That’s not true either......man understand what polling is about....and therefore do not put the “value” polling wants to have in determining the race.
I agree that it “might” at times indicate a trend...but even then considering how few they poll to the actual voting population is almost laughable.
Like the media...’pollsters’ want the ratings and engagement they get.....and it feeds the media as well as the public who likes to bet on who is to win place or draw.
so, if your theory is right, and the sample is too small relative to the general population to be reliable, then the numbers for Trump might even be higher? Right?