Clizby Wampuscat
Well-Known Member
I did a state by state analysis of the final poll forecast and the actual results of the 2016 and 2020 elections. Here are some facts:
2016:
Trump over performed the polls by an average of 7.8 points in the states he won. No states did he underperform the polls.
Clinton over performed the polls by an average of 1.9 points in the states she won but had 7 states she under performed.
Trump flipped 5 states:
Florida Polls had Clinton at 0.6 points ahead, Trump won by 1.3 points
Michigan Polls had Clinton at 4.2 points ahead, Trump won by 0.3 points
NC Polls had Clinton at 0.7 points ahead, Trump won by 3.8 points
Pennsylvania Polls had Clinton at 3.7 points ahead, Trump won by 1.2 points
Wisconsin Polls had Clinton at 5.3 points ahead, Trump won by 1.0 points
Clinton flipped zero states
2020:
Trump over performed the polls by an average of 5.7 points in the states he won. No states did he underperform the polls.
Biden under performed the polls by an average of 3.0 points in the states he won but had 21 states he under performed.
Trump flipped 1 state:
NC Polls had Biden at 1.7 points ahead, Trump won by 1.4 points
Biden flipped zero states
2024:
I took the poll forecast from 9/25/2024 and compared to 2016 and 2020 and accounted for how Trump over performed the polls and was as generous to Harris as I could and here is what I found:
Trump will flip 3 states:
Michigan Polls have Harris at 2.0 points ahead, Trump will win by 2.5 points
Pennsylvania Polls have Harris at 1.0 points ahead, Trump will win by 2.5 points
Wisconsin Polls have Harris at 1.6 points ahead, Trump will win by 6.0 points
If all this holds Trump will win by 305-233 EV's. The interesting thing is Trump just needs to flip one of these three states to win.
Also it is amazing that Trump in the last two elections never under performed a poll. This has to be an issue with the polls being biased.
If you disagree with this analysis, why would the polls be accurate now and not over the last two elections?
All poll data is from the 538 election forecast.
2016:
Trump over performed the polls by an average of 7.8 points in the states he won. No states did he underperform the polls.
Clinton over performed the polls by an average of 1.9 points in the states she won but had 7 states she under performed.
Trump flipped 5 states:
Florida Polls had Clinton at 0.6 points ahead, Trump won by 1.3 points
Michigan Polls had Clinton at 4.2 points ahead, Trump won by 0.3 points
NC Polls had Clinton at 0.7 points ahead, Trump won by 3.8 points
Pennsylvania Polls had Clinton at 3.7 points ahead, Trump won by 1.2 points
Wisconsin Polls had Clinton at 5.3 points ahead, Trump won by 1.0 points
Clinton flipped zero states
2020:
Trump over performed the polls by an average of 5.7 points in the states he won. No states did he underperform the polls.
Biden under performed the polls by an average of 3.0 points in the states he won but had 21 states he under performed.
Trump flipped 1 state:
NC Polls had Biden at 1.7 points ahead, Trump won by 1.4 points
Biden flipped zero states
2024:
I took the poll forecast from 9/25/2024 and compared to 2016 and 2020 and accounted for how Trump over performed the polls and was as generous to Harris as I could and here is what I found:
Trump will flip 3 states:
Michigan Polls have Harris at 2.0 points ahead, Trump will win by 2.5 points
Pennsylvania Polls have Harris at 1.0 points ahead, Trump will win by 2.5 points
Wisconsin Polls have Harris at 1.6 points ahead, Trump will win by 6.0 points
If all this holds Trump will win by 305-233 EV's. The interesting thing is Trump just needs to flip one of these three states to win.
Also it is amazing that Trump in the last two elections never under performed a poll. This has to be an issue with the polls being biased.
If you disagree with this analysis, why would the polls be accurate now and not over the last two elections?
All poll data is from the 538 election forecast.