Looking At The Presidential Election With FiveThirtyEight’s Election Day Forecast

Highlights from FiveThirtyEight's predictive forecasts.
Share on facebook
Share on twitter
Share on linkedin
Share on email

As part of their Election Day Coverage, FiveThirtyEight simulated the election 40,000 times to forecast what might happen tonight. Here are the key takeaways:

Democratic Nominee Joe Biden is favored to beat President Donald Trump. However, Trump still does have a 1-in-10 shot at winning, and underestimating the incumbent president’s support is “why he still has a path to reelection.”

  • Sample of 100 Outcomes: Biden wins 89 times, Trump wins 10, one yielded no Electoral College Majority.
  • Projected Electoral Votes: Biden 348, Trump 190
  • Projected Popular Vote: Biden 53.4%, Trump 45.4%

A lot of Biden’s chances hinge on Pennsylvania, where he currently leads by 5 points in FiveThirtyEight’s polling average. Losing the Keystone State doesn’t sink Biden but leaves him with around a 30% chance to win the White House.

  • Other Tipping Point States By How Decisive They Could Be: Florida (14.3%), Michigan (7.7%), Arizona (6.4%), Wisconsin (5.3%), North Carolina (5.2%), Georgia (4.2%), Nevada (3.4%), Minnesota (3.1%), Texas (2.4%), Ohio (1.8%), Colorado (1.5%), New Mexico (1.2%), New Hampshire (1.1%)

Likely Possibilities:

  1. Biden wins at least one state that Trump won in 2016 (98 in 100 chance)
  2. Biden wins the popular vote (97 in 100 chance)
  3. Biden wins more than 50% of the popular vote (95 in 100 chance)

Unlikely Possibilities:

  1. Trump wins in a landslide (<1 in 100 chance)
  2. Trump wins the popular vote but loses the Electoral College (<1 in 100 chance)
  3. No one wins the Electoral College (<1 in 100 chance)
  4. The map stays exactly the same as 2016 (<1 in 100 chance)

We’ll Likely Have To Wait: Barring a landslide, it may not be clear who the winner is tonight. It’s unlikely that neither candidate wins the Electoral College, but it’s going to come down to several key battleground races and challenges from mail-in balloting and Covid-19 could slow down how quickly a representative share of the vote can be reported.

Justin Oh:

The data going into the election points to Biden winning decisively. But beneath the numbers, doubt still looms about how accurate the polls will predict actual voter results. The 2016 election showed us that polls can be inaccurate, especially when there is late-stage drama and when many voters secretly support a candidate but are afraid to admit it publicly. 

Hillary Clinton led President Donald Trump by about 4% going into the 2016 election, and 12.5% of voters were undecided. But late deciders broke towards Trump, especially in white working-class states like Wisconsin, where the ultimate result showed significant polling errors. Much of this could be attributed to the third debate and Comey letter drama.

This time around, Joe Biden has a much more significant 8.4% lead, and only 4.8% of polled voters are undecided. I don’t see a compelling reason to predict that he will not win decisively. But if somehow President Trump does win reelection, despite these numbers and the polls having been adjusted, we’ll really have to rethink how predictive the data is with controversial figures.

Share on facebook
Share on twitter
Share on linkedin
Share on email


Your email address will not be published. Required fields are marked *

Related Posts