That's probably going to be a majority of the state.
No? There was only about 30% of the state that did that in 1994 when Weld outran Mitt Romney by 30 points. Assuming of course that there was no significant amount of Romney-Roosevelt voters, which I suspect there wasn’t.
Yeah, I was assuming, through a combination of bad data prediction and bad math, that the respective opponents are just some Also Rans that barely make a dent—there's a difference between winning 60/40 and wiping somebody off the map, which means the required crossover vote is that much lower.