All Rights reserved
The out election forecasting whiz, Nate Silver, admitted today he "screwed up."
Silver had from the start dismissed Donald Trump's chances of winning the Republican nomination, and he didn't do it based on the numbers.
"We didn't just get unlucky: We made a big mistake," wrote Silver today, explaining how he and the rest of his staff at FiveThirtyEight lost their bearings.
In August 2015, Silver estimated that Trump had just a 2 percent chance of winning the nomination, labeling him a "fringe" candidate. "Trump's campaign will fail by one means or another," he predicted.
Silver very slowly revised upward those teeny chances. By the end of last year, with Trump continuing to dominate all polling, Silver still had Trump's chances in the single digits.
"The big mistake is a curious one for a website that focuses on statistics," he wrote. "Unlike virtually every other forecast we publish at FiveThirtyEight -- including the primary and caucus projections I just mentioned -- our early estimates of Trump's chances weren't based on a statistical model. Instead, they were what we sometimes called 'subjective odds' -- which is to say, educated guesses. In other words, we were basically acting like pundits, but attaching numbers to our estimates. And we succumbed to some of the same biases that pundits often suffer, such as not changing our minds quickly enough in the face of new evidence."
They became victims of "a kind of rigid empiricism," he said.
In typical Silver-style, his column today is more than 5,000 words and comes with 18 footnotes, explaining every possible facet of what was overlooked.
Admirers of Silver's data-based approach to interpreting elections are forced to reconcile this misstep with his past success. Silver famously predicted the results of the 2008 election of Barack Obama in 49 out of 50 states. He also got all 35 U.S. Senate races right that year. Most recently, he picked every state correctly in the 2012 Obama versus Romney match-up.
Still, Silver isn't admitting he got it entirely wrong in this cycle's Republican primary, noting that basically no one saw Trump's success coming, and warning against something he calls "hindsight bias." In other words, Silver doesn't expect to dramatically revise his famous "poll-plus" model for predicting election outcomes.
"Trump is one of the most astonishing stories in American political history," wrote Silver. "If you really expected the Republican front-runner to be bragging about the size of his anatomy in a debate, or to be spending his first week as the presumptive nominee feuding with the Republican speaker of the House and embroiled in a controversy over a tweet about a taco salad, then more power to you."
He added, "Sometimes, low-probability events come through."