Rasmussen Reports has Trump ahead by 4 Points

[citood]

6. Most polls are too small to really be random

So, the pollster crafts the demographics of the sample. Rasmussen probably has fewer D's and more R's in his sample than other polls, etc.

But the bottom line is that these polls are almost an echo chamber for the pollster's sample demographics. The pollster makes assumptions...and those assumptions are reflected in the result.

Last election, most polls were pretty darn close.

This time? I'd assume that most pollsters are having a hard time 'figuring out' Trump's base of support, and who to sample. But somewhere along the way (convention and debates), he'll have to actually take a position on something...and it will become easier to poll more accurately.

Anyway, as far as I'm concerned, any pre-conventions data can be thrown out.