The "Bernie versus everyone else" vote totalling is not just silly, it's also contrary to the data.

[TwilightZone]

9. No, it's understanding how the electorate thinks.

Last edited Wed Feb 12, 2020, 11:09 AM - Edit history (1)

If a significant portion of the Democratic base thinks Sanders is going to lose in November, they will gravitate toward someone they believe can keep him from being the nominee.

We may already be seeing that happening. Sanders underperformed in both IA and NH, Buttigieg and Klobuchar overperformed in NH (and Pete in IA), and Bloomberg is picking up significant support, including minority support, the latter somewhat unexpectedly. Biden and Warren are in serious trouble, after each lead the race earlier in the cycle.

The problem with trying to predict is that the electorate isn't a monolith and can change on a whim. General election polls taken when the primaries have just started mean little. National polls at this time mean even less -- this election will be determined by six or seven states. They also can't account for months of hammering by the GOP media machine.