Popularity over quality
· 5 mins read
The web is driven by popularity. Not quality.
A story about statistics in Open Source
Both packages are open-source, but we can see some differences of statistics.
There is a pretty big difference in popularity about these packages. As the time I’m writing this article,
- the one from Uber has 3106 ⭐️ stars on Github
- the one from Alex has 461 ⭐️ stars on Github
Uber’s repository is way more popular than the one from Alex. It’s easily understandable, Uber has way more visibility than Alex.
But if we take a closer looks on statistics, particularly the number of downloads on npm, Alex’s package is downloaded near twice more per month than the one from Uber:
- Uber repository: 12k/month ⬇️
- Alex repository: 21k/month ⬇️
I’m not an expert in React and I don’t know Mapbox at all, so I can’t try the two packages to make my own opinion (which is the best way to have a good and unbiased opinion by the way, even if it’s take time). But I trust more the number of developers who download a package to use it, than a simple click on a star on Github.
So, for you, based on numbers and metrics, which one is the best library for Mapbox and React?
A lot of company try to make their products trustable and popular based on numbers and third-party services.
Don’t we all receive messages from friends about posting rates and recommendations about their new product to help them reach more people?
Don’t all Youtubers ask us to like, comment and share their videos?
Why don’t you think people on Instagram attach #tagsforlikes, #likeforlikes and #likes4likes to their Instagram photos?
It’s because of the spawn of an entire exchange economy where people pimp themselves out to the world, offering to repay insincere engagement with equally insincere engagement.
I can understand this way of doing. We all know that algorithms rule the world wide web now. And algorithms only understand numbers (yet). So we have to adapt and make our business the prettiest for the machines if we want to have the more chance on our side to succeed. The more data we have, the more the algorithms like us.
But, even if this way of doing is made to enhance quality, it’s not the perfect solution.
Here a personal example.
I try to give recommandations on Google Local Guide about places I visited. I always try to explain my opinion, which is just an opinion by the way. Who are you to say it’s not good? Better say you don’t like it or it doesn’t match your needs 🙃.
OK, back to the story. So I try to give honest and detailed opinions. One day, I noticed a guy who was at a very high level (something about level 13) on Google Local Guide. So, I visit his profile and find a lot of bad quality reviews (“good”, “average”, “bad”…) and bad quality pictures. Why? Because on Google Local Guide, you earn points mainly by doing actions, like giving reviews, add pictures…. You don’t earn points by the quality of your actions.
Strangely, this guy didn’t add places, or edit information on places. He just rated a lot of places. Of course, it’s harder and take more time to add useful information than simply rate a place by clicking on stars. 😓
Unfortunately, knowing that, I always unconsciously think if there are very good rates on a place or a product, it’s because the guys who posted these reviews are friends with the business owner. At the opposite, there are so many people who complain on recommandations. Or unlike, so many people try to give such a good opinion in their reviews that make them look fake… It’s hard to find people who take the time to give honest opinions.
I use a lot Booking. And I think Booking has a good rate system, even if they use a lot of dark patterns on their website (there is always someone who visits the same hotel as me in the same time 🙄).
In fact, the rate system of Booking is good because it’s a form, and the user is limited.
You have to answer categorised questions, and the rate is automatically created. You can write a review to give a detailed comment. But instead of a simple textarea like on a lot of websites, you have to fill 2 textarea: one for what you liked, and the other one for what you dislike.
I think it’s a good way to more qualify the reviews, because the user is limited and must justify his rate, even if it’s always possible to create biased reviews.
I think the solution will come from ourselves, human, and not from a machine or an algorithm.
We need to be aware of all the ways business are trying to screw us over and try to counter them:
- We do not have to rely on the magnitude of the ratings.
- We have to stop reading the cherry-picked reviews. If it’s possible, it’s better to sort the reviews from newest to oldest, to get an up-to-date and hopefully unbiased selection.
But these human verifications take time. And it’s easier to quickly look at rates to try to determinate if a product or a place is good than trying to take time and search about information on a product/place.
We always have to use our human brain to verify the real quality of something. Because we always have to remember algorithms rules the web… And at the end, it’s always 👱 vs 🤖.