Media Wrong Again? Near-Foolproof Model Gives Trump 91% Chance to Beat Biden

A statistical model that correctly predicted five of the six most recent presidential elections has given Donald Trump a 91 percent chance of winning reelection in 2020.

Helmut Norpoth, a professor of political science at Stony Brook University, said his forecast is "unconditional and final."

"The polls don't faze me," he told We'll Do It Live in an email exchange.

Norpoth famously called Trump’s upset victory in 2016 when almost all other modelers and pollsters had Hillary Clinton as an overwhelming favorite.
  • Since 1996, he has only erred in one election: Republican George W. Bush’s narrow defeat of incumbent Democratic Vice President Al Gore.
  • The model relies largely on primary performance, not on opinion polls.
Applied retrospectively back to 1912, when major party primaries were first introduced, Norpoth has said the model would have correctly guessed 25 of the 27 elections.
  • The only other miss would have been in 1960, when Democrat John F. Kennedy edged out another incumbent vice president, Republican Richard Nixon.
As in 2016, Trump, the Republican, is almost universally favored to lose this year.
  • Democratic challenger Joe Biden, a former vice president, has consistently polled nearly 10 points ahead of Trump in national polls and was recently found to have big leads in six swing states.
Keep America Great?: Trump and his campaign have publicly accused pollsters of again underestimating his appeal.
  • But the president and his advisers have privately accepted that he’s losing, according to reports.
  • Even Fox News host Tucker Carlson has warned Trump is at risk of defeat.
Still, Norpoth has defiantly stood by his prediction of a second Trump term, which he first announced on March 2. 

"The terrain of presidential contests is littered with nominees who saw a poll lead in the spring turn to dust in the fall," Norpoth told Mediate on Wednesday. "The list is long and discouraging for early frontrunners.”
by is licensed under