re: #43 ObserverArt
Thanks…I was the one asking about it.
I’m still trying get my head around it all and judging how full of crap it is or is not.
Relative strength in the primaries is a pretty good predictor of the general election. But it fails in a lot of cases. What Norpath did is tune his primary model with a some things that made his “predictions” work, like making two term incumbency for a party a big drag and adjusting for partisanship changes in the electorate. Also many of his predictions are within the margin of error, making them meaningless.
And don’t forget, it really crashed and burned in the 1960 election.