I am sure this has been debated here before but I keep dreaming of it anyway. Let’s say everyone’s upvotes and downvotes were public and you could independently score posts using this data with your own algorithm. If the algorithms to score posts were also public then you could use another users scoring algorithm instead of writing your own (think lesswrong power-user).
As a simple example, lets say my algorithm is to average the score of user_Rational and user_Insightful and user_Rational algorithm is just lesswrong regular score minus User_troll’s votes.
The benefits would be a better curated garden, more users, and more discussion.
Currently, the backlog to changing the codebase here is so big and there’s so little work going on it that even if there was a consensus for this change it would be unlikely to happen.
More specific to this proposal, there are at least two problems with this idea: First: it could easily lead to further group think: Suppose a bunch of Greens zero out all voting by certain people who have identified as Blues and a bunch of Blues do the same. Then each group will see a false consensus for their view based on the votes. Second, making votes public by default could easily influence how people vote if they are intimidated by repercussions for downvoting high-status users or popular arguments, or even just not downvoting because it could make enemies.
Yeah, I suspect this would just move the game one step more meta. Instead of attacking enemies by mass downvoting now people would attack their enemies by public campaigns based on alleged patterns in the targets’ votes. Then we could argue endlessly about what patterns are okay or not okay.
I agree there still would be very easy ways to punish enemies or even more common ‘friends’ that don’t toe the line.
I do think it would identify some interesting cliques or color teams. The way I envision using it would be more topic category based. For instance, for topic X I average this group of peoples opinions but a different group on topic Y.
On the positive side, if you have a minority position on some topic that now would be downvoted heavily you could still get good feedback from your own minority clique.
Public voting and public scoring
I am sure this has been debated here before but I keep dreaming of it anyway. Let’s say everyone’s upvotes and downvotes were public and you could independently score posts using this data with your own algorithm. If the algorithms to score posts were also public then you could use another users scoring algorithm instead of writing your own (think lesswrong power-user).
As a simple example, lets say my algorithm is to average the score of user_Rational and user_Insightful and user_Rational algorithm is just lesswrong regular score minus User_troll’s votes.
The benefits would be a better curated garden, more users, and more discussion.
Currently, the backlog to changing the codebase here is so big and there’s so little work going on it that even if there was a consensus for this change it would be unlikely to happen.
More specific to this proposal, there are at least two problems with this idea: First: it could easily lead to further group think: Suppose a bunch of Greens zero out all voting by certain people who have identified as Blues and a bunch of Blues do the same. Then each group will see a false consensus for their view based on the votes. Second, making votes public by default could easily influence how people vote if they are intimidated by repercussions for downvoting high-status users or popular arguments, or even just not downvoting because it could make enemies.
Yeah, I suspect this would just move the game one step more meta. Instead of attacking enemies by mass downvoting now people would attack their enemies by public campaigns based on alleged patterns in the targets’ votes. Then we could argue endlessly about what patterns are okay or not okay.
I agree there still would be very easy ways to punish enemies or even more common ‘friends’ that don’t toe the line.
I do think it would identify some interesting cliques or color teams. The way I envision using it would be more topic category based. For instance, for topic X I average this group of peoples opinions but a different group on topic Y.
On the positive side, if you have a minority position on some topic that now would be downvoted heavily you could still get good feedback from your own minority clique.