Below in an excerpt from a very good basic primer by Kristen Soltis Anderson on how to think about polls. I particularly like this part:
…it is important for us to separate out the “predictive” quality of these horse-race polls (they are not really predictive, despite everyone loving to use them as a forecasting metric) and the “validity” measures that give us reason to believe or doubt polls as they stand today.
When deciding if you “trust” the polls, I would encourage people to stop worrying about whether these polls are predictive, because they really aren’t. I do think we need to be very critical about whether or not these polls are valid measures of this current snapshot in time, and I think there are important questions to be raised on that front.
I’ll add to it that valid polls largely only become predictive ones after the fact; you can easily have a situation where a poll is accurately describing the national political environment until it suddenly isn’t. I think that it’s generally safer to keep in mind that it’s safer to analyze polls in terms of the present than in trying to use them to predict the future. And if 2012 taught us anything, it’s that trying to mutate any one result in order to fit them fit the past is not a good idea…