and the media's skewed attention to that orange buffoon was unprecedented in a presidential election

People don't really understand probability or percentages. Even a 70% chance of winning an election means roughly speaking that you win 2 out of 3 times. The third time you lose. There aren't many things that you would bet on if the stakes were high when you could lose 2 out of 3 times. But yet people interpret 70% as "near certain". Furthermore, there's something known as "statistically significant" which for simple systems is a sample size of like 35 or so. Probability means nothing for an individual event, and very little for samples of 3, 4, 5, etc. What Nate is trying to explain is that there is WAY more to his analysis than one number. Most people don't know there is a difference between precision and accuracy. Precision is getting the SAME answer. Accuracy is getting the RIGHT number. His polls, while precise, don't have nearly the same level of accuracy.

He was trying to explain too, right up until the end, that the magnitude of the "undecided" was so large, that depending upon how they "split", it could make the difference in the whole election. And that's exactly what happened. In the right 4 states, JUST enough people voted the "wrong" way to result in a shift in the outcome. They do analysis at 538 where they change the polling data just a "little" and see what difference it makes (it's sometimes called a "Monte Carlo" analysis). And as he tries to present, small changes in the polling data resulted in huge changes in outcome. Think about a marble on a "flat" table. Any small imperfection in the table is going to determine if the marble is going to roll, and in what direction. Stick a matchbook under a leg and the imperfections won't matter. But on a table that is truly "level", the small imperfections will decide everything. After the Comey letters, and the fake news the table was VERY level. Rain, a traffic jam, an election day sales at Sears on Play Stations could shift an election.

MSM media corruption is homogeneous only in its bias toward sensationalization, especially overstating problems, to maximize profits. Not to minimize that, though, as it's been very destructive.

On individual subjects, coverage can be honorable and excellent on most and still be corrupt on others. Hillary is one of the others -- almost every story was given the negative spin required of all who wanted to attain and keep prestigious jobs.

Another, ultimately far more monumental corruption, though, is not reporting on the degree of control and diversion to the far right of the Republican Party achieved by ultraconservative anti-government donors. That failure, including the reasons behind it are, is frightening.

The media most people watch cover issues and events without ever mentioning this huge factor. They now routinely refer to the Kochs' extremist "Freedom Caucus" as congress's conservatives and people like Graham and McCain as "moderates," as if by far most genuinely moderate conservatives hadn't been purged and replaced with strong and extreme conservatives long ago. A huge warping of and veil over political reality embedded in everyday language of reporting.