Nothing wrong with dreaming. For me, the numbers were never there for Trump to win. Sure he won the nomination, but he won it getting 40% of the GOP primary vote which the Republican base makes up 27% of the total electorate. Transferring 40% of 27% to 51% of 100% is not an easy task. The demographics between Republican primary voters and those who vote in the general election is completely different. But I rarely get into them, I prefer to look at party affiliation numbers which all demographics are include. What percentage of the electorate makes up the Republican base, what percentage is the Democratic base and the percentage of the electorate who affiliate or identify as independents. Then break down the independents into independents lean Republican, indies lean democratic and pure or true independents with no leans. Knowing those numbers which Gallup provides usually once a month, sometimes twice a month if an election is near one can use historical averages, percentages on how each group votes.
It doesn't matter how many whites, blacks, Hispanics, Asian are in each group, just the total number. An example is the Republican base makes up 27% of the electorate and the historical average is that base votes for their party's candidate 91% of the time. So the GOP nominee should get around 24.5% of the Republican base vote. This works great in a normal election like 2012. Where the two major candidates get 98% plus of the total vote.
But this year is so unique with the very high dislikes of each major party candidate and with third parties drawing 10-15% depending on the poll. One must then throw the historical averages out the window. But RCP provides plenty of polls which break down the percentages of how many Republicans are voting for Trump, how many for Clinton, how many for Johnson and Stein, the number of undecided and will not vote. So the numbers are there, all one has to do is averaging them out to come up with a number that should be within a percent of two of how the Republican base will vote. Do the same for the Democrats, the leans which leaves the pure/true independents with no leans.
But again the polls provide the total numbers for independents also. Just take the favorable numbers for one or the other candidate times the number of pure/true independents with no leans.
Or one can take the easy way and just look at RCP's averages in the four candidate field and you will have a pretty reliable feel of where this election stands. But I find it more fun to play with the numbers. They do surprise you every once in a while, give you something you didn't expect.
RealClearPolitics - Election 2016 - General Election: Trump vs. Clinton vs. Johnson vs. Stein
