This has been discussed to death in the comments section, and this misconception is bogus because it has nothing to do with support. Internet hits are about interest. The more interest there is in a candidate--whether that interest translates as support or not--the more news coverage he should get.
This was the case with every other candidate except for Ron Paul. The news media based their coverage on their bogus polls. His support wasn't just on the internet (that's just the only place where we have raw data to work with), it was in the real world, with Ron Paul signs everywhere, call-in shows swamped, there were many in the media who said that the person they had the most requests to learn more was Ron Paul. The only reason they thought this must have been anomalous was because of their polls. And as I pointed out in the episode, maintaining this on the air just fueled the public perception that there was no real support (even after he started out-polling McCain and others in the primaries), making it a self-fulfilling prophecy. It's like how people think that overall crime, school shootings, and other things the media reports a lot are on the rise, when really crime rates are dropping and school shootings are so rare as to be freak occurrences.
Jerry Day was being conspiratorial, and he had a clear Ron Paul bias, and I edited out a lot of that and just showed his analysis of the data, which I maintain is valid even if it doesn't point to a conspiracy. As I said, they probably believe very sincerely that their polls are accurate. That's why I downplayed the conspiracy angle: it just isn't relevant. He used good data and took it in the wrong direction; I think I took it in a better direction.
I think I did the same with Zennie Abraham. He had a clear pro-Obama bias, and he wandered into conspiracy territory himself, but I think I pulled out the good points he made and took them in a better direction.