After analyzing the polling conducted by FiveThirtyEight during the 2016 general election answer the following questions. Why should a poll be scientific rather than informal? Was Dr. Silver’s polling inaccurate? Why or why not? If you were a 2016 pollster, would you have done anything differently? If so, what?

You can see Dr. Silver's polling methodology that he used to forecast the 2016 Presidential election at the link below. He used state polls and weighted each poll average to account for how accurate each pollster's track record was and how recent the poll was. He also weighted the poll according to sample sizes. As he stated, a larger sample size helps the accuracy of the poll, but there are "diminishing returns," meaning that a poll only has to be so large to be accurate.
As Silver wrote before the election (see the link below), there were a range of outcomes possible. Clinton's lead over Trump in the polls during the days before the election was within the range of normal polling error. In addition, there were more third-party voters, which added to the uncertain outcome of the election. Also, while Clinton was popular amongst Hispanic and college-educated white voters, these voters did not tend to live in swing states.
Polling should be scientific, but it always involves some uncertainty. To be scientific, a poll should be sure to survey a wide range of voters and should include a large sample size. While Dr. Silver tried to make his polling as scientific as possible (by including weighting polling that privileged more recent data from more trusted pollsters), it is sometimes still hard to predict the victor, especially in a close election. Perhaps one way the polling in the 2016 election could have been improved was by weighting swing states more heavily. These were the states in which Clinton and Trump were close in the polls. If these states (states such as Florida and Michigan) had been accurately predicted, it might have allowed pollsters to determine the winner of the Electoral College vote with greater certainty.
https://fivethirtyeight.com/features/a-users-guide-to-fivethirtyeights-2016-general-election-forecast/

https://fivethirtyeight.com/features/final-election-update-theres-a-wide-range-of-outcomes-and-most-of-them-come-up-clinton/