Hacker News Reader

Comment by ryandvm | Hacker News Reader

Mercy no. What ever happened to letting each comment stand on its own merit?

Hacker News is already afflicted with "celebrities" and their unbearable sycophants. The last we need are additional mechanisms to support the phenomenon. Hell, if anything, we should move more towards meritocracy by hiding authorship for the first 60 minutes of a comment's life...

View on HN

Although that works for some very trivial discussions†, it's unrealistic to think that every comment has easily quantified value without respect to the author's expertise. Not all opinions are equal. That's the kind of malarky that you hear from school districts that want to teach creationism and evolution as two equal theories. It's simply ludicrous to suggest that a comment from somebody who just read 40% of the Wikipedia article deserves the same standing as a comment from somebody who has devoted two decades of his life to the subject.

† The kind of discussion where each comment can stand on its own merits seems unlikely to be a good fit for Hacker News. That trait implies that the topic is shallow and there aren't any great insights to be had — if there were, the people who had poured years of their life into the topic would deserve more deference.

"Comments stand on their own merit" ≠ "all opinions are equal".

If comments can stand purely on their own merits, then one person's opinions must be equal to another person's — otherwise something outside of the comment matters. If the only thing that matters is what you say and not the actual expertise that causes you to say it, then your doctor telling you "You need to go see a surgeon right away" is equal to a drunk bum telling you "You need to go see a surgeon right away." That is obviously not true. Thus comments cannot stand solely on their own merits. QED

What you're saying now is reasonable (though I still don't agree that HN needs identities).

Your previous comment read as implying that from "all opinions are equal" followed "creationism is as valid as evolution", which is a total straw man since we're here using "equal" to mean "equal irrespective of the person advocating" not "equal irrespective of intrinsic merit". It is not because creationism is advocated by creationists that creationism is a poor theory.

There is no intrinsic merit to the statements "God created the world" and "Animals evolved from simpler lifeforms and ultimately from non-living compounds" — their merit is entirely extrinsic, deriving from research done by scientists. It's ignoring these extrinsic factors and "letting the statements stand on their own merits" that leads to creationism being taught in school — that literally is the way the school board members usually put it. That's what I'm getting at: The more context you get for a claim, the better. In technical matters, the identity of the person telling you something is very important, because the value of an expert's opinion is extrinsic to the actual words that comprise the opinion.

That sounds like an interesting idea.

If implemented, it would also be possible to use patterns of voting to estimate (to some degree) how likely a user is to up/down-vote based on the author alone. Maybe dye a user's name purple for "groupie"?

Of course, stylometric analysis would still give away many of the users here well before the sixty minutes is up. A real solution (to burn through all usability in the name of meritocracy) would be text rewriting ala "A Scanner Darkly".

User identification with the short comments usually found on HN might be difficult. The challenge still stands if I'm not mistaken: http://news.ycombinator.com/item?id=1199592

Your last remark is not incompatible with what OP asks for. Maybe HN could test that during a trial week?

60 minutes doesn't seem quite long enough. I'd look at perhaps 0.5 * mean ttl on the front page.

This! +1

I love reading HN because arguments and reason are all that matters. Spiced with some trolling here and there to make it fun. Trolls' messages end up gray anyway.

Reply on news.ycombinator.com