polis VS VisualizingTwitchCommunities

Most of the work of good public consultation facilitators is (1) stopping those loud/frustrated/opinionated/entrenched people from distorting process, and (2) allowing some people to be heard.

Perhaps surprisingly with (2), public consultations are sometimes almost like public therapy -- allowing people to feel heard so they can then take a step forward, even if it's not a step fully in the direction they prefer. People can make better compromises when they feel heard, they just need to be contained so they don't take speaking time/attention from everyone else.

Disclaimer: highly involved in public consultation, tech, and consensus-building technology through my interest in https://github.com/compdemocracy/polis

In regards to you reference about too many nodes: https://github.com/KiranGershenfeld/VisualizingTwitchCommunities I think that's a great example. I tried looking at the code before but found it very messy to understand, but then again I feel like that is always the case when python projects become bigger. Maybe I just need to spend an hour to understand the structure or something, so I can actually understand where exactly what is done.

This is all open source. You can download my data and put it Gephi yourself to play with your own map along with the scripts I wrote. Instructions are all in the Github README. Find it here.

*according to this https://github.com/KiranGershenfeld/VisualizingTwitchCommunities/blob/CloudCompute/Visualization/ActiveNodes.csv from here https://www.reddit.com/r/LivestreamFail/comments/m3r9o6/twitch_atlas_vol_2/

You can see a list of each community and what I have identified them as here. You can see some of the raw numbers here as well as every streamer in the image (there are 1484 of them).

If you want to see the actual code its available on Github here.

Its this file btw: https://github.com/KiranGershenfeld/VisualizingTwitchCommunities/blob/main/Visualization/GephiData/Dec12Edges.csv

As always the code is here. The graph was generated directly from "Users In Chat" data taken from the top 100 streams every hour for the month of February. I calculated the overlap of these viewers between each streamer and graphed what I found. Read more about the process here.