There was recently a link to a register article covering a scientific study where not just the title, but the whole URL got switched to the original study. Now, I'm not disputing that it probably would have been better to link to the article titled "Empathy among students in Engineering Programs" originally, but the Register article contained quite a bit of satire. It's title was "Engineers are cold and Dead inside".
The Register article contained slang and sarcasm aplenty, and most of the discussion revolved around that. When the URL changed, most of the comments suddenly became not only irrelevant, but incredibly confusing if you hadn't seen the originally linked register article.
So, I don't have a very strong opinion on article changes. But I'm pretty strongly against actually changing the URL. Yes, a lot of the time the submitted link is blogspam. But a lot of the time there's some added commentary or color, and people start discussing it. Changing the URL is like pulling the rug out from under the people who already commented. I know it's intended to elevate the content on the site, but it just ends up being kind of rude.
I also find it irritating when people call everything "blog spam". Many blogs are fluff, but either we go with the votes or we don't. If a link to a blog article is posted and someone thinks it's "blog spam", they should post another link to the original content, if the original content is better, more informative, more complete, it will get voted up. If it doesn't, the "blog spam" was "better" content, as voted by the community, like everything else is supposed to be over here. There should be no need for the patronizing moderation.
If we don't trust the votes to select better content what are we using them for?
The problem isn't fluff, it's blogs that act like a broken copy of the original with chunks missing, providing no commentary either. The voting system isn't a very good solution: it requires the front page to be inundated in duplicate articles followed by people reading both versions and voting up the good one. And what if they already upvoted the spam blog because they liked the information it had and didn't know it was a cheap knockoff?
The voting system lets you compare different articles, but it doesn't let you compare two versions of the same article submitted at different points of a day.
It is somewhat annoying when a link's "natural title" is not good and the submitter improves it - only to have it reverted.
It is somewhat annoying when a link has a misleading "natural title" but is not changed.
In short, I like editing when it makes sense and don't like it when it doesn't. Hence you cannot please everyone with either absolute, so I would say err on the side of least effort - therefore no editing!
This is exactly the problem. A good title on Hacker News is itself a valuable piece of content. A good title here will concisely convey the point of an article, so that each of us can decide if we want to check it out.
Often this is directly at odds with the purpose of a "natural title", which is designed by the author to entice and tantalize in order to drag eyeballs to it. Not to communicate, but to manipulate.
If the natural title is also good as a HN title, great. But if not, and if the submitter made a good effort to construct a good HN title, then it should not be changed.
I would go as far as to say that title changing by admin is lying to the users. Submitter decided on title, submitted with that title, and suddenly, 20 minutes later, the title is changed while submitter name stays the same. That created false impression that submitter said something what he didn't said.
Why stop at title, why not change comments also, if admins think they are not good enough for HN standards?
If title in bad, the article will be downvoted, or not upvoted enough. That's all regulation that is needed.
If you have to change the title, at least state clearly that the title is not written by the submitter, and leave original title available somewhere.
Generalizing that, I think it's a problem with the Reddit/Digg model of titles only (which HN followed), versus the Slashdot/Metafilter model of short blurbs. The former gives no space to explain why content would be of interest to HN, so is biased towards content where it's obvious from the headline, which in turn both favors fairly linkbaity things, and discourages most people from reading the /newest page at all (because you can't really tell much from it).
I've occasionally found longer articles where I think HN would be very interested in some of the discussion in a way that isn't obvious from the title, and might require 2 or 3 sentences to explain. I rarely submit those, though, because there's no good way to convey that.
Did I already look at this? I don't remember this title... Why's it grey? Oh yeah, I did read this!
This website isn't really a democracy. We, the users, can complain all we want, but until a) people are taken into the back alley when they bring it up, or b) enough people at the top get fed up with the complaining that they stop doing what they're doing, nothing will change.
Not that I care in slightest (not my problem, I'm happy as a clam), but the admins are actively doing something that many people don't particularly like, which seems like it would be easier on both sides if the action stopped (unless the admins enjoy the work).
Which one would save more of humanity's time - all the time spent complaining about title changing, or the time spent changing titles?
There are three titles in play: the article's original title, the title used by the submitter, and the title after a moderator change. I'm not sure if the poll is discussing changes by the submitter or changes by the moderator.
My datapoint on rewrites: I've had several of my blog posts submitted under a rewritten title. Usually I find the submitter's title is better than mine and I change the blog title to match.
Like most people, I believe misleading or content-less titles are bad. As such I'm in favor of keeping descriptive titles and editing bad ones - same as almost everybody else. There is no programmatic approach to this, and I doubt we can come up with rules that improve the process. In the end it's always going to be a human being making a judgement call.
One thing we reasonably could talk about, however, is whether the current bias towards keeping/restoring the external title is healthy. If we can agree that it's mostly detrimental, it would make a convincing appeal to emphasize more moderatorial restraint.
I haven't noticed a recent surge, and without evidence I'd be skeptical that there has been one.
That said, my general feeling is: it's probably anti-annoying most of the time (an option which you've left out), and sometimes somewhat annoying.
It's annoying when the submitter improves on the original title and a mod reverts it. But it's anti-annoying when the submitter uses a bad title and a mod edits it. I don't think I've ever explicitly noticed one of those, but I'd still guess they're the majority. (I only notice the annoying ones when someone complains.)
Edited articles (whether the title, URL or anything else was changed) should at least show who edited it, when, and what the previous content was. Then you can see that the username associated with the post did not really write that content (making it more authentic), and any comments relating to previous titles/URLs can still be understood by seeing the previous content (making it more relevant).
True story, the Huffington Post's CMS either did or still does A/B testing on all headlines within 5 minutes. Authors enter two headlines for each story, and after 5 minutes, the headline that's garnered the most page views becomes the singular hed for that article.
To be fair, usually when a title changes post submission it's to rename it from obvious link bait to something more reasonable. If people are discussing the link bait title instead of the actual content then that's all the more proof that editing the title makes sense.
The original submitted title was "Minecraft's Monster Profits". Minecraft was something I recognized and was interested in. When I tried to find the thread again to see what people had been saying, I couldn't see it and had to hunt around for a bit. Mojang meant nothing to me, and my eyes even kept misreading "Mojang" as "Mahjong", just to add another layer of confusion.
I commented in the Mojang thread -- my top-voted comment was bumped down to the bottom of the page after about 20 minutes. Actively discouraging meta-discussion of HN is a terrible precedent to set, in my mind.
Most annoying is we have technology to show information better.
We could show our original titles next to officiated, moderated, adworded language.
We could give first words and ad words equal life.
Do we have any more senseful, mindful technology that might represent title variations as diverse as real life?
Is our loved accessibility only for special interests? Why can't technology show common ground, without hiring and firing info like we are out of space, in a world we can use real psychotechnology to truly fit more people, more info, and more views.
If it is annoying admins to show more/less words based on eachothers diverse contexts, maybe let people who want to see moderated words see it so. Too, let us all otherwise real life readers see what people really submitted as their first words, first.