15
\$\begingroup\$

On Code Review, I've noticed that there is a blind upvoting of posts. People judge the question by the heading, not by the effort.

Is this a proper way of upvoting posts?

\$\endgroup\$
6
  • 3
    \$\begingroup\$ By "heading," do you mean the title? If so, how can you be sure that people are only voting based on the title? Please provide some examples of such posts you've observed. \$\endgroup\$
    – Jamal
    Jun 27, 2014 at 17:20
  • \$\begingroup\$ Strongly related: meta.codereview.stackexchange.com/questions/75/… \$\endgroup\$
    – Simon Forsberg
    Jun 27, 2014 at 17:32
  • \$\begingroup\$ I also upvotes on the posts by its tittle, even i dont know the abc of the code written in it. \$\endgroup\$
    – Govind Singh Nagarkoti
    Jun 27, 2014 at 17:34
  • 3
    \$\begingroup\$ I also use the title as a first impression. If it's good (description of the code), chances are the rest of the post is good, and I'll upvote it. If it's a bad title ("How to improve this code"), and there's no explanation of the code that can help with improving the title, then I'll downvote the post. \$\endgroup\$
    – Jamal
    Jun 27, 2014 at 17:39
  • 7
    \$\begingroup\$ FYI I upvoted this post based only on the title, you know ;) \$\endgroup\$
    – Marc-Andre
    Jun 27, 2014 at 18:06
  • 1
    \$\begingroup\$ One thing that none of these answers address in the slightest is the problem of bad questions getting upvotes. Yes user retention is important, yes allowing askers to upvote answers to their question is important, but people ARE blindly upvoting questions to the point of upvoting questions that are off-topic for Code Review and should be deleted! \$\endgroup\$
    – nhgrif
    Jul 6, 2014 at 1:15

6 Answers 6

Reset to default
23
\$\begingroup\$

As one of this site's top voters and with one of my former display names being @lol.upvote, I feel somewhat obligated to post an answer here.

This site has been in public beta for well over 3 years (today is beta day #1254). The only reason for this, boils down to people not voting enough. Feel free to browse the meta posts on this site for more details.

has been identified as the newest most pressing concern, because when you don't get rewarded for putting effort in a thorough review, it's kind of pointless. But that's mostly about answer votes.

Question votes are different, in the sense that a lot of questions asked, are posted by drive-by users that are asking their very first question on the site, and might come back with another question in a couple of weeks or months, depending on their CR experience.

By giving a new user/asker 3 upvotes, we unlock their "upvote" privilege and allow them to upvote answers they're getting.

Any answerable on-topic question with less than 3 votes is therefore under-voted.

With over 20K visitors every day, how do you explain that we hardly ever have more than 40-some voters any given week? (that list excludes users that have less than 11 total votes spent).

Out of all askers that do get the 3 upvotes to be able to spend some votes, a given fraction will upvote some of the answers on their question - and that's where it ends.

What boggles me, is how reviewers can answer a question that they didn't even upvote.

If you can answer a question, that question is worthy of your upvote. Or at the very least, of a comment explaining what the question is missing to get your upvote.

Is it "ok" to upvote a question solely based on its title? Of course it's not!

But a question with a good catchy title will draw views and attract reviewers; a couple of answers and a bunch of votes within a certain given amount of time is all it takes for the question to become a "hot network question", and from there, anything can happen.

So should you upvote a question just because it has a good title? No. You may view a question for that reason, but then you'll upvote it...

  • If it's clear what the code is doing
  • (and/or) If the question is well formatted
  • (and/or) If you read the code and think "hmm this, I'd do like that"

Notice "and/or" - I can upvote a question even if I don't know jacksquat about . How? If no one has cast a "non-working code" close vote and the question is well-written and well-formatted and reviewers interested in PHP should get their eyes on that one, I'll upvote it.

Since the middle of May 2014, there's been more comments than question votes on the site. Recent activity (last 60 weeks) doesn't seem to agree with your assertion that "people vote blindly":

site's activity graph, last 60 days (as of 2014-06-27)

If anything, people are voting sparingly. And that's hurting us. 1402 answer votes and 1060 question votes in a week with 20K visits/day means hardly 1.76% of visitors spend ONE vote, on average.

VOTE. Then VOTE AGAIN. And then VOTE SOME MORE.

\$\endgroup\$
1
  • 8
    \$\begingroup\$ The only time when you should not try to vote some more is when you just voted on a post. Voting again will remove your previous vote if you try to do it too quickly! \$\endgroup\$
    – Simon Forsberg
    Jun 27, 2014 at 18:48
10
\$\begingroup\$

enter image description here

Like the button says, if you feel the question has shown some research effort, and it is useful, and clear, then upvote.

The point is that you should not vote based on what other people have already voted.... if the post has 100 upvotes, and you think it is useful, and clear, then add your vote too.

If you think the question is 'worth' about 5 votes, and it already has 10, then you should not upvote it, well, that's the wrong way to vote.

\$\endgroup\$
7
\$\begingroup\$

What is research effort?

On Code Review, it's a bit hard to determine what research effort is.

Are you expecting users to have researched different kinds of code smells before they post a question? I don't. If people knew about all kinds of code smells then the need to post the code here in the first place is significantly reduced. The "shows research effort" item makes more sense on Stack Overflow than here.

Are you expecting users to explain what their code does when they post a question? Good, so do I.

Personally, I sometimes give a question an up-vote just because it's on topic, especially when the poster is a new user, as a way to say "Welcome! Thanks for understanding exactly what Code Review is!".

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Basic things like running your code through linters (hlint for Haskell) could easily fall under "research effort" for CR. I follow the CSS tag myself and see far too many questions that have markup that doesn't validate. Or check out this question where the author clearly didn't bother to research what it means to be responsive because there's no media queries involved. \$\endgroup\$
    – cimmanon
    Jun 27, 2014 at 20:39
  • \$\begingroup\$ @cimmanon that's a pretty interesting point you bring up here, ..do non-validating CSS questions gather close votes? I think they would belong with non-working code questions... \$\endgroup\$
    – Mathieu Guindon
    Jun 27, 2014 at 22:53
  • \$\begingroup\$ @Mat'sMug Just because it doesn't live up to the standard doesn't always mean that it's not working. Some might be closed for it, but today's browsers and applications are often too good at fixing the brokenness for it to be considered as truly broken. \$\endgroup\$
    – Simon Forsberg
    Jun 27, 2014 at 23:17
  • 1
    \$\begingroup\$ @cimmanon I think the most common reasons for why people have not run the code through linters is because they're either too lazy or just simply not aware of them. It's up to reviewers to educate them in this aspect. \$\endgroup\$
    – Simon Forsberg
    Jun 27, 2014 at 23:18
  • 1
    \$\begingroup\$ @Mat'sMug I'm more forgiving about CSS validation issues since invalid CSS tends to either not work (and its obvious that it doesn't work) or they're using browser specific styles (eg. filter for gradients in IE8). Invalid HTML is another thing entirely and can cause unexpected problems when the browser tries to compensate for them, which in turn causes the author to overcompensate (eg. paragraph within a paragraph will render as paragraph following a paragraph, which may cause the author to reduce the margins on paragraphs). \$\endgroup\$
    – cimmanon
    Jun 28, 2014 at 0:24
7
\$\begingroup\$

It's hard to say that a question is receiving upvotes due to it's title alone. Good titles catch people's attention and draw in large numbers of views. More views generally equate to more upvotes.

Putting that aside though, it takes effort to come up with a good title and I find that generally, authors of questions with good titles have given the same effort and attention to their question.

Here are two examples of questions with excellent titles:

These are both well thought out and well posed questions beyond their titles.

(Disclaimer: The latter one is mine and I have an answer on the former, so I might be a bit biased.)

\$\endgroup\$
6
  • 3
    \$\begingroup\$ Congrats BTW, "Everyone Loves Fibonacci" was featured at the top of the weekly newsletter last week! ;) \$\endgroup\$
    – Mathieu Guindon
    Jun 27, 2014 at 17:55
  • \$\begingroup\$ @Mat'sMug Could you give a link to it ? \$\endgroup\$
    – Marc-Andre
    Jun 27, 2014 at 17:55
  • \$\begingroup\$ @Marc-Andre Can't link to an email... have you not subscribed to the site's weekly newsletter? (go to your profile / preferences) :) \$\endgroup\$
    – Mathieu Guindon
    Jun 27, 2014 at 17:56
  • \$\begingroup\$ Well I don't know why I was thinking about the podcast :S ! And no I don't like newsletter :P \$\endgroup\$
    – Marc-Andre
    Jun 27, 2014 at 17:57
  • \$\begingroup\$ Really? I wonder if the newsletter got caught in my spambox. Thanks @Mat'sMug. I didn't know. \$\endgroup\$
    – RubberDuck
    Jun 27, 2014 at 18:01
  • 1
    \$\begingroup\$ "Look 'ma, I can read code!" got ~2k views in 24 hours. Good catchy titles are very important for a post's visibility, there's no bias in this answer :) \$\endgroup\$
    – Mathieu Guindon
    Jul 2, 2014 at 13:50
7
\$\begingroup\$

Voting in Stack Exchange

It is worth mentioning that voting is at the discretion of the voter. Sure there is preferable ways to vote like @rolfl and @Simon are explaning, but we can't force you to vote in some specific way.

Here is a good general resource for voting : When should I vote?

What is your definition of effort ? Could you qualify a good title and a good description like effort ? Are you talking about the effort that it takes to write the code ? Voting on a question require a lot of variables to take into account.

\$\endgroup\$
5
\$\begingroup\$

my voting stats are:

Voting Stats

I had more down votes than most people for a while, I don't just hand out votes, and I agree with what @Mat'sMug (@RetailCoder, @lol.upvote, ...) said

What boggles me, is how reviewers can answer a question that they didn't even upvote.

If you can answer a question, that question is worthy of your upvote. Or at the very least, of a comment explaining what the question is missing to get your upvote.

it bugs me to see questions that have 3+ answers and no up votes, I will occasionally blind vote on those questions, because they are getting a lot of attention or they should get a lot of attention.

A lot of my down votes I have removed because the questions or answers are cleaned up, or the questions/answers were deleted.

\$\endgroup\$

You must log in to answer this question.

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