haxxiy said:
Their worst result was the last 2018 governor's poll, but back then there were 7% undecideds that could have broken 5-to-1 for the Republican candidate. Now it's just 3% undecideds plus 2% who didn't reveal their vote. Even if they are *all* Trump voters, and everyone saying RFK just votes Trump instead, he's lost a precious 2-3% vs. Harris compared to Biden in the Midwest that will shore up margins nicely in the other states she needs. |
Oh sure but my point was that, would the methodology be different for president versus governor/senator polling? Do we put more weight in the governor result or more weight in her presidential specific accuracy?