It's Not A Bad Restaurant, It's A Bad Dish

You go to a restaurant with great reviews, but find the food completely mediocre. What happened?

Of course there's lots of options: other people have bad taste, the reviews are fake, the place went downhill.

But I think an underrated explanation is that there's high variance in quality within a single restaurant: the best dish the restaurant makes can be 10x better than the worst dish.

I figured this out while going out for group dinners where we shared a bunch of dishes and I realised 1) the best dish was extraordinary, 2) the rest of the dishes were mediocre, 3) if I'd gone by myself I would have only tasted one dish, 4) statistically it would have been a bad one, and 5) I would have then thought to myself "this restaurant is bad", where a more-accurate assessment would be "this dish at this restaurant is bad."

I think the pattern is more general: the variance in quality within a single producer is often extremely high, whereas (I think) most people implicitly assume the opposite.

I notice this often with my favourite authors (or perhaps better to say: the authors of my favourite books). I'll read one book that blows my socks off, and go around telling everyone what a genius the author is. But when I excitedly buy her next book, I disappointedly discover that it's just-ok.

Obviously this doesn't negate the author's genius – it's hard to write even one great book. But it does change e.g. how heartily I recommend the author, rather than the book.

The takeaway from all this is (perhaps too) simple: don't assume consistency from producers. If they make something great, their next thing might suck; if they make something sucky, their next thing might be great.

Subscribe to Atoms vs Bits

Receive our weekly posts by email