5

I've taken a shine to the Programmers Stack Exchange lately, but one thing I've noticed is it seems to have what I think anybody looking at the questions list would classify as excessive down voting of questions (not answers, the answers seem to get fairly normative treatment from what I've seen)

My questions are I guess: Can anybody say why? Can anybody propose a solution?

I feel like there are a lot of questions which are programming related and have distinct answers which are voted down because they may not be very high level, or sometimes confusion among the poster leads people to downvote instead of commenting/editing to guide the question to make more sense. Again I ask, does anyone have ideas about what can be done to help solve this? Feel free to comment if your assessment of programmers is that the level of downvoting is completely applicable based on the questions being asked, perhaps I am giving the questioners too much credit.

I would very much like to hear from the moderators there on their opinions as well as any frequent participants.

Edit:
Just a few quick ones to give examples of the things I'm seeing: What's the difference between overloading a method and overriding it in Java? I grant the english seems broken, but it is a language agnostic question about a software concept. https://softwareengineering.stackexchange.com/questions/164342/which-design-pattern-to-use-when-using-orm Again it has specifics but those are for the background, the question itself is about design patterns in relation to ORM tools which is a generalized design question.

That's just off hand right now, perhaps the downvotes are correct though. Again I am asking what we might do to get less downvoting, perhaps that means increasing question quality, to which I submit this request for ideas to that ends? Is there perhaps a way to put the FAQ more directly in front of people before they ask on there?

  • 7
    The answers found here may apply to this question too: Why are so many questions closed on Programmers Stack Exchange? – Rachel Sep 10 '12 at 14:05
  • 3
    Can you specify exactly what questions you think are programming related, meet the guidelines for subjective questions, and answerable, and have been downvoted? You don't need to link to all of them, but a few recent examples would be helpful. – Thomas Owens Sep 10 '12 at 14:16
  • of examples referred in recent question revision, "overloaded? overridden?" question is tagged Java - I wouldn't call it language agnostic – gnat Sep 10 '12 at 14:33
10

Part 1: statistics

(Skipping over the lies and damn lines)

Programmers has one of the highest all-time average score on questions of all sites. Out of 34 launched sites (I'm excluding metas), only Skeptics, RPG, SF&F and CSTheory have a higher average. If you count the all-time average number of votes (this query conflates questions and answers), Programmers comes 4th after Skeptics, CSTheory and SF&F. If you only count downvotes, Programmers comes 3rd after Skeptics and CSTheory. Programmers's all-time downvote/upvote ratio (again, conflating questions and answers) is 5.6%, rank 22 out of 34.

The recent trend is different, however. If you only count votes cast in 2012, then Programmers's voting amount is average, and it has the second-highest proportion of downvotes at 12.7%. Looking at monthly trends, the proportion of downvotes has been steadily increasing during the lifetime of the site. (Note that entries before 2010-09-01 are solely made of votes on questions migrated from SO, cast on SO before the migration.)

It's normal to see more recent downvotes, because the total number of downvotes includes those on highly-downvoted posts that eventually get deleted. For example, Wordpress (the site with the highest proportion of downvotes in 2012) has a recorded proportion of downvotes of over 30% in the last two months for which there are statistics, against 6% to 10% in previous months — I think deletion of old crap accounts for the disappearance of many downvotes.

So I retract my initial analysis that Programmers isn't an outlier: while it is not an extreme case, there has been a rather large proportion of downvotes recently. The increase is more marked than on other sites with a large proportion of downvotes.

I'm not sure how to interpret these statistics. SEDE data includes only the posts that weren't deleted as of 2012-06-27. Does this mean that, as of the last SEDE data dump, Programmers had been lax in deleting old crap (where crap is defined by having had downvotes)? Or have there really been more downvotes and fewer upvotes? It would be interesting to see similar statistics taking deleted posts into account.

If the large proportion of recent downvotes is not just an artifact of deletion dates, the next question is the interpretation. You force an interpretation in your question by speaking of “excessive downvoting”. Is there really excessive downvoting? Or is the downvoting a reflection on the quality of incoming questions?

The first step of finding a solution is finding what the problem is. Too many downvotes? Not enough upvotes? Too many bad posts requiring downvoting? Not enough deletion (and perhaps not enough closure) of downvoted questions?


Part 2: anecdotes

You cite two questions. Let me have a look at them.

  • https://softwareengineering.stackexchange.com/questions/164342/which-design-pattern-to-use-when-using-orm — score as I write: +1/-2

    Web and database applications are far from my expertise, but I have a hard time believing that this question is interesting and useful.

    The scenario is: “I am writing a small ASP.NET Web Forms application. (…) I want to use PetaPoco micro-ORM.” Uh, ok, so what will this application be doing? Is this something that has to be rock-solid by the time it's deployed, or something that's going to be continually evolving?

    The question is: “Do I still need several class library projects to separate the concerns?” That's a very broad design question. With such an unclear scenario, it's hard to give a generic answer to such a broad question.

    Furthermore, the question does not have any redeeming value of being important. If the initial decision on the breakdown in wrong, for a “small” project, refactoring isn't going to cost much.

  • What's the difference between overloading a method and overriding it in Java? — score as I write: +3/-0

    It would help your complaint against downvotes if all the examples were actually downvoted. Presumably there was a downvote on the initial question, which was written in barely-comprehensible English.

    Downvoting a post where the poster has made no effort at making himself understood is perfectly legitimate (“is unclear” in the downvote tooltip). We aren't just talking about grammatical mistakes here, there's a marked lack of effort. There's one single sentence here, followed by a few related words. It's not clear what is asked; I'm not convinced that the edit respects the original meaning of the question.

    Setting aside the presentation, the other thing that makes this question a candidate for downvoting is that it “does not show an research effort”. Here's how a good question might look like: “I read about overloading and overriding in this book/this blog post/…. I understand that this example [code] demonstrates overloading and this example [code] demonstrates overriding. But I don't understand why this example [code] is overloading. The author claims without explanation, but I think it is overriding because [explanation]. Why is that last example overloading?”

So my provisional conclusion is that there is no problem with good questions being downvoted. There are mediocre questions that are downvoted, which is as it should be. If you think that's the case, give examples of good questions that have downvotes.

  • 3
    Those numbers are very skewed from the site's early days. This query will show you a more accurate average post score for questions since 2011. The average question score has actually been dropping quite a bit. If you remove the piece of the query that limits questions to 2011 or later, you'll see the huge score averages from pre-2011 which skew the numbers from your query. Also, Data.SE does not include deleted questions, and I know a process auto-deletes low scoring questions after a few weeks or months. – Rachel Sep 10 '12 at 15:22
  • 2
    @Rachel Only questions without upvoted answers are automatically deleted, I think this is negligible. On the other hand, the historical variation is a good point, especially as Programmers has a lot of hugely-voted-on old crap migrated from SO which probably skews the averages (feel free to write a query with medians if your SQL skills are up to it). Let me see if I can get more relevant recent statistics. – Gilles 'SO- stop being evil' Sep 10 '12 at 15:38
  • Your revised query would probably be more accurate if you divided up/downvotes on questions by the # of questions, and up/downvotes on answers by the # of answers. Your current method of dividing all up/down votes by the # of questions or answers regardless of what type of post got the up/down vote will make the numbers skewed based on site traffic and voting patterns (for example, do answers get a lot of upvotes while questions do not? or vise versa?) – Rachel Sep 10 '12 at 17:18
  • @Rachel have a look at this query, this speaks more to the topic of growing question downvoting: data.stackexchange.com/programmers/query/edit/79473#graph – Jimmy Hoffa Sep 10 '12 at 17:26
  • The second one which is +3 was in the negative when I posted it as an example, it became negative mere minutes after it was posted as a question, but luckily someone took the time to edit it to fix the broken english instead of it being ignored then closed which is more common of negative voted questions, I'm guessing because I posted it here it got the attention it should have had to begin with. – Jimmy Hoffa Sep 10 '12 at 17:36
  • Programmers had been lax in deleting old crap This is somewhat true. – yannis Sep 10 '12 at 17:55
3

The downvote arrow tooltip says:

This question does not show any research effort; it is unclear or not useful

Latest questions with negative score:

No research effort, tick! Unclear, tick! Downvote!

Don't know what this one is about. Skip!

Someone please delete this. Downvote!

At first I thought the OP was quoting a reputable source, then I realized he's quoting himself, so there's no actual research effort. There might be, but I'm not going to read his blog to find out. Would have downvoted if this one was at 0 or -1, skipping it now.

No research effort, tick! Pointless question, tick! Downvote.

Extremely broad, unanswerable, uninteresting. Useful to anyone but the OP? No! Downvote!

What the hell is this I don't even... Downvote!

Meh. Skipped.

Which are the basic steps for learning Iphone development? Ah, Google must be broken for this guy! Downvote!

Android Holo UI might be a specific enough parameters for this one to not be yet another "pick for me, please" question. Meh. Skipped.

Meh. Skipped.

Flagged for deletion. OP posted an off topic question here by mistake, a moderator explained why it's off topic and had to be closed, OP seems to have understood why his question is off topic. No point in keeping this around anymore.

Profoundly confused question, OP doesn't seem to have even visited jQuery's website once. No research effort, tick! Unclear, tick! Downvote!

I don't even... Downvote!

Too long to read. Skipped.

Too long to read. Skipped.


I don't agree with you there's a problem. I'm very new around here, but I can tell you that one of the things that lured me in was the "promise" that this site is about "professional programmers":

Programmers — Stack Exchange is a site for professional programmers who are interested in getting expert answers on conceptual questions about software development. If you have a question about...

I've even written about this, on a different but related Meta discussion:

Programmers is not supposed to be a beginners site, and a lot of closed questions come from beginners. I don't think there is anything we can do here it is only natural that the site will have more closed questions than others. I like the advanced aspect of the site, and the work that is being done in the disciplined aspect of it, questions like this one should be the norm not the exception.

As a newer user, representing no one but myself, this is what I have to say:

  • I rarely upvote a negatively scored question. Not because I go with the flow, but because usually I find the negative score justified.
  • I like it that Programmers is trying to be a bit more disciplined than other sites.

I am not a professional programmer, bit young for that, but I have learned a lot since the first day I've joined the site. One of the more important lessons: A professional environment is tough, and it takes skills to survive in it. The scrutiny poorly researched questions face has made me a far better researcher, to the point that I now find 99% of the answers I'm looking for on Stack Overflow or Programmers just by searching.

Perhaps you want more of a "let's all hold hands and pat ourselves in the back" kind of site. I don't see anything wrong with that, but it wouldn't be a site for professional programmers who are interested in getting expert answers on conceptual questions about software development..

I'll close with Gilles' conclusion, with which I agree 100%:

So my provisional conclusion is that there is no problem with good questions being downvoted. There are mediocre questions that are downvoted, which is as it should be. If you think that's the case, give examples of good questions that have downvotes.

  • 2
    I wonder if this was downvoted from someone who passionately advocates posting helpful comments when downvoting. – Roc Martí Sep 10 '12 at 19:31
  • 1
    -1 The answer is unclear. Not really sure what point the example questions/responses are trying to make. – psr Sep 10 '12 at 21:10
  • 2
    @psr The example questions isn't the only thing in my answer though (and I was answering the question as it was before the latest edit). The point I'm trying to make is that the downvotes are justified (imho), given the site's scope and expectations, at least for this very limited set of questions (the negatively scored questions that were in the first page of the newest questions while I was writing the answer). Although the set is limited, is far larger than the OP's two questions, and I think it supports my opinion that there's no problem. Crap questions get downvoted, the system works! – Roc Martí Sep 10 '12 at 22:09
  • 1
    +1 reflecting on how you vote is the way to go: one of the most productive ways to learn at Programmers imNSho. And not going with the flow is the right attitude. 'You may find that highly voted posts have a certain appeal that kind of makes you compulsory follow the "majority vote". Resist that appeal, because making a habit of blindly following the score may damage your ability to evaluate content. When you see a highly voted post, don't just click the up or down arrow...' (active voter tips) – gnat Sep 11 '12 at 8:10
  • 2
    I passionately believe in posting helpful comments when downvoting, but I don't think it should be forced on anyone, and I do agree with nearly all of your votes here, and your conclusion. – Joris Timmermans Sep 11 '12 at 10:27
2

Programmers gets a lot of crap, down-votes and close votes are are how crap is dealt with. We try to avoid letting crap stay too long here, not all sites are like this.

Those of you that think down votes are too prevalent go out there and up vote the good stuff, quit trying to defend the indefensible.

1

I think I can identify some of the reasons why

  • The site scope has changed dramatically from what it originally was supposed to be. A lot of users (particularly on SO) will refer people here to ask their questions, not realizing that the question would no longer be on-topic for the site. This leads to a lot of questions getting asked that the community views as off-topic, or "bad".

  • There seems to be a lot of confusion over what is on/off topic on here by new users, and older users have gotten tired of explaining the same thing repeatedly. This leads to them just downvoting and moving on instead of repeating the same explanation they just gave on another question.

  • Some close votes give an automatic downvote, and we get many close votes

  • Question downvotes don't cost anything, so people are quite willing to vote questions down.

  • Downvoting seems to be encouraged on meta and in chat

    I watch meta and chat a lot, and I see a lot of users promoting downvoting, or bragging about their downvote record. Downvotes are often recommended when someone asks about what can be done about low quality questions, although I've also seen it recommended as a way to push questions off the front page too.

  • Some active users just see a few key words or phrases, (such as "Books" or "What's the best...") and automatically downvote and vote to close without actually reading the question. For example, see this meta question.

What can be done about it?

Some users don't actually view the number of downvotes as a problem. Others do.

Personally I think it's a problem, but I'm honestly not sure what one person alone can do about it. Most solutions I can think of would require either a community effort, or for SE to step in.

A few things I can think of:

  • To fix the number of downvotes we see, we really need to fix the quality of questions that get asked. The easy solution is to make new users understand the site scope better. The not-so-easy part is figuring out how. If you have any ideas, feel free to post them here: What can we do to help users understand our site better?

    I know I've suggested this many times and SE has denied it, but one thing I think would help would be to rename the site to match it's scope, or change the scope to match the site name

  • When promoting downvotes on meta, make sure to include instructions to leave a comment about why you're downvoting. Sure users don't have to follow it, but the guideline will be there for users that are only trying to "do the right thing".

  • Promote the SE app which lets you automatically insert generic comments. If you find yourself downvoting a lot and being tired of leaving the same comments, I'd suggest checking out this SE app which lets you quickly and easily add generic comments.

  • This only applies to a few users (you know who you are), but stop bragging about the # of downvotes you have. You shouldn't be proud of the fact you have contributed more negative votes than positive ones to your site. You guys are high rep users of the site, and people will look up to you and try to follow your example. Downvoting can be good, but I feel your attitude about it takes it to an unhealthy level.

  • Probably not a good solution, but SE could make question downvoting cost -1 again. I'm pretty sure SE removed the -1 downvote cost in the first place because they didn't want people to be hesitant in downvoting questions over 1 reputation, so I don't think this will happen and probably wouldn't be a good solution.

  • 3
    I am a new user, but I'm neither a drone nor an idiot. There might be some drones and idiots out there who "follow leaders" instead of using their brain, but I don't think it's a significant problem or we can blame the "leaders" for it. – Roc Martí Sep 10 '12 at 18:28
  • 4
    I'm looking at the people you blatantly call out (with direct links, and yet pretend to speak about "some people", passive aggressive much?), and I don't see them as either "leaders of zombies" or as doing anything wrong. The one guy has edited and fixed an impressive amount of questions, and is always around and helping noobs, the other one is also one of the most helpful commenters around and he called himself out on Meta, possibly because he already knew what he did wasn't so wrong. – Roc Martí Sep 10 '12 at 18:30
  • 3
    @gnat, who has also edited and improved an insane amount of posts, silently went on to fix this one as well, although he's also one of the guys you specifically point out to as "bragging about their downvote record". You are also a "high rep user of the site, and people will look up to you and try to follow your example", but looking at your profile I see that you are mainly interested in complaining and plugging your proposals, rather than actually helping people. Remember, all your comments and edits are public... – Roc Martí Sep 10 '12 at 18:33
  • 2
    -1 for a vitriolic, passive aggressive polemic, that isn't useful to anyone. – Roc Martí Sep 10 '12 at 18:34
  • 1
    @RocMartí I'm not entirely sure what you're saying, but I'm sorry if I wasn't clear in what I wrote and that lead you to misunderstand it. I wrote what my point of view was as a long-term P.SE user for the reason for the large number of downvotes on P.SE, and tried to offer some constructive suggestions for helping to reduce the number of downvotes. I'm really not sure where you're coming up with "drones" or "idiots" from, and I don't think I said anything against any particular user, only their actions. – Rachel Sep 10 '12 at 18:43
  • 3
    Nothing constructive about your suggestions, you start by plugging your long forgotten and largely irrelevant proposal, then move on to blantantly call out others (even if I agreed with you that their behavior is problematic, it still wouldn't be a significant part of the "problem"), then you move on to a counterproductive suggestion (auto comments? Right, that's what's going to help people, canned messages), then again you come back to people "bragging about downvotes" (Oh, sorry, your other point was about "promoting" not "bragging") and lastly you suggest something that is unfeasible. – Roc Martí Sep 10 '12 at 19:07
  • 2
    There's absolutely nothing constructive in there. Please stop, this is plain noise. – Roc Martí Sep 10 '12 at 19:08

You must log in to answer this question.

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