How to save the world... One question title at a time

Question

If you got here because you googled "How to save the world," I apologize. This post really has nothing to do with saving the world. I picked the title specifically to illustrate my point.

This question...

How to call function with variables of a form

...really has nothing whatsoever to do with passing variables of a form to a function in Microsoft Access. Instead, it is a mundane question about a routine syntax error.

I use Google for assisting me with my software development all the time, and this is seriously starting to impact my Google searches. The title was exactly what I was looking for, but the question itself had nothing to do with my problem. Stack Overflow results appear high in the list of results in Google, so the titles need to be accurate.

How do we solve this problem?

Answer 1

Stack Overflow is so big now, this is a major issue. Perhaps the site should start voluntarily withdrawing questions of questionable value from the Google index.

For example:

• Ask the visitors. When a user comes in from Google, ask them something like

  We see you found this question while searching for how to flobber the gargle. Was it helpful and relevant to your query? Yes / No

  If a question receives too many "unhelpful" votes, either remove it from the Google index through robots.txt, or add it to an "unhelpful" review queue specialized on editing titles and such, or both.

• Alternatively, use the existing anonymous feedback data (if it's any good) and exclude questions from the index that get overwhelmingly bad feedback.

• When a user comes from Google, measure the time they spend on the page. If they leave it relatively quickly with no action at all, interpret that as an "unhelpful" vote.*

Also, a bad generic question title usually goes hand in hand with a bad / too localized question. To address that,

• bring back "too localized" and streamline closure of those questions (but that's a different discussion); and

• exclude questions with one or more close votes (except of course duplicate votes) from the Google index. That would go a long way towards cleaning the search results from trivial garbage, too.

---

_{* Google are in a much better position to do this accurately, though, as they can see whether the user returns to the results and searches on, suggesting that their problem wasn't solved.... Not sure whether doing this on Stack Overflow's end is ever going to be effective.}

Answer 2

Some of this problem is probably due to authors not having a clear grasp on what their questions are about. How can they, when they write their titles before the questions even exist?

When a title is written before anything else, it captures the essence of what the author assumes the post will be about. But it's common for writers to reorganize things as they go. When writing questions, in particular, they may make some realizations through the miracle of rubber ducking and change course. The final version of a question may end up being very different than the original concept, and it's easy to forget to update an already-written title after finalizing a post.

I've already proposed one possible solution: encouraging askers to write titles after they finish up the question bodies. We can do this in practice by moving the title text box below the body text area. I know this seems odd on the surface, because when you read, you start with the title, and then the move on to the "real content." But why should writing take place in that order?

Several of my old English teachers and professors, drawing on years of experience, recommended writing the titles (and theses) of essays last, or at a minimum reviewing them after all other edits were done. Perhaps more convincingly, this is standard practice in the newspaper industry. Reporters don't submit headlines with their articles (or if they do, they're just suggestions). Editors write the headlines towards the end of the layout process. Doing so allows them to review changes that are made to stories throughout the process and ensure that headlines accurately describe the versions of articles that actually go to print.

This problem can also be caused by authors not having good "command of language," or not understanding the purpose of titles, but those are harder problems to solve, and at the moment, I have no brilliant ideas.

Answer 3

Step 1. Filter the questions having more than 100 (1K... 5K...) views.

Step 2. For filtered questions, measure relevance of the title to text.

Step 3. Pick 1% (5%, 10%...) questions that are worst relevance wise.

Step 4. Put troublesome questions into a new review queue.

Step 5. Let community sort it out.

Answer 4

Badgees? We don't need no stinken badgees!

While sometimes badges as rewards work well to entice people to flag/edit/vote etc, with title edits I think these rewards will attract too many bad badge whores.

It's easy to change something minor, so they'd change any old thing to gain their badge.
Reviewers can deny the change, but as it needs reviewing this causes more work for reviewers and adds to review queues.

While it will also attract badge whores who actually suggest worthwhile title edits, by their very nature, badge whores seek out potential opportunities. So even the decent ones will likely suggest edits which while are arguably decent and so accepted, they're also still unnecessary and time could be spent elsewhere for them and the reviewers.

Tiny pointless edits from badge whores introduces more work for reviewers, unnecessary edit history, more data to store and show so more server strain, things for people to view and potentially rollback, etc etc.

It also provides badge whores and robo reviewers to feed from each other, with badge whores posting pointless title changes and robo reviewers allowing them.
Who benefits?

Yes, it's currently easy to post tiny edits in the question body for rep, and with robo reviewers there too, so the point is don't introduce another badge(s) to add to the current issues.

In summary, I think all you'd mostly do here is provide more people with more badges...

Interesting ideas

The goal is to get the questioners to make decent titles, and avoid others having to edit, flag, review, etc.

1
Add a new vote function (up/down) for the title. So users vote on question as per normal, and additionally and separately the title.

Rep gained is 1 reps per positive vote, -1 per negative vote and just added/removed from their usual site rep. (or, -5 rep lost for negative.. hmm..real incentive..)

I know this is a bit fiddly, and will make the currently simple and easy voting system a bit more complex and faffy.
I also know this is a lot of work for devs, and a big change to the site for people to get used to, but I left the idea here as it is a good one that will very likely return better titles, from the want for rep/avoid rep loss.

It makes questioners aware that it was their title that got them the downvotes and so they edit and improve it, and as this becomes common knowledge people write better titles from the start.

2
Change the ask a new question form title place holder to something more, persuasive.

"What's your programming question? Be specific" is not really pushing people to write a decent title. In fact the first part is suggesting they can be vague about it as they think "I'm about to write that further down so this bit doesn't matter..".

Something like:

Be specific and summarise your question here (Bad titles give bad rep)

As people posting bad titles are often just after a quick answer:

Good titles return decent answers, and quickly. Bad titles get downvotes and slow or no answers

Or:

-This is very important- if you want decent and quick answers. Summarise your question carefully

etc

3
Freebie SO hats for good titles.

4
Ability to edit the title separately from the (current) question edit.

Clicking "Question edit" remains the same, and still has the title in there so one can do all in one go if necessary.
If the question is ok but title needs some love, one can click the newly implemented link "edit title" and get a single text field to quickly edit and done with.

Min char limit 3 perhaps, as titles having a small amount of text are likely to need a small edit.

No badges or rep gained, so while rubbish edits will be suggested (nature of the beast(s)), they will be minimal (or at least lacking rep/badge whoring).

This just makes it quicker and easier to edit titles, which might push more people to do it as the HUGE page full of stuff on the current edit is daunting, and fiddly with min 6 chars in body, XX chars in "reason for change" do this, don't do that, etc.

5
Change a lot of the system so good people doing good things can continue, and bad people doing bad things lose (relative) privileges indefinitely.

IE
Someone found robo reviewing, no more reviewing, ever, kthx.
Someone with a poor accept to deny flag ratio, no more flagging.
etc
This would allow introducing rep gained and/or badges to various things, such as title edits.
Then, again, poor accept to deny ratio for title edits, no rep or badge for you! Lose the badge if you gained it, do not pass go, etc.

---

While change requires people to learn something new, it also freshens the site and shows people new ideas are being implemented in the site for improvements.

I know a lot of these things wont be implemented here, but they should be. People get away with lots of shit around here, and continue to do it.
This drops morale in those doing good as they feel demotivated as while there's incentive to do good, there's no loss for people taking the proverbial.
To give incentive for people doing good, rep and badges is not enough, being told thanks you've done good, but this person didn't so we've punished them, is what drives a lot of people.

Well, that and money, if we got cash instead of rep...

Answer 5

We need another editing badge just for titles.

You can never have enough editing badges, there's always more to be done.