High threshold for "well-asked" questions

Question

Various posters have previously complained about the voting patterns on SuperUser and the standard response is that down votes are accorded to "poorly asked questions". Well, it would seem that "well-asked" questions are rarer than unicorn farts, because I literally have to search to find any questions on Super User that are upvoted. For example, this is current list of active questions:

 0  Make Ubuntu computers phone home for remote SSH
-2  How can get micro torrent working?
-2  How can i send an analog signal trough an hdmi port
 0  The print queue loads very slowly
 0  WMP 12 - .wmv video/audio freezes when pausing/unpausing. 
 0  Notepad++: is there any add-on or regex or macro to split long text
-1  The input line is too long
 0  Split a large file into small pieces
 0  How to diagnose computer fan refusing to spin?
 0  Command line to disable trackpad on Macbook Pro 15 w/ Touchbar
-1  I can't install hyper-v in win10,error 0x80070057,here is my cbs log file
 0  How to join multiple files in Notpad++
-1  High Latency of DirectX Graphics Kernel (dxgkrnl.sys)
 0  Windows 10 Installation - Missing required files and Blue Screens

Okay, so 95% of the questions asked "poorly asked" questions. Hello, this is web site for computer users, many of whom might be new to computers (duh) or new to a technology, who are asking questions. Requiring somebody to write PhD nobel-prize winning research paper, foot noted, peer reviewed, with help from their professor and character references before they get an upvote seems kind of like a steep threshhold to me.

I would add that NONE of the downvoters here have posted comments explaining why they are down voting the question, which is a violation of SE guidelines. The general practice seems to either ignore questions or downvote them, then not explain why the question is being downvoted. The guidance for downvotes is that they should be used when the question is "egregiously sloppy", which none of the above listed questions are. So, clearly we have a lot of downvotes occurring that are not only unexplained, but are against our own guidelines on the help page.

I counted the upvoted to downvoted questions for the most recent 150 questions and these were the results:

down    25   17%
up      38   25%
zero    87   58%
----------------
total  150  

Maybe we should self-examine a little and re-consider the policies for downvoting. When 95%+ of posts are not receiving upvotes, that is of concern, at least to me, because it creates a hostile climate for new users and sends the message that questions are unwelcome. To those who say, only "poorly asked" questions are unwelcome, I think when the downvote/ignore-to-upvote ratio is approximately 1-to-1, that does not really hold water. When only 5% of questions are receiving upvotes, that is a problem and indicates a policy-level problem.

In most classroom/learning environments (and yes SE is supposed be a learning environment) a common motto is "There are no stupid questions." But on SU the motto seems to be the reverse: "Every question is a stupid question."

I suggest a discussing policy changes to change that.

Answer 1

Here's a snapshot of the semi-public site analytics page - you need to be a trusted user for full access. I've taken a snapshot from 21st april 2015 to 21st april 2017. It kind of shows something interesting. On a bad week, we have under 2k downvotes and the worst I've seen is 5k upvotes. Clearly, your statistics are off. (You can probably get the same data for other periods from clever SEDE queries but I'll leave that for someone else. ) On average though, we probably have a better proportion. Once again, I'll leave it to others to crunch the numbers, but the graph makes it pretty clear.

So, the whole premise of this question - that voting is overwhelmingly negative is incorrect. I'd prefer to comment but as voting is anonymous (even mods can't see them outside general trends) this is optional rather than compulsory.

Most SE sites are a balancing act. Good questions attract experts - and sometimes users. While we try to be welcoming - if we have a load of questions that don't quite cut the mustard, we lose the folks who actual generate value.

And while broad policy changes sound very grand, I don't see any detailed, actionable suggestions that balance out those two things. Sometimes we choose to answer or salvage. Sometimes its the downvote. Both approaches have their advantages, but I don't see what policy changes will help.

Answer 2

I ran some numbers on questions specifically.

This query (based on some of the work I did on a very similar question) counts the questions with negative, zero, and positive scores asked by users with less than a certain rep in the most recent period of a certain length. In the last 30 days, users with 100 rep or less posted 286 negatively-scored questions, 1350 zero-scored ones, and 493 positively-scored ones. Note that these results are skewed due to automatic deletion. Setting the maximum question age to 9 days (the wait on the quickest roomba task) shows 122 negative, 426 zero, and 92 positive. Upvoted questions from new users are not the most common, but they're not rare either.

This other query is very similar, but counts individual votes rather than questions' scores. On less-than-9-day-old questions by users of 100 rep or less, there were 158 upvotes and 232 downvotes cast. Downvoting is more common, but again, upvotes can be found as well.

A bit of data handling caution is advised when lowering the rep cutoff. If you set it to 1 (to get only new users), you're almost guaranteed to get purely downvotes, since users who received upvotes no longer have only 1 rep.