This started out as a discussion over in salesforce.meta.stackexchange (SFSE for short from here on), where I'm the most active.
The general idea
Downvoting, and its prominence on questions & answers, is a tool of the utmost import.
No matter how much you may think that being downvoted shouldn't make you feel bad or be taken personally, there are people who will feel bad or take it personally.
I feel that this is even more likely to be the case for new users.
Instead of discouraging people from downvoting new users (though I think I'll always exercise greater restraint with new users), what if there were a way to turn a downvote into an opportunity for more engagement?
Things like downvoting, voting to close, and voting to delete low-quality questions/answers are very fire & forget (or set it & forget it, if you have a fondness for a particular USA infomercial). Once you take that action, unless you bookmark the question or keep it open in a tab, chances are pretty slim that you'll ever come back to it.
How can we improve the outcomes of downvotes (with an emphasis on new users)?
The goals here are:
- help ease new users into what's still an intimidating landscape
- increase the net score for new users
- overall increase the number of new users that become long-time contributors.
Possible implementations / Prior Research
- One of the most interesting ideas I came across in seeing whether or not there was already something to tackle this issue was the mentorship experiment. Based on what I've found in meta from the DAG team, this idea seems to have fallen by the wayside. I'm not sure if SFSE (or other SE sites) would have enough active, knowledgeable people to make this work well
- Push a notification to downvoters of a question when the author makes an edit
- Allow users to opt-in to receive notifications after certain actions (leaving a comment, voting to close, voting to delete, some combination thereof) when an edit, comment, or maybe both are made. Notifications could be triggered by author edits/comments.
From the research I did prior to posting this question...
This appears to have been asked before in 2009 with a very positive receiption. It was declined back then, but as some of the more recent comments/answers have suggested, perhaps it's time to reconsider in the wake of https://stackoverflow.blog/2018/04/26/stack-overflow-isnt-very-welcoming-its-time-for-that-to-change/
Automatically removing downvotes is not a popular solution.
Tweaking the reopen system was previously suggested but, again, the proposition of automatically undoing downvotes is not popular.
Based on Letting receivers of downvotes ask why they got it, automatic, non opt-in notification is problematic. Maybe the problematic bit here is that this would have been based on what the OP wants to have done (with the implication that everyone would question downvotes) vs handing more control to the downvoters.
Mandating comments to accompany downvotes is a non-starter because it makes an already (comparatively) rare action harder to perform.
(link to some study I remember about how people feel bad about negative messages being sent, and how making the negative option more neutral led to more use of the negative option)
Possible outcomes and side effects
More opportunity for engagement
If people stop engaging on a post, the game is over.
Keeping people engaged can lead to more people becoming long-term contributors, which is important for the long-term health of a stack exchange network site.
Discover, and retain, more people who could become valuable contributors
There are people and posts that fall through the cracks. This could be a good way to address issues of "this post has potential, but needs some tweaking".
I don't know if there would be many other good options to address these posts and try to improve quality and encourage new users. Other options may be more work to implement, or be excessively obtrusive to the StackOverflow/StackExchange flow
Not only more engagement, but more timely engagement
The longer a question sits, the higher the chance that the response you've already gotten is the response you'll end up with.
All 3 of the suggested solutions would be able to address the issues closer to the time they arise (or to the time they could possibly be resolved).
This doesn't necessarily give an advantage to hit & run posting
The idea here is that this is only a shift of some of the responsibility of posting from the OP to the responder.
#1 and #2 require action from the OP (and my imagining of #3 would involve that as well), at which point things can be taken into the hands of an experienced contributor who better knows what the community is looking for (and if more work is needed still)
No action from OP = no notification to downvoters.
Downvoters could be overwhelmed with notifications, or punished for not taking action
An opt-in solution here, like #1 or #3, would mitigate this completely.
The mitigating factor for a solution like #2 would be (downvote frequency) * (edit frequency). There's probably a few queries that could be run to work out some hard numbers for how likely a downvoter would be to receive a notification, but my gut feeling here is that this probability is not likely to be high.
There are a lot of questions/answers that are completely unsalvageable, which would cause the SNR to tank
Yes, there are a lot of extremely low quality posts. Such posts wouldn't take much additional effort to ignore, certainly less effort than keeping a bookmark/tab and refreshing until you give up. The important thing to me here is that this would increase discoverability, and give people who are worth it another shot.
This could be gamed
Well, almost everything can be gamed... but what if we limited this to "new users" &/or to the Nth question/answer?
"New" users probably stand a lesser chance of knowing the system well enough to game it. Hell, it's common for me to need to tell some new users to edit their post instead of adding a comment multiple times before they actually get it.
Experienced users are very likely to know how to @mention in a comment.
What about upvotes on low-quality answers?
I'd argue that's a separate problem, and not one that this request is meant to address.
The point here is that downvoting discourages new users, more active users is a good thing, and how can we make downvoting less of a death kiss without making it harder to downvote?