> I'd focus on the people who predicted it perfectly
It appears that I have a view not being mentioned here: Sure, can pay attention to issues, using data on the voters, crafting messages based on the data, making predictions, etc. But to me in this election, all of that was small potatoes compared with what really caused Trump to win.
Here's my guess:
(1) Analysis and Arithmetic. Broadly Trump did see from somewhere some statistics on states, counties, and voters. Then he did some simple arithmetic on votes and the votes in the Electoral College. Okay, we all can agree so far.
(2) His 757. Then Trump got in his Boeing 757 and started flying to selected locations to give speeches, rallies. Eventually a favorite approach was to rent a hanger at an airport, arrive in his plane, park his plane in front of the hanger with the crowd, give a speech, and then fly to the next stop. That was an efficient way to motivate 5,000 to 30,000 (in a sports stadium, etc.) potential voters at a time. And a lot of those voters communicated with others and influenced them. That he did this is easy to see.
(3) Polishing Lines and Delivery. Now we come to what appears to me to be the big, crucial point: At each of the rallies, Trump delivered one by one his then favorite lines, looked at the audience, and just observed, no numerical data, just with his two ears and eyeballs, how good the reaction was to the line. Based on such reactions, he selected and refined his lines and his delivery.
(4) Application. Then he took that technique in (3) and applied it like his life depended on it, maybe five rallies in three states in one day, then four hours of sleep, maybe on his plane, and do it again.
(5) Targeting. Crucially near the end of the campaign, he raised the intensity in (4), carefully selected his rally locations, and, then, applied his polished lines and delivery as in (3).
E.g., why was the election so close? Because really early on the mainstream media was correct -- Hillary was ahead in both popular and Electoral College votes. Then, for his attack, especially near the end, one candidate swing state, one promising county, one good rally location at a time, Trump did JUST ENOUGH to win the popular vote in that state and, thus, the Electoral College votes.
In a word, the Trump secret sauce was hustle as in super-salesman drive and 20 hour work days, 7 days a week, determination, in two words, insightful hustle, in the words of Ross Perot, personal "finger tip feel" for the people at his rallies but along with careful attention to the Electoral College map and the maps of the candidate voter characteristics of the states and the counties.
Any good salesman constantly watches his candidate customer, second by second, during the pitch and adjusts instantly, appropriately accordingly. Trump watched and adjusted, a week at a time, a day at a time, a rally at a time, and at each rally a second at a time.
All of that Trump super-selling skill, hustle, and determination is tough to quantify, analyze, and use to make projections, but IMHO that skill, etc. was the key to his victory.
Or, setting aside big issues of the future of civilization, Trump wanted to win, went to where Electoral College votes were potentially available, and got the votes by doing well with what lines would please the corresponding audiences.
Net, actually, in some important ways, Trump is a bright guy, e.g., he is good at filtering information from others, keeping only the best stuff, and understanding that; he has good intuition; although he has some rough edges that hurt him at times, he is good at being a super-salesman, one person at a time, one audience at a time; he is highly determined and takes a lot of personal pride in winning; he can go for months working 20 hours a day, 7 days a week.
Tough to use analysis of numerical data 12-18 months ago to predict the outcome. Better: In the last weeks of the campaign, look at some maps, county by county in the swing states, observe how good Trump was at using the Internet to announce his rally times and locations to candidate audiences, observe how good Trump was at getting his rally audiences up on their hind legs, look at where Trump was giving rallies, see how hard and effectively Trump was working, and then do some more vote counting.
Democracy is supposed to be about getting votes, for POTUS in the Electoral College, and that is what Trump concentrated on and did.
For the future? Maybe like a real super-salesman, Trump is continuing to sell to his audience, e.g., Twitter, meetings televised on YouTube, a weekly report, a big rally in Florida, etc. Trump has some bitter opponents, but he can continue to get his audiences enthusiastic and applying pressure to politicians, the media, etc. Likely before the end of 2017, nearly all the bitter opponents will tire of fighting, be discouraged from losing, calm down, and do something else, and the Trump Administration will be less contentious and more normal. E.g., for al the screaming about immigration, Trump can simply say that he's just actually enforcing long standing US immigration laws, policies, and procedures, and eventually that will be enough to get nearly all the opponents to calm down.
It was a bitter campaign. The Trump opponents tried hard to scare a lot of people about Trump; a lot of people are still scared but will calm down as they don't see actual big things to be scared about.