I get it when Nate Silver and his ilk suggest that even with odds of 10 out of 100, we can’t say Trump will definitely lose. Silver even suggests that he won’t say that even with odds of 5 out of 100. In that case, I don’t see the utility of these models. Just average the polls!
@pratik The entire concept of probability becomes confusing when you're talking about unique events.
@MitchWagner @bix I’m a data nerd but yes, this doesn’t seem suited for predicting election outcomes that occur once every 4 years especially in polarizing times. Sports predictions to a lesser extent suffer from these same issues.
@pratik Nate Silver gets a bum rap for failing to predict the 2016 election outcome, but I remember even as he was making his predictions, prior to the election, he was cautioning that yes, his models gave Trump 1 chance in 7 of winning the election – but things with only a 1:7 probability of happening happen with great frequency. @bix
@MitchWagner I'm actually fine with his polls and polls+ average. I think a lot of mission creep has happened with his prediction model. The growing gap between electoral college and popular vote hasn't helped either.
@cambridgeport90 Someone once likened it to parliamentary democracy where each state is a constitutency. I was like, ok, but that makes no sense especially when you have 435 Reps in the House
@pratik it's times like these where it might be good to go back to our roots. Only the people. Might make more sense.
@cambridgeport90 I'm not surprised. Most immigrants understand the USA better than Americans. Perhaps because we actually have to study it become a US citizen. I emigrated from a country with a parlimentary system. The current Prime Minister has been in power since 2001 and is the he longest continuously serving head of government since the country became independent from Great Britain in 1979.