Sometimes the polls are tricky to follow.
One has Trump ahead.
In another, Clinton could be in front; and the latest one from Reuters actually has the two candidates pretty even right now.
So what can you believe?
Dr. Jason Husser, director of the Elon University poll, says not all polls are created equal.
"If you just take one single survey you're probably not going to get the whole story," he explains. "What you need to do is take the average of all the polls that are our there."
Dr. Husser gives three tips for weeding out the more reliable polls:
1) Make sure the poll is transparent. Who paid for it? Which is the sponsoring organization? Husser says good polls could be off by a few points, but are typically not biased or skewed, so you'll want to know where they came from.
2) Make sure the pollster calls people on cell phones - not just landlines, which is easier to do. That ensures you get a broad, diverse sample.
3) Make sure the poll samples at least 600-750 people.
"By sampling a random group of enough broad people that they're able to simulate what it would be like to talk to every single person in society," Dr. Husser says. "You might think you need to talk to 5,000 people but really once you get over 600 or 750 you have a pretty good sample of what's going on."
As for how influential polls can be, Dr. Husser says people vote on what they believe is right and wrong, not necessarily on poll numbers.
He says they'll start surveying for the next Elon University poll next week.
(© 2016 WFMY)