By "oversampling" democrats and/or various minority groups, pollster after pollster kept rolling out predictions that seemed utterly ridiculous to us but were gobbled up by complicit media outlets.  Here is just a small sample of our headlines from the past couple of months:

"ABC/Wapo Effectively Admit To Poll Tampering As Hillary's "Lead" Shrinks To 2-Points"

"How Reuters "Tweaked" Its Latest Poll (Again) To Show A Clinton Lead"

"Statisticians Warn Of "Systemic Mainstream Misinformation" In Poll Data"

"CNN Goes "Full Reuters" - Attempts To Rig VP Debate Poll With Too Many Dems In Sample"

"New Podesta Email Exposes Playbook For Rigging Polls Through 'Oversamples'"

Now, with the official results in, we can finally confirm that pollsters were either intentionally manipulating their data or were simply clueless about how to structure their sample pools.  Per Bloomberg, going into election night, only two polls, IBD/TIPP and USC Dornsife/LA Times, showed victories for Trump while all of the other major polling outlets showed a 3-4 point national lead for Hillary.  And, while Hillary may still scrape up a modest popular vote edge it will be no where near the 3-4% margin predicted by these polls.