I’m a big fan of the crowd. I often rely on the opinions of other people when making certain kinds of decisions. For example, if I need to find a great restaurant in an unfamiliar area, I use Yelp. If I need help shopping for quality products, I use Amazon. And, if I’m trying to look for the right hotel, I’ll use TripAdvisor.
All of these companies have one thing in common—reviews and lots of them. They all have individual and aggregated ratings for the products and services they cover. And, almost always, the crowd gets things right.
If I need to buy a good pair of scissors on Amazon, I can look at the overall rating and some individual reviews to help me make my decision. This works because the crowd has more information about scissors than I do. I’m no expert when it comes to scissor quality or manufacturing, so I will happily defer to the crowd.
But, there have also been moments where the crowd failed me. And every time this happened, the reason was the same—the crowd did not have sufficient expertise. They simply didn’t know enough about the topic at hand.
The wisdom of the crowd only works when everyone knows something about the thing they are evaluating. For example, if I asked a group of Americans to guess the weight of a Tesla Cybertruck in pounds, the answers would vary. Some might guess as low as 1,000 pounds while others guessed as high as 30,000 pounds. Nevertheless, I would expect the average of their estimates to be pretty close to the actual answer (which is 6,600-6,900 pounds depending on the configuration of the Cybertruck).
But if I asked that same group of people to guess the weight of a Cybertruck in grams, I’d expect the average answer to be far less accurate. For the record, 1 pound is 453 grams (yes, I had to look it up), meaning that a Cybertruck weighs about 3 million grams.
The reason why the crowd fails when asking in grams vs pounds is because of a lack of expertise. Americans (myself included) don’t work with grams often enough to give an accurate answer with this measure. As a result, the crowd would unintentionally mislead you.
This explains why I have a problem with Reddit, especially when it comes to financial advice. I’ve seen far too many cases where the crowd upvotes an answer that is either wrong, financially dangerous, or both. Exhibit A:
Following that post in September 2017, GE stock declined by 75% (at its lowest) and went on to underperform the S&P 500 by over 100% through today:
I would’ve recommended diversification, but diversification rarely gets upvotes.
When it comes to taking financial advice, you have to be careful whom you ask. If you want to know which bank has the best customer service, the crowd will provide a great answer. But when it comes to topics where some knowledge is required, you have to rely on experts.
Most recently I experienced the ignorance of the crowd during my trip to Japan. For those that don’t know, the thing that I spend the most money on every year is dining out. I don’t spend money on fancy clothes or nice cars, I spend it eating at restaurants (mostly in NYC). As a result, while traveling I try to find the best dining establishments possible.
However, this creates some pressure. Since I only have a limited number of nights to eat out while traveling, I have to be very selective about where I go. With only 9 nights in Japan, this means I would have just 9 dinners (or 9 chances) to eat at the best restaurants I could find.
When traveling, finding the best restaurants usually isn’t a problem for me because the crowd tends to get things right. However, when it came to Japan, the millions of tourists that have visited over the years couldn’t help all that much.
The reason why is two-fold. First, virtually all of the food options in Japan are high quality, even the inexpensive stuff. I had a $3 egg sandwich at 7-Eleven (yes, the convenience store) that was incredible. It had a creamy, light texture unlike any egg sandwich I’ve ever eaten before. A reader of mine described Japan’s food culture as the “excellence of the ordinary.” And I couldn’t agree more. Typical food in Japan is the 80th percentile everywhere else.
But, this high quality has a downside—it’s much harder for the crowd to differentiate between food that’s pretty good and food that’s amazing. The problem is that most people don’t dine out often enough or live in a city with a vibrant enough food culture to know any better. As a result, when people visit a place with a thriving food scene, they are blown away by just about everything.
I can’t tell you how many reviews I read for restaurants in Japan that said, “BEST SUSHI EVER!!” (or something similar). I don’t think any of these people were lying when they left these reviews, they probably just never had top tier sushi before. Because of this, I was unable to separate the great restaurants from the best restaurants in Japan based on online reviews and data.
Before you judge me as some sort of food elitist though, take a moment to understand my plight. New York City is considered to have the best food scene in America. It has 72 Michelin Star restaurants. Tokyo, on the other hand, has 170, which is the most of any city in the world. Kyoto has 100. Osaka has 85. The third most popular city to visit in Japan has 13 more Michelin Star restaurants than NYC! While the number of Michelin star restaurants in a city isn’t the only measure of its food scene, it’s definitely correlated.
This is why the crowd wasn’t as helpful as I would’ve hoped. There are simply too many options. So, what did I do instead? I listened to people I trust for my food recommendations.
The best steak I had in Japan was recommended by David Chang, the famous chef and restauranteur. The best burger joint came from a friend who is a huge foodie. And the best sushi came recommended from a wonderful chef I met in NYC. I put my trust in experts and that decision seems to have paid off. Though the crowd gets most things right, when you need to go a bit deeper, look elsewhere.
Merry Christmas Eve and thank you for reading!
If you liked this post, consider signing up for my newsletter.
This is post 430. Any code I have related to this post can be found here with the same numbering: https://github.com/nmaggiulli/of-dollars-and-data
