President Donald Trumpin the probability of re-election does not look good based on polls and analyzes.
Before this week’s election debate Joe Biden led in nationwide support measures by 7-8 percentage points. According to the FiveThirtyEight site, which is lowering its own prediction of polls and emphasizing polling companies on the basis of reliability, Trump’s distance is now greater than anyone’s since 1996 and Bob Dolen.
But elections are not decided by the number of votes in the whole country, but by the electorate. Also in that look, Biden is on the neck. In many important Libra states, including those that Trump won in 2016, Biden is 5 to 7 percentage points ahead. These include, for example, Pennsylvania, Michigan and Wisconsin.
FiveThirtyEight is now analyzing its fourth presidential election, and on all previous occasions, there has even been some ambiguity as to which candidate is in the lead. John McCain visited at one point Barack Obaman above in 2008. The race between Obama and Mitt Romney in 2012 was extremely tight, with Romney leading virtually throughout October and falling below one percentage point in the final predictions. Four years ago, at this point, in the stages of the first debate, Hillary Clinton led by only 1.4 percentage points.
This year, Biden’s management has been at its worst 6.6 percentage points in monitoring the site. Another Gallup-averaging player, RealClearPolitics, which does not value reliability, had a difference of 4.5 percentage points at its lowest in May and now 6.6.
The watch is also ticking against Trump. There is only exactly a month left until the election, and many voters cast or have already cast their votes in advance.
What if the gallup error recurs?
One common perception — as is typical of populists, in part, one fed by Trump and his supporters — is that there is a big mistake in the polls, and that mistake is in favor of a sitting president. The rotator of FiveThirtyEight Nate Silver knock out this.
– There is no guarantee that the mistake will be to Trump’s advantage. Historically, the direction of the Galllup error has not been predictable. The same polls that underestimated Trump in 2016 underestimated Obama and Democrats in 2012.
His site and prediction are criticized by claiming that it assumes the polls are right.
– The whole purpose of the forecast is to assess the probability that the polls are wrong, Silver says.
For example, four years ago, the site gave Trump a remarkably good 30 percent chance of winning compared to other forecasts. Now the forecast is 21 percent. If the election were held today, the figure would be only nine percent. So FiveThirtyEight predicts that Biden’s leadership will shrink by election day.
Admittedly, SIlver estimates that the poll error will be greater this year than in the last election due to the wider postal vote brought by the coronavirus, and will seek to take that into account in his forecast as well.
The opposite possibility also exists.
– The wider popularity of advance voting means that there is now less time for some big, changing event, Global Strategy Group’s Vice President of Research Mario Brossard tells Newsweek.
In any case, Trump would clearly need a bigger mistake four years ago to win. Even if Biden’s margin were cut by three percentage points in each state, he would win the previous election that clearly settled Pennsylvania, Michigan and Wisconsin and would be ahead of Arizona as well.
The New York Times ’election data department, The Upshot, has calculated that if the poll error is the same in all states as four years ago, Biden will win, but narrowly (280 against 258 voters). If, on the other hand, the mistake is like 2012, Trump will suffer a sled loss.
Both Silver and the authors of the polls are able to point out that the possibility of error – even such a large one – always exists. In 1948 Harry Truman was five percentage points behind Gallup in the measurement and won by almost five percentage points. In 1980 Ronald Reagan the meager lead turned into an urn crush in the urn.
– Given how eventful this year has been, I think it makes sense to exercise restraint in all attempts to predict the future, Ipsos polls Chris Jackson said recently to Newsweek.
Jackson points out that this election should not be compared to the last election. Four years ago, a new FBI investigation into Clinton’s emails rocked the situation significantly.
A significant difference from four years ago is the number of undecided voters. According to various measurements, they are now between 3 and 11 per cent, compared with 20 per cent at their best four years ago and 13 per cent on election day. In 2012, on election day, 4 percent were insecure.
Four years ago, the media and many forecasters announced Clinton’s victory, but ignored the large number of uncertainties. Trump, who made his last-minute choice, clearly won by the ballot box.
This was also the case in 1980, when 17 per cent were insecure on election day: many of them jumped on Reagan’s sled and made the polls look miserable. Indeed Jimmy Carter in that they liked their itching really well. The pre-election forecast was 43-40 in favor of Reagan, with a final score of 51-41.
According to data collected by FiveThirtyEight, lower numbers of uncertain voters predict more accurate polls.