>Discussion will be about some piece of controversial research, e.g. GMOs or vaccines. People who want to discuss the merits of the research are shouted down by what I can only describe as 'science fanbois' -- anything that is 'against science' must be wrong.

Discussions on reddit (and all sorts of other social platforms) are being influenced and manipulated by people who are being paid to do so, and anyone who denies that at this point is delusional.

Source: I know people who run firms that provide this service.

> Some piece of controversial research, e.g. GMOs or vaccines

Who or what determines what is controversial?

To me, vaccines are not controversial at all. You can pick and choose specific ones to argue about, but the concept of vaccines is literally proven, beyond a doubt.

As for GMOs, there's a lot more room but even still, the fundamentals are nothing controversial. This is also proven in pretty much anything you can buy in a grocery store today.

Controversial literally just means that there's strong, polarizing disagreement. It's fairly well-defined, and vaccines definitely qualify. This doesn't mean that examining the actual positions of each side would lead you to conclude that they're of equal worth: it just describes the significant disagreement that exists. (As opposed to, say, the heliocentric model)

I don't think it's really about science, it's more about reddit. You'll find the same thing on any discussion on any somewhat popular subreddit regardless of the subject. There's almost always a reddit hivemind sanctioned Good Answer to any issue and if you stray away from the Good Path you'll get buried in downvotes, regardless of how much effort you put into arguing your point. Conversely if you make a completely useless one-liner comment agreeing with the Good Answer then you'll be showered with upvotes and propelled to the top of the page.

Upvote/downvotes are de-facto agree/disagree buttons on reddit. As a forum it's completely useless to discuss anything beyond cat pictures and funny gifs.

HN isn't all that different, so it's probably about more than just reddit.

For example, you can post the same remarks criticizing functional programming in a post about C# or Java and get upvotes, while the same post in a Haskell discussion will be downvoted. You can try the experiment yourself to verify it.

This applies to companies as well - you will be upvoted for pro-Microsoft or anti-Google posts in a news topic about Microsoft, and the reverse in a news topic about Google. This is because people who use Microsoft products will be very interested in Microsoft news, but not interested enough in Google news to go and downvote opinions they disagree with there.

> For example, you can post the same remarks criticizing functional programming in a post about C# or Java and get upvotes, while the same post in a Haskell discussion will be downvoted.

It's simply unreasonsble to think that the value of a comment is independent of its context.

Certainly, although it's clear that the largest part of the context would be the audience that the comment is exposed to. And it also shows that the upvote and downvote feature is not used correctly here either - it's used to promote or hide comments that the viewer agrees or disagrees with.

Just the same as Reddit. It's a deeper issue than just being 'reddit', and it's more about online discussion in general.

It's significantly better in my experience. It's not perfect but you can actually have a plurality of opinions on a subject. I think the fact that the community is tighter and that low karma accounts can't actually downvote help massively. Also no visible scores which reduces the "gaming" aspect. I kind of wish HN would go a little further still and ration downvotes for everybody, it shouldn't be something you throw willy-nilly.

That's true. Maybe the issue is that when there's only a 'down' and an 'up' button, people infer everything into it. Maybe we should have an 'up' button for interesting posts, a 'flag' button for posts that don't add to the discussion, and 'agree' and 'disagree' buttons to make it clear that it's a different thing from upvoting a comment.

Then people could turn on or off some way to see which posts are heavily agreed or disagreed with without downvoting interesting posts just because you disagree. You could even have a system where you mark people you respect and then only see which posts they agreed with to get a more personal feel? Might be an interesting discussion board idea to pursue.

The community is smaller, the barrier to finding the place are higher.

There are several factors that ensure better Signal to Noise ratios in communities.

Protective Moats, specific technical subject matter, for example.

Essentially it boils down to difficulty to engage, or non trivial effort required to engage.

The factors you mentioned have little impact on the behavior of the sub, which is mostly influenced by the people and norms of behavior, over the UI.

It's always fascinating to me how little some of these people understand science. There was a forum I frequented where there was a prolific "scientism" poster, and pretty much every time several of the working scientists on the site, myself included, were like "Um...that's not how this works."

I was referring to the work one has to go through to get your research recognized in the scientific community, for example. I'm not "worshipping" science.

I had in mind people who go on planes to disprove the earth being round, who believe crystals have actual healing powers, or people who refuse proven cancer treatment.