Really, I have to give them credit where credit is due: by repeatedly pointing out errors (however trivial) out of the tens of thousands of news stories that are published every day, they've managed to get their supporters to the point where they'll trust a new story on www.siteiveneverheardofbefore.com/newishstuff/hillaryclintonpedophilering.html more than they will an actual newspaper. It's a real masterstroke in terms of controlling the narrative. "Anything negative you hear about me, it's fake, because there exist cases where newspapers have made errors, and we've selectively presented you only with those cases to create a narrative for you that newspapers are packed full of fakery." Not just newspapers - fact checkers, peer-reviewed articles, even official government statistics - all fake, because they've been presented with every case people can get their hands of of error, without the balancing context of the 10000x more that wasn't in error.
In the words of XKCD: "Dear God, I would like to file a bug report".
It's the same thing that contributed to the Challenger explosion. They had a nice clean graph in front of them that plotted O-ring failures vs. temperature. There was no clear trend visible on the graph. The problem was that they omitted the successes, the cases where there were no O-ring failures. Here's what it looked like with that added in. All of the sudden there's a very clear trend of failure increasing at low temperatures - in fact, every low temperature launch had had O-ring failures, while very few high-temperature launches had. By being selective in what data you present (accidentally in that case, on purpose in the present case), you can get people to believe precisely the opposite of what is true.