Welcome to DU!The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards.Join the community:Create a free accountSupport DU (and get rid of ads!):Become a Star Member
Furthermore, 'approval' polls are inherently different (and more likely accurate) from predicting election results, because you are just measuring 'how someone answered' as opposed to relying on people 'following through with something' ... such as 'actually voting' rather than just 'being likely to vote'.
Final 538 odds of Hillary winning the morning of the 2016 election was about 7-3 odds. That absolutely does mean their polling was inaccurate because she didn't win. Things with 30% chance of happening ... obviously can happen.
This is especially so given that the election was decided basically on a mere 80K votes over 3 states.