.

  16. It does happen in animals. We have two cats at home. We feed them dry food and wet food. The cats notify us that they are hungry by waiting until we’re in the kitchen, then whining pathetically until we fill their bowls. Note that the whining only happens when we’re in the kitchen, and only before we’ve fed them – once they stop being hungry, they are content.

    Up until a few months ago, we only fed them organic dry food of a particular brand. They ate it happily. One week we ran out, and got the generic (non-organic) stuff from the corner store, and they loved it. Since then, when we give them the organic food, they sniff it, and keep whining for the “good stuff”.Eventually, one of two things will happen – they get hungry enough to eat the organic food, or my wife will break down and give them the non-organic. They clearly have a preference, in that they will go hungry rather than eat the inferior substitute. Before they knew there was a better alternative, though, they ate it without complaints.

    About a month ago, we ran into the same situation with the wet food, and yeah, they now do the exact same thing with the wet food.

  17. Here I have confirmed a finding of my own. Use the cheapest tea you can find for making kombucha.
    Off topic sure, but kombucha is mentioned in the article.

  18. As many others have noted, it really needs to a double-blind if you want to exclude price as a factor.

    CLB and others note that price matters a lot – and I think the study that CLB quoted makes it clear that this is NOT unconscious snobbery or social climbing – price is actually a flavor – it is perceived as taste at the sensory level. More expensive things taste better because rarity is delicious.

    The point in excluding price as a factor is not because it obscures reality (it *is* reality), but because you are stingy and cheap and want to minimize expenses 🙂

    Since all taste is arbitrary anyway, you could go all the way and cultivate an appreciation of cheap flavors.

  19. @Gnu – it was Meow Mix, wasn’t it? At my house, and at my friend’s it’s the same situation. Once they tasted the Meow Mix they didn’t want anything else, much to our dismay.

  20. I agree that finding that the most expensive brand of foodstuff is always best reveals a built-in bias. Taste testing must be blind as to brand and comparative price to eliminate this bias.

    Anecdotally, while spending time in Santa Fe, New Mexico, USA we visited various “Southwest” restaurants for lunch and dinner. The best one was away from tourist districts in a blue-collar neighborhood, serving a wide variety of authentic central-American foods from various cultures.

    It was neither fancy nor expensive, but the food was well prepared, the recipes were widely varied, the ingredients were top quality and fresh, wonderful.

    Can’t remember the name, couldn’t find it now without a lot of driving around that part of town, but it was really popular with the local folks.

  21. I would like to point out that your intention was originally to maximize pleasure, but your results indicate you minimized pleasure derived from at least one tea. It sounds like you used to be happy with any old green tea, but now you’ve limited the range of acceptable teas by nurturing a disgust for some of them. It is only compounding the problem that the ones you do still enjoy (and would have enjoyed regardless) are more difficult to access. To think, you could have been content with the cheap stuff if you’d never compared.

    At least you have derived pleasure from the experiment!

    1. Nick, I have vastly increased the pleasure I get from tea. I just spend more money to get it. It is easy to get more expensive teas and I can afford them. It’s true, it could go the other way if I couldn’t afford the expensive stuff I liked.

  22. Another reference:

    As blueberries go, the wild variety (“lowbush”) are the rock stars. The cultivated kind (“highbush”) can be planted anywhere, and grow in huge fields in places like New Jersey and Michigan. You’re as likely to find either topping your cereal. Sometimes you’ve got these big, fat berries bobbing in your milk, and other times you’ll have tiny bold nuggets on your spoon. Do a taste test someday. The cultivated ones are watery and mealy compared to the tiny wild ones—intense bursts of candy-like fruit. Once you notice the difference, you will never buy the fat ones again.

  23. I quoted your post in application to mathematics education: http://www.naturalmath.com/blog/example-spaces-hedonic/

    Apparently, there are some delicious complexities there. Check out this quote from a study: “Surprisingly, induction pro?ted from spacing, even though massing apparently created a sense of ?uent learning: Participants rated massing as more effective than spacing, even after their own test performance had demonstrated the opposite.”

    1. Thanks, Maria. I think of using the Willat Effect to teach young children science. It is fun to make the comparisons, the results aren’t obvious (e.g., how long to brew tea), and the comparisons are tiny experiments. It is mini-science.

  24. There are classic (in math ed circles) experiments where kids determine which of the mixes of orange juice and water will taste more like orange juice (for example, 1:2 vs. 3:4). The classic study is done on paper, presented as symbols or diagrams, which is sort of funny, in a sad way.

    In the math club, we once played with making simple shakes out of milk, honey and cocoa. That was a lot of fun, and good math. Please blog if you work with kids on tea experiments – I’d like to know how it works.

  25. Just a quick note: it seems like you’re working with a taste-level instance of a broader phenomenon discussed by epistemologists: we need to see contrasts in order to see similarities. If I have a very dark blue pair of pants, I’ll tend to see them as black if I wear them with a white shirt. But, if I lay out the dark blue pants against a black shirt, I’ll see it as blue. Same, even with the color black itself. When I lay out blue-black against red-black against green-black, I can see the differences against each other. But, if I take one at a time and put them against white, it will be harder to see. Same with the color white!

    1. Brian, that’s very interesting. You write, “if I lay out the dark blue pants against a black shirt, I’ll see it as blue”. That sounds like an instance of seeing similarities (dark color) brings out differences (blue vs. black). Not needing to see differences in order to see similarities (which is also true). An instance of that is comparing blue-black to white brings out the similarity between blue-black and red-black (both are dark).

  26. Interesting concept to apply on the spiritual level as well. Do we get used to certain thought processes that are higher in nature and therefore spoil us into keeping to those processes, therefore making us better people?

  27. That sounds like an interesting observation. I wonder whats the best way of making use of it? Developing a liking for tea is good if you want to drink more tea, but at this point in my life I dont see how I would benefit for developing connoseurship for wine, movies, or other trivial things.

    However, there can be certain qualities that would be very useful to develop.
    Developing a connoisseurship for clothing (something that most women aquire naturally through shopping) could allow one to rise in society faster. A connoisseurship in writing could allow one to become a better writer.
    The same could be said about speeches and any type of art or crafts. You could argue that developing a good taste in a field is necessary to advance in it (eg you need to be a connoisseur of indie films to make good indie films).

    But what other hidden willat effects could one exploit? Could you become a connoisseur of productivity? A connoisseur of conversations? Why not see if you could compare two days of your life to each other?

    In theory you could compare any two things so long as you had them side by side and were mindful of the differences.

    There’s so much potential here, where would you even start?

    1. The same could be said about speeches and any type of art or crafts. You could argue that developing a good taste in a field is necessary to advance in it (eg you need to be a connoisseur of indie films to make good indie films).

      Yes, good point. A friend of mine suggested that musicians be trained by being exposed to side-by-side versions of the same piece. His underlying idea was the same as yours. I like your generalization of it.

  28. Usually when I cook, I often ask for feedback from my wife and or other guests and compare their experiences. I try to elicit as much description as possible and try to force comparisons with past memorable shared dishes. I make small changes each time I cook mostly with substitutions and not volumes. I usualy see much more changes than they do, but I also miss some observations. In a tapas I made recently my wife and her father tasted a sardine like flavor independently and without the others awareness. I didn’t include any sardines but did use fish roe. Was there a genetic connection? I didn’t taste sardines either. I am a foodie but forcing them to make comparisons on a regular basis has made them foodies as well. I find with non foodies in general mixing multiple brands of the same thing usually gets the response of “it’s better than usual” . For example mixed rice, and mixed coffee brews.

Comments are closed.