pull down to refresh

Are you going to share your thoughts about the aggressive downzapping?

193 sats \ 7 replies \ @k00b OP 10h

I don't like it and found it incredibly frustrating yet interesting. We don't have immediate plans to deviate from the money is the moderator experiment.

The problem is one of outliers. Mods and trust are the usual way forums deal with outliers, which we can always introduce. For now, we want to deal with problems as surgically as we can. Stackers can see when there are aggressive upzaps but aggressive downzaps hide their effects so this new sort is aimed at helping that.

I'd be happy to answer any specific questions if you have them.

reply

I'm supportive of continuing the experiment, and believe in the idea at a conceptual level.

The solution to trolls should be more tools to tailor your own experience, and sorting by downsats is a good step.

I suppose some decisions need to get made about what the default experience will look like, esp for new and not logged in users. But that's a lot of money for a troll to spend to influence the experience of non-commital users only. And every bad-faith downzap is actually a net resource contribution to other stackers, so while we may dislike the effect it has on rankings, we can't complain about big reward pools.

reply

Ideally the larger reward pool will elicit more zaps and offset the downzaps.

One of the problems with that is we know the targets of the downzaps, which makes the Keynesian Beauty Contest dynamic more problematic.

reply
reply
reply

The last X items didn’t get any downzaps. But they still show up.

reply

This is awesome, but also reminds me of something that (mildly) annoys me when looking at my (or anyone's) history. For posts, I can mouseover and see the breakdown of sats, but for comments, mouseover won't do it, and I have to click, go to the comment in the post, and then mouseover to see the breakdown. Is there a reason for the different behaviors?

reply

Do you expect the corrective mechanisms that already exist to be sufficient for this or do we need some new tools?

reply