How could the mainstream media and their pollsters once again get the U.S. election results so wrong?

The election wasn’t supposed to turn out this way, according to months of American election polling. Survey after survey and breathless media commentary predicted that Joe Biden would beat Donald Trump in popular vote by upwards of 6, 8, even 10 percent. The same polls and media commentators projected the Democrats with winning margins in the key battleground states of Florida, South Carolina, North Carolina and Idaho. Bullish predictions included a Democratic sweep of Senate, more Democratic seats in the House of Representatives, and Biden flipping the traditional Republican strongholds of Georgia and Texas. So far, none of these predictions came to pass and, instead, a Democratic cakewalk into the White House has morphed into a contested election with possibly weeks, if not months, to go before a winner is officially declared. How did this happen? Why, after 2016, is much of the media, and seemingly the majority of pollsters, so clueless when it comes to fathoming voter’s intentions on election day? Is it time, once and for all, to give up on public opinion polling as predictive tool? What is the effect on democracies of faulty polling and a media only too happy to widely publicize survey results that 2020 would suggest have little real bearing on what voters actually think?