I feel like reddit ends up playing whack-a-mole with toxic subreddits because it's not like the problem users disappear - they just move onto another subreddit and slowly turn it into some flavor of the place that got banned/quarantined. This might be an unsolvable problem long-term without something like real IDs tied to user accounts, which brings its own slew of problems.
> I feel like reddit ends up playing whack-a-mole with toxic subreddits because it's not like the problem users disappear - they just move onto another subreddit and slowly turn it into some flavor of the place that got banned/quarantined.
On the contrary, at least one study found that what you describe does not happen. Rather, a quantitative reduction in hate speech was observed when Reddit banned a number of toxic subreddits in 2015.
Inhabitants either moved off the platform entirely (accounts that frequented those subs ceased to be active) or those that stayed appeared to modulate their behaviour to conform to the norms of less-toxic subs.
Among social sciences a single study has very little weight. Given how often it is quoted despite this signals a flaw in popular cultures relationship with the field.
> Among social sciences a single study has very little weight.
Sure. "I feel like ..." has absolutely no evidentiary weight, though, so in this case, a little weight is strictly preferable to nothing.
> Given how often it is quoted despite this signals a flaw in popular cultures relationship with the field.
I'd say rather it signals a blind spot in the social sciences, where researchers are failing to investigate emerging online phenomenon wrt communities and moderation in significant numbers.
More studies confirming or disproving the results of this one would, of course, be preferable. But we should hardly apologize for turning to what little study and evidence there is rather than pulling "this will cause X to happen" assertions directly from our asses.
>I'd say rather it signals a blind spot in the social sciences,
Apologies, it appears I wasn't clear. I meant the extent that psychology and sociology studies in any part of their fields are quoted when there is only a single study, not just in relationship to internet/social media.
>But we should hardly apologize for turning to what little study and evidence there is rather than pulling "this will cause X to happen" assertions directly from our asses.
The difference is in the latter case we are well aware of the origin, while in the former case many can mistaken think there is the full weight of science behind the findings comparable to the theory of gravity or evolution. They shouldn't make the mistake, but I've seen it made enough times.
You can only draw conclusions from available data, whether that's one or a dozen data points. Being a single data point doesn't invalidate the study though.
The dynamics may be different with 750,000 subs (on T_D) compared to ~20,000 in the ones mentioned in the study. Political-related speech won't disappear from Reddit, especially with an election year coming up.
> The dynamics may be different with 750,000 subs (on T_D) compared to ~20,000 in the ones mentioned in the study.
They may be, they may not be. So far we have evidence that, in general, banning a sub doesn't result in it simply diasporaing into other subs. "Something different will happen this time because I feel like it will" isn't a terribly compelling counter-argument in and of itself. Absent new evidence, and given the evidence already in hand, this seems like a reasonable move to me.
> Political-related speech won't disappear from Reddit, especially with an election year coming up.
I don't think Reddit is attempting to "eliminate political-related speech" in general so much as the "I'm ready and willing to put a bullet in an Oregon cop's skull" speech that was being given free-reign in this subreddit?
> I just think people may be less likely to give up on sharing political speech
But again, that's probably not what they're trying to do so much as get people to give up on talking about organizing ad-hoc assassination squads of law enforcement officers
One of the hardest things to do, I think, for a platform like reddit is to get users to curate and specialize the content and create real communities that people can be involved in.
For reddit, that's their specialty. That's a solved problem. In their specific case their subreddit user groups don't act as filter bubbles. By some miracle they largely act as intended... As communities for like minded individuals to share their interests and largely in a positive context.
The money and advertising and focus is always on the "front page" and the jockeying for position to make headlines and drive traffic to affiliates but this is a huge distraction.
Reddit should focus on investing in the positive and great communities that are built on the site. Build tools to help these communities do what they do even better and use more front page real estate to drive people to these positive experiences since they already exist today. That's the real value of reddit.
Some censorship is inevitable, some bad actors need to be expelled from the site, but ultimately you need to lift up the good examples, not just play "bop a troll"
I have the feeling that organizational energy isn’t cheap, and that repeated banning or asymmetrical resource burning will eventually cause organizations to decay.
When did reddit change their frontpage behavior? Previously you'd only see posts from the default subreddits on reddit.com, now you have a chance of seeing something gamed from an alt right subreddit? Why did they ever ditch the whitelist model?
"It’s very easy to remove spam, bots, racial slurs, low-effort trolls, and abuse... But once you remove all those things, you’re left with people honestly and civilly arguing for their opinions. And that’s the scariest thing of all."
