The problem with your “scientific poll” charts is that they include such clunkers as American Research Group (ARG) and Quinnipiac. Take those two biased surveys out and the results are astoundingly different, aren’t they?
The charts that you are looking at on this thread are for Rasmussen only and are labeled as such. The other charts I post and which are on my profile page incorporate all poll results and use a 4-period moving average to round off the "clunkers". I use the same exact method that RealClearPolitics uses and it is a valid method. If you don't like the results of certain polls, don't look at them. If you don't like my charts, don't look at them either.