Their two forecasting model’s both favor Joe Biden winning on election day ….
This compliments almost every other forecaster’s call….
(FiveThirtyEight ‘s projection)
It has been tough for them….
They have had to get over the 2016 Donald Trump ‘surprise’ win over Hillary Clinton polls got it wrong….(They didn’t….Clinton won the popular vote by 3 million)
But the situation 6 days from the last Presidential election is vastly different….
— A presidential election forecasting model based on the president’s approval rating, first published at the Crystal Ball in early August, continues to show Joe Biden as a favorite in next week’s presidential election.
— An average of current state-level polling produces a very similar Electoral College projection.
— These forecasting methods produce projections very similar to a more complex model published by FiveThirtyEight.
Forecasting models point to Biden
Models for forecasting presidential elections fall into two broad categories. Many political science models, including my own “time for change” model, use election fundamentals such as the state of the economy and the incumbent president’s approval rating to predict either the popular or the electoral vote. By using measures of these fundamentals from the spring or summer of the election year, such models can make a prediction several weeks before Election Day.
In contrast, the second type of forecasting model uses polling data and, sometimes, past election results and other factors to predict the outcome of the election at the state level and then combines these state forecasts to predict the national electoral vote. Perhaps the best-known example of such a forecasting model is the one published by Nate Silver’s FiveThirtyEight website. These forecasts based on state polling results are typically updated on a daily basis right up until the day of the election….
…
Conclusions
Two different methods of forecasting the 2020 presidential election, one based on an aggregate level model of the national electoral vote and one based on individual state polling data, yield almost identical predictions of the outcome. The aggregate level model, first published in early August, predicts a Biden margin of 345-193 in the electoral vote. A forecast based on simply combining the results of recent state polls predicts a Biden margin of 350-188. Both predictions are extremely close to the latest forecast from the much more complex FiveThirtyEight model.