Is npm deliberately lowering quality scores to favor certain packages?

[ericmorand]

This is a following to the discussion started there:

https://github.com/npm/feedback/discussions/66

I'm creating another discussion because my question is not related to npms.io or any other scoring tools, but to npm's policy itself.

I have the feeling that some packages are intentionally made harder to be found than they should. I have been discussing with a few developers recently about that - read some forum and twitter posts too - so I decided to check it by myself using my own twing.

Searching on the term "twig" and sorting the results by quality gives some results that makes me suspect that npm is messing with the quality score to favor certain packages against some others. Look at the first page:

https://www.npmjs.com/search?q=twig&ranking=quality&page=0&perPage=20

twig.js is the first result and twing is not even there.

They both mention the twig keyword in their package manifest, so we can remove this metric from the equation.
They also both come with README, LICENSE, a documentation website, a link to their repository and a link to their issue tracker. And they are both open source, of course.

So, let's focus on this:

• twig.js is not covered at all - i.e. it does not execute any coverage tool nor does it publish any coverage metric; twing is 100% covered and publishes its coverage to coveralls.io (https://coveralls.io/github/NightlyCommit/twing)
• twig.js doesn't embark TypeScript definitions; twing is written in TypeScript, is strongly typed, and publishes its type declaration as part of the package.

Keeping these into consideration, let's have a look at the quality score of both packages according to npm:

• twig.js: 92%
• twing: 70%

So, except if there is something obvious that I'm missing here and that could make the Twing's metric collapses to 70%, I can't help but think that there is something fishy at stake: twing is arguably a better quality package based on test code coverage and TypeScript definitions but it still lag 22% behind twig.js.

Can someone from npm jump in and explain what is happening here?

EDIT:

And to be totally transparent, I've been wondering for years why Twing was not attracting more popularity considering its tremendous level of quality - objectively. Now that I'm trying to search it using npm search and seeing how it is under-highlighted by npm, I understand why it is stagnating to 5000-10000 downloads a week: it is simply hidden from sight by npm. And this is a very sad discovery.