The science of predicting who wins an election is imperfect.
The challenge of making it perfect doesn't faze one Roanoke College professor who has been polling voters for years.
Chances are many of you have been pestered by phone calls asking which candidate will get your vote.
As Professor Harry Wilson pointed out, the numbers always change even though the methodology of polling doesn't.
For Wilson and his landline warriors, from college students to retirees, this is go-time.
Roanoke College's polling maestro says his methods are the same as everybody else's.
“We statistically weigh the results at the end pretty much the same way everybody else does. There are some minor variations from poll to poll,” Wilson said.
For a man conducting polls year-round, November is the time they matter most.
Wilson views polls as a snapshot of who would win if the election were today, that past polls aren't as important as the one he's doing right now.
“You have to take those with a grain of salt,” Wilson said. “Even from someone who does polling you have to take them with a grain of salt. As we get closer, we should be better in terms of predicting that.
Wilson and his team collect data from 600 people.
They work to ensure an accurate representation of demographics, asking things right off the top like age, location, gender and race.
Even with all that, some polls still yield bad results.
“You can get a sample that's disproportionately conservative or disproportionately liberal,” Wilson said. “It does happen every once in a while. It happens to everyone, no matter how big your operation or how small it is. “
To get 600 people, Wilson says the automatic dialer he uses calls 12,000 people.
So when all you see or hear is a percentage of who is winning and who is losing, there's always more than what meets the eye.
“I see polling as an art and a science,” Wilson said. “There's a science to it. We're not just making up these numbers. We're actually calling these people.”