House polls are generally not perfect.
Oh, aye. The reason being that you can't take a scientifically unbiased sample of a given human population and that professional pollsters know how to cheat and get a representative-on-the-questions-they-care-for sample anyhow, in populations they know their way about. If House Districts had easily defined borders and stayed the same over lengthy periods of time, there'd soon be better polls of them.
Siena came up with interesting crosstabs for this Gibson poll. Gibson is winning 2-1 in his old district and only by 5 in the new parts.
I wonder if upstate is going back to its old holding pattern where incumbents hold marginal districts with ease.
I hope that's the case; that would be good for Hochul too.