Because the accumulated weight of the polls is to the contrary. I believe in the averages to be more reliable even considering the history of the last election.
But let us assume that I am wrong, that the polls you cite which show a virtual tie or Trump slightly behind are accurate, the situation even so it is not much improved for Trump because of his daunting task of overcoming the Democrat advantage in the electoral college. Trump must virtually run the table of the battleground states, if he stumbles in the wrong place, or stumbles twice virtually anywhere, he loses. With the polls you cite, Trump runs a very high risk of failing to put together 270.
A loss in Florida would be fatal, a loss in North Carolina nearly fatal. Do you like his numbers in those states? Do you like his numbers in the other battleground states? Are you sure he will stumble or do you believe the odds are that he will somewhere?
Because the accumulated weight of the polls is to the contrary. I believe in the averages to be more reliable even considering the history of the last election.I believe that the polls with a track record of being accurate are the ones that are reliable.
IBD is the best. That is the history. Check the sidebar. It shows that "Trump Leads Clinton By 1 Point Going Into Debate -- IBD/TIPP Poll".
Rasmussen also has a pretty good track record. Check the side bar again. It shows, "Rasmussen: White House Watch Trump up 43-40".
The poll I'm less familiar with, the USC Dornsife/LA Times Poll hows Trump +0.6.
What has a long history of not being accurate is the RCP conglomeration that contains numerous push polls.
even so it is not much improved for Trump because of his daunting task of overcoming the Democrat advantage in the electoral college.
The latest round of state polls has Trump winning in OH, FL, NV and IA. Beyond that, if the popular vote says Trump will win, then mathematically it is likely that the EC totals will following.