Monday, September 07, 2020

Predicting results of our presidential election.

I want to pass on an email from retired historian Lou Wangberg, a friend from my Fort Lauderdale period who leads a history discussion group there. He gives links to analyses by The Economist magazine and Mind Matters A.I.
Between now and the election on November 3rd you will see literally countless polls and forecasts. You see, as humans we hate to live with uncertainty. We want to know in advance what is ahead for us. And it is a legitimate question to ask, "Which poll(s) should I pay attention to? "Which poll(s) will be most on target?" the answer is NONE OF THEM. In the end virtually all of them will be wrong for one reason or another. That is why we hold elections. But I have found one that is more sophisticated than the rest and uses 20,000 factors. It also goes day by day so it is interesting to watch. So give this one a try and you will at least have something you can quote in your discussions (arguments) with friends. I'm going to share an additional link that will explain why most predictions are bogus. Enjoy! Here are the links:
The Economist - Forecasting the US elections
Mind Matters - Election Models: Predicting the past is easy - and useless

No comments:

Post a Comment