@BGMAN: "As I said before, 2011 and 2015 samples are not statistically valid. The sample size is too small. Two points on a graph will always give you a straight line. The whole world cannot be wrong and only the PNC is correct. "
How could 2011 and 2015 not be statistically valid ? Why is the sample size too small? Django is talking about the entire elections results . That's the population on which voting patterns could be discerned.