At a ‘big data’ conference in San Francisco held in Oct 2015, Nate Silver announced - ‘Big Data has Peaked, and thats a Good Thing. Unbeknownst to Nate Silver, his own ‘predicting’ ability (based on ‘big data’ or statistics) would become the biggest example to prove his statement right.
In 2008, Nate Silver was considered by Time magazine among the “100 most influential people” for correctly predicting US election outcomes in 49 out of 50 states. He repeated his performance in 2012, but this year, he is having a horrible time.
Mish wrote in December -
Silver concludes his “Dear Media, Stop Freaking Out About Donald Trumps Polls” article with …
So, could Trump win? We confront two stubborn facts: first, that nobody remotely like Trump has won a major-party nomination in the modern era. And second, as is always a problem in analysis of presidential campaigns, we dont have all that many data points, so unprecedented events can occur with some regularity. For my money, that adds up to Trumps chances being higher than 0 but (considerably) less than 20 percent. Your mileage may vary. But you probably shouldnt rely solely on the polls to make your case; its still too soon for that.
Mish was right that Silver blew up big time in December.
Mish again came back to the topic in January -
Nate Silver Off the Mark on Donald Trump Nomination Odds
I am a big fan of Nate Silver. His calling of the last two elections was nothing short of brilliant. However, I just cannot accept Silver’s current assessment of Trump’s chances of winning the Republican nomination.
In footnotes to his January 8 article Three Theories Of Donald Trumps Rise Silver expressed belief that “Trumps chances are about half of what betting markets say they are. I think theyre about half that - 12 or 13 percent.”
On the Democratic side, Silver has been equally clueless.
Nate Silver and his team are intelligent, reasonable people. From his perspective, there’s really no reason his statistical methods won’t lead him to another successful election prediction. Most of the time, his analyses have worked and it’s reasonable to expect they’ll work again. And that is where the logical flaw in his statistical analyses lies.
Nate Silver hasn’t come out and blatantly predicted that Hillary Clinton will be the Democratic nominee, but here are a few titles of articles he or his team has written: Bernie Sanders Could Win Iowa And New Hampshire. Then Lose Everywhere Else, The Bernie Sanders Surge Appears To Be Over, and What Happens If Bernie Sanders Wins Iowa. The articles don’t specifically say Sanders won’t win, but the real evidence of his prediction is in his Silver’s Polls-Plus Forecast.
What is going on?
In our understanding, Nate Silver has been following a methodology that extends the existing trend. We have seen ‘smart’ people at Bears Stearns and Lehman doing something similar in the housing market in 2008, and they failed miserably at the turning point. Similarly, Nate Silver’s methods are failing, because the country is at a turning point.
Where Silver Went Wrong
Silver has been given high praise, and deservedly so for calling past elections, Nonetheless, Silver has surely blown Trumps odds of winning every step of the way this go around.
Silver went wrong by going with history rather than attitudes. That was a huge mistake for which I do not believe Silver has admitted.
Usually attitudes and history align. This time they didnt. And it was obvious at least three months ago.
Maybe it is time for Nate Silver to drop out of race :), as people are asking in twitter.