I agree that this is quite annoying. I can't count the number of times that I found a truly awesome article that I hadn't read before, I submitted it, and found that it was submitted 3 days ago but had a single upvote (probably a day later).

Here I take a fundamentally different attitude to jacquesm. Please, if you submit a truly awesome article, the kind you know everyone will love, please, TELL SOMEONE ABOUT IT. Otherwise, unless it comes from an already popular news source that will get multiple submits, it will probably sink into forgetfulness.

One place to tell people about it is, as has oft been mentioned, the #startups channel on Freenode. Another place would probably be on Twitter. Here's a pledge: if you @swombat me a really awesome submission and I happen to see it in time, I will upvote it (if it's not worthy of an upvote, I will try to tell you why in 140 characters).

It seems that once something scrolls off the first page or two of 'new', or sits too long at a low count, there's no chance of it catching a flurry of new votes -- and even if it did, I don't think the rating function considers momentum enough for it to reach the front page.

(That is: if a 3-day-old article that was initially overlooked gets 4 votes today in rapid succession, while a brand-new article gets 3 votes today in the same period, the older article is still punished by averaging its new votes over the days when it wasn't even really being reviewed. At least that's how it seems to work.)

There needs to be a better way for items to get a second wind when the right readers show up.

A possibility would be to replace 'new' with 'newly worth a look'. Essentially, this would be the combination of new items (which have an implicit upvote from their submitter) mixed with older items that just got a fresh upvote, minus items that are already on the front page. That is, every upvote would bump an item back to the top 'worth a look', even if not to the 'top' rankings, so it has a chance again of a self-reinforcing rise.

Probably, once a user has upvoted or 'passed' (hypothetical new option) on an item, it would no longer appear in 'newly worth a look'.

I don't think the system needs changing - sometimes submissions just get unlucky and/or have a bad title. Just try again a while later, or hope someone else does. Nothing wrong with that as long as you aren't spamming the post (imo).

My highest voted submission of recent times was an article I thought was great, that someone else had submitted a few days earlier, which hadn't got any upvotes. So I made the title more appealing and tweaked the url, hey presto 100+ upvotes. The system works! (well enough)

Could also have been a better title. What title did you use? The repost improved on original page title in a small but useful way. There is a lot of value generated by a good title.

I took the guidelines on titles to imply that if they were adequate then leave them alone. Perhaps I'm misreading them. http://ycombinator.com/newsguidelines.html

just have a bunch of friends vote your stuff up....that's what most people do. Otherwise it's really a crap shoot on whether or not you'll make it to the front page, unless you are submitting something that other people would try to submit(i.e. techcrunch articles)

This kind of defeats the point of any of these peer driven vote sites if you can get your friends to vote stuff up (cos you asked, not on merit).

Yes, yes, yes I know that is EXACTLY how it end up working on HN, Digg, Reddit and everywhere else. I'm just a crusty old purist.

As a purist of a different kind, I see it as more of a fundamental problem in the way HN/Digg/Reddit/etc. voting works, in that it's heavily sensitive to voting in the initial period, which is often pretty random, and then stuff that happens to get a few upvotes in that early period tends to snowball due to the feedback loop of being on the front page -> more people see it -> more people upvote.

Apart from its slower speed (too few stories per day), I actually see the Slashdot model, warts at all, as somewhat superior; or the Kuro5hin model as a vote-driven alternative. How's that for crusty purism. ;-)

I try to personally insulate myself from the fundamental brokenness by not using the HN front page, and only looking at the /newest page.

the problem is that the entire life of a story depends on a 5 minute window during which you absolutely must get an upvote.

A better window for a story's quality should be based on a # of views. i.e. at least one upvote in 100 views.

I don't have friends with HN accounts :(

Come hang out on #startups on Freenode!

Thanks, but I already spend too much time on HN as it is! :)