Hacker News Reader

Comment by bulleyashah | Hacker News Reader

Let me chime in on the voting issue (or I think what you mean is mob voting issue?)

Imagine a discussion on Linux and Torvalds comments on it. But due to randomisation, it doesn't get enough traction. Note that this problem will become bigger the more top level comments you get.

This is the same problem with democracies also. Everyone gets to vote but the outcome might not be the best.But the alternative of randomly selecting people to govern also has its problems.

Do you know any platforms which have successfully done the randomisation thing?

View on HN

> Imagine a discussion on Linux and Torvalds comments on it.

This seems like a rarity. The vast majority of HN discussions don't have this situation, and it seems odd if the only purpose of HN voting is to upvote "celebrity" comments. There are much better places to follow the comments of celebrities than on HN.

> the alternative of randomly selecting people to govern also has its problems.

I strongly believe this is actually the least bad form of government, and vastly superior to elections, which are glorified high school prom royalty pageants.

> Do you know any platforms which have successfully done the randomisation thing?

No, though I have no idea what "algorithm" Twitter uses to determine the order of replies in a thread. (Probably not totally random.)

I didn't mean celebrity but more like knowledgeable people. A random system, just like random election, does not guarantee best or betterness in any form. It's not important however, because I agree it's unlikely and I do see where you're coming from (I also share your sense of cynicism about democracy).

I don't use Twitter that much but I hear that twitter is really toxic. Assuming some part of twitter tweet section is random, does that inspire confidence that such a system might work.

> I didn't mean celebrity but more like knowledgeable people.

That's my point though. A celebrity like Torvalds will likely get upvoted, but in my experience, non-celebrity knowledgeable commenters often get downvoted by people who are much less knowledgeable.

> A random system, just like random election, does not guarantee best or betterness in any form.

I don't think any system guarantees betterness. :-) But random seems to be at least pretty fair and least subject to abuse.

> Assuming some part of twitter tweet section is random, does that inspire confidence that such a system might work.

There are different parts of Twitter. The Twitter timeline is definitely not random. It's either reverse chronological or "algorithmic", depending on your settings. But any given tweet can have any number of replies, and I don't know how Twitter determines the order of display of replies to a tweet. But it's overall a very different format from HN, so comparisons are difficult.

I see. My only contention would be that abuse is stopped but use is also equally degraded (due to randomness).

But I see what you're saying. Some combination of voting and randomness might br worth it. Also, another thing is maybe some sort of sentiment analysis can help (abuse mainly comes from trolling, virtue signalling etc).

I don't know, if ther was a way to figure out what value a comment adds (or inverse), then that, combined with voting and some sort of randomness might make the system fairer and better?

Well, there's a lot of forms of randomness...

One could use Thompson sampling. Every comment starts with 1 up and 1 down vote. The total number of up/down votes determine a beta distribution. When displaying comments, draw from the beta distributions for each comment, and present them in that order. High quality comments drift reliably to the top over time, but other comments have their own chances at the top to accumulate votes and better determine their place in the stack.

Reply on news.ycombinator.com