[abetusk]

People are focusing on the Reddit mod narrative but if you go down in the article further, there's some interesting ideas.

The author talks about Dunbar's number [0] and does a speculative approximation, saying that good faith communities are about the maximum size of the "friends-of-friends" and some of "friends-of-friends-of-friend" graph. That is, say you trust your friends to have good faith discussion (around 150 of them) and you trust your friends of friends to have good faith discussions (150^2) and maybe some portion of your friends of friends of friends (now we're at 150^2.5), then the community size is about 300k.

The data size is small and anecdotal but the idea is pretty interesting, at least to me. It at least tries to answer what the critical online community size is where good faith discussions break down, "scaling issues" start to creep in and at what point online communities change. A lot of folks here on HN talk about things like how to social media at scale (for example, Shirky's "Three Things to Accept [about online communities]" [1] has been no doubt featured on HN before). I also wonder if this ties into the "1000 true fans" idea somehow.

Whatever the community size is where scaling issues crop up, this leads to practical advice on which communities to focus on if you want to engage in a meaningful way, promote your work, or find a niche community for a business idea, say. That is, look for communities whose size is large enough to be impactful but small enough so that it hasn't gone through the scaling phase transition. This number is most likely in the range of 20k to 300k, maybe with an upper bound of 3M.

[0] https://en.wikipedia.org/wiki/Dunbar%27s_number

[1] https://web.archive.org/web/20131130191257/http://www.shirky...

[dblack12705]

Thank you! The Reddit mod thing was really just supposed to be a catalyst for some of the ideas I had about group size limits, not really the point of the article. Though I WAS a little annoyed by the mods lol.

Edit: rereading your comment this is a very good summary of my article and maybe I should have put something like this up front to stave off the inevitable focus on Reddit mods…

[johnnyanmac]

>look for communities whose size is large enough to be impactful but small enough so that it hasn't gone through the scaling phase transition. This number is most likely in the range of 20k to 300k, maybe with an upper bound of 3M.

The issue is like Earth in the solar system; way too small of a "goldilocks" zone and in the days of the modern internet it seems almost impossible to hit it.

Reddit is undoubtedly too big, but look for any other competitor and the discourse can barely even hit the 10k mark, where you're browsing maybe once a week hoping someone made a post or left a comment in a post. And any modern aternative that does hit that mark (discord, tiktok, Vine back in the day) will quickly balloon past that 300k range.

Your only solace seems to be seeking older communities that have stood the test of time. Even 4chan these days seems to have cooled down outside of the larger, more infamous groups.