““On October 18, before the numbers had been re-weighted, the campaign’s internal election simulator, dubbed the “Battleground Optimizer Path to Victory,” gave Trump a 7.8 percent chance of winning the 270 electoral votes he needed. His long odds were mainly due to the fact that he trailed (albeit narrowly) in most of the states that looked like they would decide the election, including the all-important state of Florida. Re-weighting the model to reflect an older, whiter electorate changed the polls by only a couple of percentage points in most of the major battleground states. But it shifted the odds of victory substantially in Trump’s direction because it either put him ahead or made him””