This discussion has been had before and people didn't like the idea of it being public. What I am pointing out was there too: that pseudo-privacy is not great either especially when anybody can be a moderator or host an instance (but not to the point where I will expect others will know my dissenting reply is not the source of the 1 downvote).

It would be more understandable if this were semi-anonymized data that could be cross-referenced, but there's already other issues (nebulous or subjective hidden thresholds, moderators moderating their own posts). Unless it is incredibly cautious (detecting mainly scripting or obsession-level voting) I expect it'll be a repeating annoyance especially if the approach is widely used.

[–] jet@hackertalks.com 1 points 1 day ago* (last edited 1 day ago)

I think its a core design issue; if votes are free and unlimited then they will be gamified, one method to provide a feedback mechanism here is to make votes accountable.

There are other moderation styles, like the slashdot random moderation points, which show some promise.

I've seen other gamified models where downvotes are scarce, or cost time, or cost money, so they are reserved for bad behavior.

The design of showing votes on posts makes it a signal, and that signal will be gamified. I think for smaller communities having outside third parties chill participation is bad for the overall growth of lemmy. In the long term I imagine a community level system selected by the moderators would fit best.

examples

Only people who have posted or commented can downvote / upvote
Only people who are subscribed can upvote/downvote
only people who have repetitional guarantees can upvote/downvote
posts only visible to subscribers

Just trying to get local representation into communities.

etc. etc