People will say you need to be cautious about election polls. They don’t always divine the results of these events properly, and if you give your heart to one because it complements your political hopes, you run the risk of looking foolish. What could be worse than that?
The example you will invariably hear in any discussion about polling is the 2016 race for president. Hillary Clinton was leading throughout the contest, so critics say the polls were inaccurate because Donald Trump won, despite his many shortcomings.
So was the 2016 presidential sampling wrong? Most horse-race polls measure popular vote support for candidates. That year the data agglomerator FiveThirtyEight.com averaged all those available. Their election eve prediction had Clinton besting Trump 48.5% to 44.9%.
The final election numbers showed Hillary defeating Donald 48.2% to 46.1%. Similar results, right? Of course, Trump somehow won the Electoral College vote and went on to…