What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 130 Stack Exchange communities.

There have been previous discussions on newbies being able to upvote on their own questions, and although the request was denied, the arguments for and against can be summed up as

  • They should be allowed to upvote as it encourages people to answer questions from new users and the new users should be able to reward all the folks who post helpful answers (as they can accept only one); users don't always return to upvote old questions once they gain the privilege.
  • They should not be allowed to upvote as it could lead to voting cabals and they've not earned the site's trust to be given the right to vote.

There are merits to both sides of the argument, and hence I propose a solution below that meets the two arguments in the middle.

SHADOW UPVOTES

  • New users can upvote helpful answers on their questions (and only their questions). This encourages users to reward answers when it is still fresh in their memory. People seldom revisit an old question to upvote helpful answers if they hadn't done so/couldn't do so then.

  • The upvote doesn't trigger a rep increase for the answerer until the said user gains the privilege to vote across the site (15 rep), which can only come from upvotes to their question from other users. This adjustment can take place during a system wide rep-recalc or user triggered recalc.

  • Answerer/other users do not see the upvote, whereas the user can. i.e., the vote count displayed for the new user and the community is off by 1. This is just to deter answerers from upvoting newbies solely to bump them up to 15 rep after knowing that they have a pending upvote, without looking into the merits of the question. Of course, the new user can always mention it in a comment, but that can't be avoided.

  • Once the user gains 15 rep, the shadow upvotes become permanent and are visible to everyone (again, only the counter will be visible. It still will not be obvious that the upvote came from the user. Perhaps the rep page can read +10: Shadow -> Permanent, without mentioning the question it came from)

Some of the benefits of such a system are:

  • New users (question askers) get to participate & upvote early on, bringing them a little closer to feeling like being part of a community than the current system. Also, they get their voting badges sooner.

  • All answerers (not just the user with the accepted answer) who have invested their time in answering and have been helpful to the question asker (i.e., asker has upvoted) get rewarded, if the new user stays on to be active in the community and crosses the threshold at some point (this in turn is encouraged by giving them limited privileges)

  • This is also harder to game using sockpuppets (as was the concern in the previous discussions). For e.g., if a user creates sockpuppets A & B, and asks questions and answers with puppets and gives upvotes all around, all they're giving out is shadow upvotes which have no effect and no rep increase until the community decides that the user can be trusted (i.e., upvotes their question(s) 3 times). Gaming the system once the user has gained the reputation is no different from gaming the current system, and there are other checks in place.

One possible drawback is that it might encourage users to upvote a question they have answered in the hope that there's a hanging upvote. But, I don't know if its that big a concern.

I'd appreciate a discussion and comments on this suggestion.

share|improve this question

4 Answers 4

I'm curious if this is worth doing. Using this query you can see the following data

countid  newuser  expuser  Type                    Ratio of New to Experienced
-------  -------  -------  ---------------------   ---------------------
160,179  20,082   140,097  Zero Accepted Answer    14.3%
728,619  25,287   703,332  NonZero Accepted Answer  3.6%

Ratio of zero to non-zero answers 20%

You can see that

  • New users represent 5% percentage of all accepted answers. 45,369/888,798
  • New users represent a larger portion 14.3% of the zero accepted answers than of the non-zero
  • If this system could improve the non-zero answer of new user from 14.3% to 3.6% it would only affect < 2% of accepted answers.

Seems like a lot of effort for little in return (for the second listed benefit anyway).

share|improve this answer
    
I had not considered looking at data dumps. Perhaps that's a decision for the SO team to make as to whether to implement or not. I was looking for a discussion on the merits of the suggestion, as it could possibly evolve into something that has a higher return on effort. –  Lorem Ipsum Apr 26 '11 at 23:38
    
voter != answerer. –  Uphill Luge Apr 27 '11 at 1:53

Another pretty large drawback is that it's confusing to the new users in that they are giving out upvotes but not having a visible impact on those users' reputations. I enjoy reloading a page and seeing a rep bump when I upvote/accept, especially when I help them hit a privilege milestone. Not having that would be confusing and might discourage users from sticking around because it would appear that their actions don't have any tangible impact.

You could mitigate this somewhat with notifications to the effect that these are shadow upvotes, but that too could confuse and drive off new users as well.

share|improve this answer
1  
Yes, but as you say, a notification: "This is a temporary upvote and will be made permanent when you get 15 rep. You can always mark this as the accepted answer by clicking the check mark" can help. The added bonus is that new users are made aware of accepting answers. That is something unique to SO and new users are not aware of it, but upvoting is very common among forums/social media and hence everyone tries to upvote first. I've seen several Q's where the user posts a comment saying "I tried to upvote, but sorry :(" and leaves and doesn't come back to learn that they can accept answers. –  Lorem Ipsum Apr 26 '11 at 21:18
1  
If someone is still so new that they can't upvote yet, I doubt they're going to notice that someone's rep didn't increase by 10. Or if they do, they'll assume that there's a delay before it takes effect. –  Cam Jackson Nov 14 '11 at 23:34

What I took out from those discussions on why it isn't necessary is that it shouldn't happen because accepting is much more valuable both rep-wise and as feedback overall, and I agree with that. New users should earn the right to give votes around. If they intend to stay, then earning that right should be relatively easy the more they participate.

I would much rather they comment on the other answers that are helpful stating that. (a good indicator that they do intend to stay) That sort of feedback is much more valuable to me than anything else. That tells me that the asker declared this useful information, something that an anonymous upvote can't do. Having unresolved questions never feels good (whether that comes from the asker, answerer or anyone else interested). That would also mean that answerers shouldn't expect additional upvotes to come from the new user either. Let those upvotes come from those who have already earned it.

Side-note:
What would happen if a 21 rep new user falls back below the voting threshold then? No simple way to handle that (that isn't confusing to everyone).

share|improve this answer
    
"What would happen if a 21 rep new user falls back below the voting threshold" - what happens now, anyway? I'd say just be consistent with that (though it is an interesting edge case to consider) –  Daniel DiPaolo Apr 26 '11 at 21:50
    
Well, it's not really so much about the answerer expecting as it is about the new users being introduced to the system in a systematic, yet limited way. See my comment on Daniel's post on why I think this might also be a good way to increase answer acceptance. –  Lorem Ipsum Apr 26 '11 at 23:08
    
@Daniel: I think certain rights should remain once earned... like the right to vote. Upvoting rep doesn't really mean the community "trusts" the person; it is just an anti-newuser-upvote-clique measure. –  Lorem Ipsum Apr 26 '11 at 23:11

There's another face to this, and that is that the experienced users have a responsibility to upvote the interactions between and by newbies. Too often I see a reasonable question, reasonable answer -- accepted, and both question and answer have zero upvotes. Maybe the prevention of new users upvoting is why this happens; if so, the only cure for it is for current users to upvote these things.

Maybe we should keep track of "first upvotes" and give a badge for it. Nah.

share|improve this answer

You must log in to answer this question.

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