My 5 cents on fivethirtyeight as I follow the site closely:
fivethirtyeight is not biased
I think the correct statement would be: fivethirtyeight tries to be as unbiased as possible.
We can't claim or know that they are unbiased. Just like most polling firms and data aggregation sites, they have done a lot right and made some mistakes.
They try to look at polls as scientifically and statistically rigorous as possible, but that doesn't mean it's automatically correct. They are probably the most likely to be correct, but that doesn't mean they automatically are.
It's fun to claim that they did so well in previous elections, but Nate Silver himself would be the first one to tell you that this doesn't mean much. The sample size of American elections is small and polling misses happen regularly (see the latest British elections, the Brexit was NOT a polling miss, any reasonable polling aggregator had it a very close race and it ended up being very close (the margin of error was only about 2%))
Also not that fivethirtyeight is nor was claiming that Clinton would win the elections. Last time I checked, she was pretty constant at about 85% to win. This gives Trump about 15% to win which was about the same pregame probability that San Antonio would defeat Golden State (which they did). Again, this paragraph is loosely based on something Nate Silver himself wrote.
On a more general note, the election still needs to happen. Clinton is more likely to win than Trump at this point based on the data that we have. But the range of outcomes is wide (see fivethirtyeight on this, there are multiple articles in which they explain this again and again). Trump could win if some things go right for him, including some of the claims made in this topic. Clinton could win in a landslide if some things go right for her, including some of the claims made in this topic. It is impossible to perfectly model everything that could influence the results and particularly turnout among demographic groups, the driving force behind a lot of the unknown variation. Fivethirtyeight nor most other sites are claiming Clinton will win, not even jokingly. They (correctly) claim she's the current favourite the win but that the outcome is subject to this unknown variation.
My advice for Trump supporters would be to accept that he is behind, but that the election is not lost yet. If you want him to win, you waste your time yelling on the internet about rigged polls and media. Instead, try to invigorade the base and make sure you and your fellow supporters vote.
My advice for Clinton supporters would be to accept that the election is not won (yet). If you want her to win, you should not try to argue on the internet about how Trump supporters are desillusional. You may be equally desillusional. Instead, go out and make sure that everyone who supports her actually goes out to vote.
Note, this is exactly what the Clinton campaign is actively doing the last few weeks, but the Trump campaign isn't. This is a reason why some media claim Clinton may be more likely to win than polls indicate (her likely voters would be more likely to vote according to that theory). An opposing theory would be that Trump supporters are already more riled up and ready to vote and don't need an extra push.
