I'm not going to beat a dead horse at this point - I can see that the management has no intention of removing the voting limitations altogether. However, I'd like to see them consider a compromise: Limit the "on-topic" votes if necessary, but increase the limitation for off-topic votes by 2-3x.

This might sound strange, but hear me out. I think we're all treating on/off-topic voting as part of the same process when they're really quite different.

Here are the facts that we know so far:

  1. Ever since the new voting rules were introduced, there are practically zero dissenting opinions on high-voted (on or off-topic) questions.

  2. On most proposals, there are more on-topic examples than off-topic examples. This is reflected in both the votes and the questions themselves.

  3. Most of the highest-voted off-topic questions are trivial examples that don't really help define the boundaries at all. For example, in the GIS proposal there's one recruitment ad and another request for pirated software. Well, duh. In the Astronomy proposal there are two questions about Astrology. Again - duh.

I explain these phenomena as follows:

  1. People are reluctant to give up their precious off-topic votes, saving them instead for another highly-voted off-topic question in order to help get the proposal off the ground. (Note: NaGE isn't the same thing - it's more likely to be interpreted to mean that the question is on-topic but just isn't one of the best examples.) Strategic voting is mutually exclusive with honest voting.

  2. It's hard to come up with good off-topic examples that aren't so far off-topic as to be meaningless. The thought process for coming up with an on-topic question is, simply, "What is something I'd like to learn or would enjoy answering?" For off-topic, it's "What is something that somebody else is likely to ask that would be rejected by the community?" The latter is a far more complicated process, you have to be creative and try to put yourself in several other people's shoes. Combine that with a disincentive to vote off-topic, and it tends to not be worth the effort.

  3. Because it's so unlikely to get votes for a borderline question, people who want to create off-topic examples are likely to go for the jugular. The off-topic questions most likely to be voted up are the ones likely to be interpreted as trolling and/or produce a knee-jerk "get it off my site!" reaction. This is useful, but doesn't really encourage deep thought about where the true borders are.

The more I think about it, the less it seems to make sense that on-topic and off-topic are treated equivalently. A proposed site will be defined by the best on-topic questions; the off-topic questions aren't going to appear on the site and likely aren't going to appear in the FAQ. They're mainly a way to prove that the proposal is supported on its own merits and not just attracting a lot of attention because it's all-inclusive.

Looking at the big picture, any well-focused Q&A site ought to be defined by a very small list of questions that do belong and a very large list of questions that don't belong. What we're actually seeing tends to be biased toward on-topic questions, especially when you ignore the off-topic questions that are self-evident on account of being illegal or completely idiotic.

I understand the need for some better filtering in determining what a proposed site is. But we're not doing a good enough job of determining what a proposed site isn't. There should be more encouragement to think of, and vote on, good boundary-testing off-topic questions.

Try this little thought experiment: Go to the highest-ranked proposals, look at the highest-voted (off-topic) questions, ignore the ones that are just silly, and try to come up with a list of subjects that are off-topic. If you can come up with more than one or two for any given proposal, I'd be very surprised. In my observation, the new voting system is working "OK" for on-topic questions, but rather poorly for off-topic questions.

Can there be any harm in loosening up the restrictions on off-topic votes? We still have the NaGE vote, so I don't think that it would polarize the questions/voting or result in a lot of pointless troll questions; users are still limited to 5 questions and stupid questions would theoretically just get "downvoted."

What do you think?

  • 1
    I'm for removing limitations at all... but I agree finding good off-topic questions is a lot more difficult.
    – Massimo
    Jun 9, 2010 at 16:06
  • The purpose of the first phase is more to get the most exemplary on and off-topic questions out there and have people show interest in the proposal. The next phase is more suited for your goal, which I support completely
    – Ivo Flipse
    Jun 13, 2010 at 16:08
  • @Ivo: What phase, the commitment phase (which isn't even really a phase)? Or the beta phase (at which point the site's already been defined)? Neither really seem appropriate for this. Honestly, no offense, but I get the impression that you only read the title and not the content of this post. What I've been saying is that (a) the adjective "exemplary" does not make as much sense for off-topic the same way it does for on-topic, and (b) even if it did, the system isn't producing a lot of useful off-topic examples (even if Joel says it is).
    – Aarobot
    Jun 13, 2010 at 18:27
  • If you look at my Web App proposal, you would see that I tried to do exactly what you are referring to: propose a lot of borderline questions and get the discussion going. However, the team decided to limit this to 5 exemplary examples. You're point is that exemplary off-topic questions should be borderline/grey area question. So what's stopping you or anyone else from suggesting these and vote on them?
    – Ivo Flipse
    Jun 14, 2010 at 7:00
  • If you have 100 followers you can have 500 suggestions for off-topic questions and 500 votes to distribute over them. So you may only vote 5 times, but in the end we have more than enough votes to include more off-topic questions, they just won't have your vote on them. So what? However, I think we need more than just some comment spamming to settle discussions on those borderline questions, which I hope will be done in the next phase. @Aarobot
    – Ivo Flipse
    Jun 14, 2010 at 7:02
  • @Ivo: Regarding your first response, I'm not criticizing the people creating the off-topic examples, I'm pointing out that the system is flawed; everything I wrote in this question was dedicated precisely to explaining what's hampering the collective effort to create and vote on really good off-topic examples. And yes, theoretically having 100 followers makes plenty enough votes available, but in practice, this isn't resulting in the best off-topic examples floating to the top. Surprisingly, it is working with on-topic examples, but the off-topic examples were better before the restrictions.
    – Aarobot
    Jun 14, 2010 at 13:51

1 Answer 1


I completely agree that the current system isn't doing a good job of driving good, edge-testing, "off-topic" questions, which I got into in some detail here.

But here's the (slightly) shorter version:

The combination of limiting both the questions one can submit and the number of "off-topic" votes that others can cast, while leaving the "meh" votes ~unlimited disincentivises posting good, boundary-defining "off topic" questions.

If you happen to nail it (and it's a lot harder - you're aiming for that contentious realm), you might get one of someone's precious "off-topic" votes. But you're a lot more likely to get "Meh" votes (partly because most people still can't tell the difference).

How to Fix it:

I think what the votes mean should be clarified:

  • Great Question - clearly on-topic and exemplifies everything the site is about.
  • Tough Call - this replaces "off-topic". It represents the questions that would likely get asked but are near the line. They may be just subjective enough to be closed, or just helpful enough to allow it. But they are not:
  • Terrible Suggestion - this is the "new Meh", and should be used for questions that are ridiculous, contentious, exact dupes, etc.

The fundamental change is that borderline questions would be encouraged, even though you can't be sure if they're just before or just past "the line".

  • I agree with the analysis but disagree that changing the wording of the definitions will alter the asking/voting behaviour in any significant way. Also, "great question" and "tough call" sound like "definitely on topic" and "maybe on topic", which further skews the vote toward on-topic; that's exactly the obstacle I'm hoping to overcome.
    – Aarobot
    Jun 9, 2010 at 21:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .