Orbis Tabula
Member
Meh, I think if you checked, you could still find this poll to be more suspect. I prefer aggregates, but using them comes with the tacit assumption that each poll is at least going to be worth adding to the aggregate in the first place. If I ran around and did my own poll that showed Trump winning 90-10 (and then repeatedly did this, getting the same results) I think you would rightly be suspicious instead of just trusting an aggregate with my polls in there.
You're probably right. My point is that people aren't going to check. Or at least, they're less likely to. If Rasmussen polls that show Trump up are worth disecting, so should be Rasmussen polls showing Trump down.