BTW the debate poll thing gives me an opportunity to lament the way in which the media are incapable of presenting poll results in a way that provides proper perspective. First of all, they just can't seem to understand that the "margin of error" for the poll is the "margin of error" for the percentage favoring a candidate. It's not the margin of error for the difference.
In this case, the result of the CNN poll that had 48 percent of respondents saying Pence won while 42 percent said Kaine won is a "statistical tie." While the margin of error for the percent favoring an individual candidate is 4.5 percent, the margin of error for the difference between the two candidates is about 8.5. The 6 point difference in respondents for who "won" the debate is within the margin of error.
The other thing is that CNN kept saying the sample "over represented" Democrats. But the sample they got was pretty consistent with what exit polls have said about the electorate in recent previous elections. Here is a table to illustrate what I mean:
The distribution of party affiliations in that poll was not inconsistent with the distribution of party affiliations in the past two general elections. The difference between 41 and 38 or 39 percent, for instance, can easily be random sampling error. There was no bias in favor of Democrats.
Anyway, the point is that the media have created the impression that Pence was clearly perceived as the winner by the majority who watched the debate when that is not the case and CNN itself created the impression that it's own poll was biased in favor of Democrats when that ALSO is not the case.
I don't know if it matters but in some cases it might.