Strange that the article fails to note that, even during the weekend before the election, almost every polling outfit was predicting a significant Hillary victory, and Trump reversed that by taking 6 major states that he was not expected to take.
NOV 8 2016 actual results:
WI, MI, PA, NC, and FL all flipped to red... that's 90 of 538 (almost exactly 1 out of 6, or 16.7%).
Also... Time had Alaska as merely "likely"... it went 51-36...
and they had Utah as merely "likely"... and Hillary almost finished third (45-27-21)
And let us not forget that Maine (47-44), Minnesota (46-45), and Nevada (47-45) were all in play... and New Hampshire (46-46) was a toss-up. Even Virginia (49-44) and Colorado (48-43) were in range.
Hillary got less than 28.95% in six states. Trump was not that low in any state (worst was 30.03% in HI... although he got only 4% in DC, of course).
Hillary could have been under 190 EVs without much of a change. 270,000 votes was the difference, total, for ME, MN, NV, NH, and CO... VA was 220,000 more. A half million well-placed votes, out of 132 million, makes it a 350-188 final tally.