Not only is this a really cool and novel (as far as I know) use of machine learning techniques, there's a lengthy footnote that goes into some detail about the method. And the presentation is great, very slick modern HTML.
So it seems like an excellent fit for HN, apart from the title which will unfortunately put a lot of people off.
[1] https://github.com/fivethirtyeight/data/tree/master/subreddi...
[2] https://bigquery.cloud.google.com/table/fh-bigquery:reddit_c...
https://bigquery.cloud.google.com/table/fh-bigquery:reddit_c...
They do it constantly to stories deemed "not worthy".
Truth doesn't really matter unless it supports us, everyone who says good things about us is right and everyone who doesn't is a corrupt MSM cuck libtard. To get banned you don't even have to be critical of Trump. If you're even ambiguous you'll get banned. In a sense this is self-selection in that the only people that haven't been banned are lunatic-level supporters.
Basically, it's Donald Trump times 380,000. You have to experience it.
This is just how Americans are these days. You have to pick a camp, and you need to agree 100% with everything in that camp or else you'll be shunned. And there's only two main camps too; if you try to pick an alternative camps, or even a camp that's really a sub-camp of one of the two main camps, you'll be shunned and derided.
shrug It's just NeoGAF with some surface level differences.
GAF is wider focused, has a different position, and uses different memes to be dismissive of disagreement. But overall this describes it's type of political discussion and moderation of wrongthink to a T.
