Pollsters re-weight their poll results to match a pre-determined demographic breakdown, which varies from pollster to pollster based on their own assumptions of the race. If they didn't, the raw numbers would be all over the place, since the sample size isn't large enough to capture the proper demographic proportions. This is not the same thing as the "registered" versus "likely" voter issue--that's based on the questions you were discussing earlier, but not the demographic breakdown.
"Unskewing," at least as it pertains to the action often derided here and in political polling circles in general, usually involves saying you don't believe a poll because they didn't sample enough Latinos, or blacks, or something along those lines. That is not the same as what's being done here, which is arguing that the poll's demographic weighting seems unreasonable.
#uneducatedUnskewersHillary2016
I'm talking about the pollsters' demographic turnout breakdown. That being said, I am not sure where I remember reading about that (FiveThirtyEight?), and it's very possible I'm remembering wrong. Memory is a fickle thing. If that's the case, disregard my previous message.
That being said, your attitude is out of line (if I'm wrong, a simple correction would have sufficed). Have a nice election season.
It would be nice if the moderators enforced some level of civil discourse, wouldn't it?
