Ratings for discussions

I was thinking, in particular, that if Discussions could be flagged as open/closed/whatever (status, see also https://www.loomio.org/d/qK1PbEGp/mark-proposals-as-passed-or-failed), then a rating-based sort, plus a status-based filter would basically allow you to ditch trello, and use Loomio for the vast majority of it's features. You could probably even implement a third-party Kanban-style UI on top of Loomio to make it even more convincing :)

I can imagine cases where this would make sense and cases where it wouldn't. Like, in this subgroup it would make sense, and in one that was for organizing a process it wouldn't - everything there needs to be looked at even if some isn't very juicy. But I'm sure people will use it if it's there, so you'd want to think about the downsides of having people use it in an ill-fitting case. Maybe it's harmless?

So then let's think about the cases where it makes sense. I'm not sure it's a good thing there. If the goal is to have unpopular feature ideas fall down and stop grabbing attention, then it would short-circuit proposals and decrease the chances of people considering new information and changing their minds. So it seems like it would undermine the goals of Loomio.

I like the idea. An indicator which would indicate what others think about the discussion would be usefull especially when you have tens of discussions and don't know where to start. An alternative could be a trellis chart showing activity in the discussions over time.

Quantifying and summarizing the subjective opinions to help people find their way around and show where the main discussions are and how the decision process evolves might be the next big space to be explored by Loomio.

If the goal is to have unpopular feature ideas fall down and stop grabbing attention, then it would short-circuit proposals and decrease the chances of people considering new information and changing their minds

I think that is somewhat of an exaggeration. The discussions would not disappear, they would just be pushed below more favoured discussions. In fact, the rating itself doesn't have an impact on that, just the sorting algorithm, and you can do that however you like - the rating just adds more information to the discussion, which the sorting algorithm can use or ignore.

One problem with simple up/down voting is that the meaning is not always clear. One person's upvote can mean something totally different to another person's. One way around that that I like the idea of, but have never seen properly implmented is a multi-variate orthogonal voting system. For instance, you could have voting axes like:

• urgent/can wait
• interesting/trivial
• agree/disagree

Which lets people vote something as urgent and interesting, but still allows them to disagree (at least, if they're honest :). But I think that that's probably overkill for Loomio, which is why I proposed just up/down.

I'm really interested in this stuff, but I share some of @presley 's concerns. Also, I think Reddit (and similar systems) can be improved by ensuring more 'upwelling' of content which may slip through the cracks due to a poor start.

If a ratings system were ever developed for Loomio, I would certainly hope for it to be personally optional as a sorting method. Interested in any other opinions here.

Totally agree with you @gregorycassel, optionally applying ranking systems on a public discussions index seems like a great way to allow the good parts of many sorting algorithms.

I think a starting point is just a public discussions index ordered by some activity metric. Number of comments in past 24 hours or something. It would not be perfect but would immediately expose lots of stuff and provide discussion for other rank tuning options.