Reply #36: Political predictions more than a few days out are like long-range weather forecasts

[andym]

Political predictions more than a few days out are really like long-range weather forecasts or possibly like the Farmer's Almanac's weather forecasts. They may sometimes be right, but it's usually a matter of luck. Polling, especially more than a few days from an election is just not a reliable guide into how the thoughts of the American voter will translate into votes. Too many things (events) can change at a moment's notice and people are fickle. That's why it is nearly impossible to predict what will happen in the General Election. For example, Jimmy Carter was leading Ford by 33% in the polls in 1976 after the convention but only won by 2%.

Of course, your post really deals with propaganda/spin. The prevailing media spin based on the post Pennsylvania polling results is that Clinton will win IN. However, to be fair, it's only now (about 3 days before the election) that the polling has even a chance of being accurate. The Obama's campaign's predictions from February, the Clinton's, and really anyone's predictions from 3 months ago are just not relevant, and would not be taken seriously by anyone except for spin purposes.