Hacker News new | ask | show | jobs
by nzmsv 1644 days ago
Downvotes on HN are basically random.

I've had well-written comments downvoted into oblivion in threads that also included another comment stating the exact same opinion slightly differently. I've had zingy one-liners upvoted to 100 or so. I've had discussions where the back and forth statements alternated between negative and positive karma.

I'm treating it as a study in human behavior around communications. So far my takeaways are:

- a zingy one-liner that agrees with the zeitgeist will be upvoted to the sky

- the same type of comment expressing a contrarian view will be buried

- timing matters: if you get a few downvotes early on, the downvote brigade will help ensure a continued slide

- downvoting to disagree is natural (meme theory and all that). What's interesting to me is how can one overcome the impulse to downvote and potentially seed a new idea in someone's brain.

3 comments

> timing matters: if you get a few downvotes early on, the downvote brigade will help ensure a continued slide

I think the comment kind and timing matter together. If the comment was reasonable but contrary to popular views, it will usually attract early downvotes, but over time I’ve noticed the score will normalize as more measured people visit the story over the day.

You really nailed it. I agree 100%.

It still riles me, when well reasoned opinions are downvoted.

Edited to add: Your comment is a brilliant example and I really don't get why it's downvoted <shrug>

I downvoted it because "a few outliers" does not remotely imply "100% random". Votes are pretty well correlated with comment quality, and being able to cite extreme exceptions is not a refutation of that.
You are right, it's not really random, but it can certainly feel that way.

I can throw the pedantry right back at you though: I didn't quote a percentage of randomness so I made a true statement. Some degree of randomness is present (if nothing else, it would be the percentage of readers having a bad day at that moment). So there :P

> I didn't quote a percentage of randomness

Ah, you're right, I blurred it with the comment that said they agree 100%.

>so I made a true statement.

No, you didn't, and it's not pedantry to say "don't dismiss a rating system because you can find the occasional outlier". That would be a big mistake, and merits being called out. Even speaking loosely, none of that justifies calling it "basically random".

The test of a discussion forum is whether good comments are generally pushed up (and vice versa) and how well that does compared to discussion forums in general. The occasional pathological case does not suddenly make it worthless or "basically random".

It's interesting that we had a back and forth discussion here. If we were talking in person, this kind of error correction would be natural (me: "it's random", you: "this is not what random is", me: "oh yeah, I'm using it in the colloquial sense, how silly of me, my main point...") but comments don't have that immediate feedback. At the same time they are not essays: one cannot spend as much time editing and forecast every small negative reaction.

Thank you for an interesting discussion!

That's not what happened. You're wrong in the most charitable, colloquial sense too, and that should have been clear from my previous reply. You didn't just "goof" or speak casually, you mischaracterized what the voting patterns look like, or, at the very least, you need to back up your view with more than "hey! Look at these weird cases!"

"Basically random" looks nothing like what we see on HN.

I was actually half expecting some downvotes and thought it would be a cute bit of post-irony to downvote this to near-background color.
Ironic that your comment gets downvoted. The comments I see being downvoted usually are either offtopic, self-promotion, or ignorant.