Accurate poll numbers ~ or, at least, poll numbers whose results can be interpreted or adjusted to bear some resemblance to reality ~ are useful to candidiates on both sides, since they gain knowledge upon which to base decisions about where and how to spend their resources.

Now, exit polls on Election Day, that's another matter. I absolultely support a total gag order on any hint af actual voting results or "trends" before the last polling places close in the westernmost time zone. But that's not the same thing as these post-debate polls under discussion here, and other research conducted during the long drawn-out campaign season.

I do agree that it is extremely unfortunate if and when individuals base their vote upon who they perceive as likely to win, instead of upon who represents one's own interests best among the candidates.

It's one thing for "vested interests" to study the likely winners and adjust their corporate-giving policies accordingly ~ that's probably unavoidable ~ but there's no reason for private individuals to follow that same mindset. Private individuals are not going to be in a position of influence in the same way as giant corporations and their lobbyists.