In 2012 there was a bunch of gee-whiz about how software built by a dream team of engineers drove Obama's re-election: https://www.theatlantic.com/technology/ar

In 2012 there was a bunch of gee-whiz about how software built by a dream team of engineers drove Obama's re-election: https://www.theatlantic.com/technology/archive/2012/11/when-the-nerds-go-marching-in/265325/ https://www.theatlantic.com/technology/archive/2012/11/when-... I suspect that it's fun to read and write about new tech applied to Presidential elections but that the new doesn't actually matter more than the same old. If you wanted a really simple model to predict Presidential elections, this would be it: incumbents get re-elected, and then after 8 years American voters are sick enough of one party to try the other party again. One-term presidents are rare. Maybe you remember the idea of a "permanent Republican majority" from GWB's first term. That wasn't very permanent at all. Maybe you remember the post-2012 fear among conservatives that demographic changes had given Democrats a permanent lock on the presidency. That was short lived too.