My article on an idea for a new comment rating mechanism based on Anatol Rapoports rules of constructive criticism for #WeCo, a #platformcoop replacement for Reddit.
"Each reply to a comment could be upvoted as in Reddit, but in addition, they could be given a rating based on
* How fairly the reply portrays the original commentators stance
* If there was any acknowledgement of learning something
* Whether there was an attempt to seek a common ground."
@LeoSammallahti read the posts. all very interesting.
There is some deep complexity to actual achieving what is intended but any improvement on online discourse as it is today is a good thing.
just a couple thoughts which I'm sure are already being explored;
Doesn't seem like one form do rating will do - why not a mix of them?
What are the ways the ratings will be gamed?
How will the rating system address dissenting and marginalized voices that may go against prevailing group thought?
We've thought about having communities set their own rating systems.
The gaming question is good, we could compare ratings that the person replied to gives and the ratings that other users give, to see if there's a gap. This could indicate that either OP or the users in general are not fair in their ratings.
Hopefully it would foster an atmosphere where people would be careful not to mischaracterize anyone, which is something that disproportionally affects marginalized viewpoints.
What would happen is that the huge number of trolls would downrate marginalized people, making them even more marginalized.
This could be prevented if the rating would be only given by the person replied to, in which case both participants would be entirely equal regardless of other users opinions.
So legitimate criticism would always be downrated then, right?
If OP wouldn't see the criticism as portraying OP's stance fairly, if there was nothing that the person replying to learned or aknowledged to learn from the OP and if OP doesn't see attempt to seek common ground, it would be downrated by the OP.
I'm not sure I understand what you are proposing or how it would work... is this a rating system which applies to initial posts which are not comments, or would only comments in reply to another's post be rated?
One idea would be that there would be a specific community titled something like "Constructive Discussion", where people could rate replies to their comments according to the 3 principles laid out it in my original toot.
We have not set details yet in stone :), so we might allow all users to rate all replies according to the principles. Or allow everyone rate replies, but give extra weighting to the rating by the person replied to.
Hope my explanation is not too messy :D
So how would this work to alleviate cases where someone was being brigaded by hundreds or thousands of people, such as wilw recently for instance?
@hhardy01 @rbenjamin
1. I comment in support of nuclear energy
2. You make a reply to my comment in opposition to nuclear energy.
3. I rate your comment on basis of how fairly it portrayed my stance, whether you acknowledged learning anything from me, and whether you attempted to seek common ground.
If only I can give the rating to your comment since you replied to my comment, there's no possibility to brigade.
@LeoSammallahti @hhardy01 how close is this to being deployed. Seems like it really needs some real world implementation to learn from.
One thing still missing might be how does this overall reward better behavior on a community wide scale and not just individual discourse.
Also how it's paired with other types of moderation will be critical.
When you say "how does the community reward better behavior" that amounts to saying "how does the community punish minority positions and dissent."
@hhardy01 @LeoSammallahti no not all. the internet at as a whole and most every social media out there are set up to reward engagement regardless of its form so it can monetize it.
The leeds to the most divisive and reactionary content/engagement crowding out all other forms.
The better behavior I'm referring to is simply that which doesn't intentionally/unintentionally hurt someone or just not intentionally/unintentionally being assholes to each other.
@hhardy01 @LeoSammallahti even when it's not (currently) been monetized as in the case of Mastodon this kind of content still tends amplified the most.
Maybe part human nature but also part a missing reward system for better discourse.
I by no means to do support legitimizing or softening hateful shit.