Is there a specific set threshold for the number of times various users have to ask "How do I vote to migrate questions to a site that doesn't appear in the list?" before even, say, an "Other" button that links to the "other reason" form is added to that list?

It's obviously confusing, please fix it already.

Why does vote to close have static "belongs on" list?

More options when flagging for migration

Off topic questions and redirect to affiliated sites

Should dba.stackexchange.com be an option when closing "off-topic" questions

Add a "more" link to the Close as Off Topic window so we can migrate to *all* Stack Exchange sites.*

Off-topic migration site targetting

Migrate shortlist based on tags?

Context sensitive "it doesn't belong here" options?

What is the reasoning behind restricting the number of migration paths?

And on, and on. If you have an actual specific, finite number of complaints about the same exact thing that you need to see before you are convinced that it's an issue, please let me know, and I will search harder to collect them all here until I reach that threshold and you fix it.

  • Oh, wait, are you talking about the flagging screen for <3kers? – Undo Nov 5 '13 at 0:25
  • When you state that there are numerous times users have asked, it's usually helpful to provide links to at least a few of those previous posts so we know what you're talking about and don't have to guess. It's also usually helpful to indicate which site is being referenced. – Ken White Nov 5 '13 at 0:26
  • There's a small fraction of the list of questions and confusion related to an inadequate "migrate to" target list. I'm talking about SO, and about every other site that displays a "migrate to" target list in the migrate vote dialog (vote to close -> off topic -> migrate). There is only one place users can vote to migrate a question, so I presumed that didn't need clarification. Sorry. – Jason C Nov 5 '13 at 0:34
  • Glad you find it helpful, Jason. – jscs Nov 5 '13 at 0:47
  • Related: meta.stackexchange.com/questions/203318/… – Shog9 Nov 5 '13 at 1:02
  • @Shog9 Thanks, I'll add that to my list. – Jason C Nov 5 '13 at 1:03
  • Are Meta, SU, Tex, DBA, and SharePoint really the most common migration paths from SO? I would have guessed Meta, SU, CodeReview, Programmers, and maybe DBA. – Jason C Nov 5 '13 at 1:04
  • They were, @Jason. As of late, Code Review has replaced SharePoint (which has dropped significantly in the rankings). – Shog9 Nov 5 '13 at 1:06
  • They should really be tag-based... I can't imagine questions tagged as java ending up in Tex. In any case, a fixed list of any type here is confusing and half-useful, as evidenced by the volume of posts on meta about it (and a minimal amount of thought about how humans behave), and is a limitation that is purely a product of UI inadequacy rather than a limitation of underlying functionality (a question could be migrated anywhere, a UI that prevents a user from expressing reasonable intentions is a UX fail almost by definition). It's really time to fix that... – Jason C Nov 5 '13 at 1:10
  • 2
    @JasonC "expressing reasonable intentions" that's just it; they aren't reasonable. Most people are very bad at suggesting migrations. Most people understand the scope of the site that they're on enough to know when a question doesn't belong, but the don't understand the scope of every other site well enough to know where it does belong. Because of this migrations as a whole on the site have had a long history of problems in which people are constantly migrating questions to sites that don't want such questions. – Servy Nov 5 '13 at 1:14
  • @JasonC Migrations are allowed to the small set of sites that have shown to have a strong history of frequent and successful migrations. All others are routed through moderators who can be trusted to only migrate questions as is appropriate. – Servy Nov 5 '13 at 1:15
  • @Servy That's not a case against limiting migrations, that's a case for not having a blind, inadequate, game-y reputation system that's also used to control site privileges. If people who can't be trusted to migrate questions are given power to migrate questions, that's a different failure, with limited migration paths as a shoddy band-aid (among many others). – Jason C Nov 5 '13 at 1:27
  • 1
    @JasonC But the current site behavior is very effective, that's the point. Your proposed change in allowing migrations to any site would be a colossal failure, to which there is quite a lot of history from actual experimental evidence, not merely speculation. That leaves the only other option as removing a feature that has a proven successful track record just because you don't like it. That makes no sense at all. You can claim that the current behavior is shoddy, but it is demonstrably successful. Your assertions are not supported by evidence, merely your personal opinion. – Servy Nov 5 '13 at 1:29
  • The posts and discussions on meta contradict the claim that these things have a successful track record, which incidentally is the whole point of my OP. You don't get to believe something has a "proven track record" just because you don't want to acknowledge consistent complaints (I mean a general "you"). Look at all the posts about, e.g. poor edits, people pointing out problems with gaming the review system, etc., etc. If you actually look through meta, you see that the current behavior is only marginally effective, otherwise those problems wouldn't exist. – Jason C Nov 5 '13 at 1:34
  • @JasonC You have posts of people asking for the feature to be changed. That's not saying that the feature is unsuccessful. I'm sure almost all of those people would strongly oppose your request to remove migration entirely. They want to migrate more, not less. They also largely haven't looked through past discussions (which is why they're creating duplicates) and are not aware of all of the problem such migrations have caused in the past. As for evidence of it's success, one can look at, for starters, the migration stats, to get an idea of the percentage of useful migrations. – Servy Nov 5 '13 at 1:39

It's not the number of complaints, it's the arguments that it is a problem.

Clarifying the UI is easy enough to do, especially if we just do something like a add a textbox to the close dialog that converts input into a flag. However, migration is a complex issue and opening it up more isn't necessarily a win. So, we (as in folks like yourself who raise this as an issue and us the employees) are at a bit of a philosophical stalemate here. We think the disadvantages of making it easier to flag/vote for migration to any site in the network outweigh the benefits.

Where's the harm in closing an off-topic question as off-topic and leaving a comment pointing the author to a (potentially) better site when a migration option isn't directly available? (Or even when it is.)

| improve this answer | |
  • I wanted to flag a question for migration to CodeReview, which was extremely appropriate. Instead I was presented with Meta, SU, Tex, DBA, and SharePoint? Those options don't even make any sense. How do you feel that is a "win"? Rather than closing the question as off-topic, I bypassed the close vote system (probably not your intention) and directly suggested to the poster that he repost on CodeReview, which he did, and now there's a cross post. Fail to see the "win". – Jason C Nov 5 '13 at 0:50
  • @JasonC We don't offer migration paths to beta sites at all. (I personally think that we should reevaluate that particular stance since some of our beta sites are very mature, but on the other hand even what looks like an established site could turn out to have major problems. Either way, that's an entirely different discussion.) The options presented right now are the most popular migration paths (that aren't to beta sites). – Adam Lear Nov 5 '13 at 0:52
  • @JasonC Why did you choose to suggest posting on another site while not voting to close on Stack Overflow? Migrations aside, why keep a question open on SO that doesn't belong? – Adam Lear Nov 5 '13 at 0:52
  • Well I'll be sure to suggest more questions be migrated to Tex in the future, and when I do, I'll just skip the rather useless voting path and tell it to the poster directly. Perhaps SX users should simply be encouraged to ignore the migrate vote "feature" and make direct suggestions instead (<- facetious), clearly a UX win. – Jason C Nov 5 '13 at 0:54
  • @JasonC I'm afraid I no longer follow your argument. Not being able to vote to migrate to Code Review is why you will "boycott" the existing migration paths? Okay. – Adam Lear Nov 5 '13 at 0:55
  • It's not fair for you to ask me why I didn't take extra steps to work around a poor UI. If you don't want that to happen, make the UI useful. Anyways I ultimately flagged it as "other". – Jason C Nov 5 '13 at 0:55
  • Obviously that's not why I would "boycott" migration paths, that is a gross over simplification. It's highly obvious that the available migration paths are confusing and less-than-useful, based on the sheer number of questions about it here and the fact that this conversation is even happening. Sh*t or get off the pot, I don't support half-useful features and I prefer a method that is consistently useful. It either needs improved to be complete, or it needs to go away and e.g. be replaced by simply always using "other". – Jason C Nov 5 '13 at 0:56
  • @JasonC Anna is talking about the arguments that have been made here. – Andrew Barber Nov 5 '13 at 1:07
  • @JasonC The arguments so far haven't convinced us. Pure repetition isn't likely to change that. Do you have new ones or proof backing the existing ones beyond stating that the problem is "highly obvious"? – Adam Lear Nov 5 '13 at 1:07
  • Even on a surface level: When my only options for migration targets for a question tagged java are e.g. Tex and SharePoint, and I have to bypass the system to choose something sane, how is it not obvious that there is a UX issue? – Jason C Nov 5 '13 at 1:18
  • @JasonC I fail to see where there is much "confusion" at all. People understand that they can't migrate to other sites, they just don't agree with it. That's not confusing. Just because a lot of people complain about it doesn't mean it'd be a good feature to implement. What about all of the people complaining about the existing bad migrations; what about all of the people who would start complaining about the bad migrations caused if this feature were implemented; what about all of the past evidence that this feature would be abused more than it would be used properly? – Servy Nov 5 '13 at 1:18
  • 2
    @JasonC You're right - it is a UI limitation. It is deliberate because we don't want the average user to be able to migrate a question everywhere. We limit choices on purpose and it's already hard enough for folks to decide how to close a question between several OT options, other close reasons, and the existing migration targets. Varying the shortlist based on tags was proposed before as you note in your question. Our reasons for declining that are outlined there. I believe they trump UI convenience. – Adam Lear Nov 5 '13 at 1:21
  • @JasonC Your proposal would harm site usability in the name of consistency. Again, the issue is not that it's confusing; it's rather clear to most users what the functionality is, they just don't all agree with it. Those handful of sites with both a lot of valid migrations and a proven history of successful migrations are allowed, and this is helpful. The moderators would add a lot of needless work to take on those additional proposed migrations. Just as this is a clear issue in your mind, to everyone who's looked at the stats over time the current functionality is clearly the best. – Servy Nov 5 '13 at 1:24

To back up what Anna wrote: I went through 86 migration flags yesterday on Stack Overflow, and migrated 41 questions in response to them. That's already less than a 50% success rate, but a day later 5 of the questions I did migrate have already been rejected...

But that's me, fumbling around on a weekend. Let's look at some rough statistics for migration flags handled by all moderators in the past 90 days:

Migration flags: 1971
Migrations in response to flags: 359 (18%)
Migrations rejected: 19
Overall migration flag success rate: 17%

Is this really something that's worth building a UI to encourage?

Don't get me wrong - there are certainly things we could do to improve the UI and streamline the experience for folks who know what they're doing. But right now, migration just doesn't make sense in the majority of cases. You've never even created an account on Code Review, but you're certain you know what sorts of questions they want? I tend to doubt it. Folks get this Mother Teresa complex and try to salvage each and every off-topic question, and it just ends up wasting time for little or no benefit. It's perfectly understandable - but, let's not encourage more of it.

| improve this answer | |
  • 1
    You don't need to create an account on Code Review to know what goes there, and that is an unfair criticism; I have never had a desire to post code for a review or review other people's code. Anyways, thanks for gathering these stats. I don't really understand what you are getting at; this seems to be a case for removing migration ability completely based on high work and low success rates. – Jason C Nov 5 '13 at 1:39
  • 4
    Not at all, @Jason: note that the predefined migration paths all have a very high success rate, as they require 4 voters to agree on a destination in order to actually migrate... And when even that fails, we kill the migration path. There's a huge difference between a small, curated list of good options and a massive, unmoderated one. – Shog9 Nov 5 '13 at 1:42
  • Then what is the problem with allowing voters to select any of the network sites and still requiring 4 voters to agree...? It's the same theory as duplicate question flagging. You have every single question on the site to choose from, a huge unmoderated list, yet you have the flexibility to pick ones that aren't in some preset list and the fact that previously selected questions are displayed encourages subsequent voters to choose the same question. It's not a crazy free-for-all, it's an unmoderated list that works really well. – Jason C Nov 5 '13 at 1:51
  • 1
    Because with > 100 sites to choose from, it's really not hard to find 4 people who also have no idea what SiteX is actually about. Now, combining this idea with one that let you migrate to any site you were reasonably active on might have legs - but that's a whole 'nother level of complexity and... Well, you still wouldn't be able to migrate to Code Review. – Shog9 Nov 5 '13 at 1:53
  • You say "it's really not hard to find 4 people who have no idea what SiteX is actually about" as if that would be a new problem introduced with the idea I just described, yet you have a low 17% success rate already. I'm not sure how that's an argument against anything, the current system has the same issue. With 100k's of questions to choose from its also not hard to find 4 that misflag a duplicate, but it still works well. I don't really know what else to say. I'll start posting arbitrary blocks of code on CR now to prepare. – Jason C Nov 5 '13 at 1:56
  • And, of course, this all still sidesteps the off-topic but fundamental issue that all of these systems are a workaround for giving users who don't have enough knowledge to migrate the power to migrate. – Jason C Nov 5 '13 at 2:00
  • @JasonC The problem is that there are lots of people that think they know what a site is about, but just don't. Programmers and Server Fault are probably the two worst examples of this. Lots of people think they know what belongs on those sites; enough to have quite a lot of people trying to migrate questions to them, but to which the majority of those users (on SO) are wrong. When each of those sites were options it resulted in tons of crap being migrated to those sites, and even now there are lots of flags for migration to those sites that get declined. – Servy Nov 5 '13 at 2:41

You must log in to answer this question.

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