High school class's electoral predictions model is a model for electoral predictions


Originally published at: https://boingboing.net/2018/10/24/stats-cards-on-the-table.html


Back in the day, a friend of mine was on the rifle team at Monkey Blair.


The open-source nature of this is really interesting. As someone who abandoned 538 & Nate Silver completely after 2016, can anyone tell me if he’s made any significant changes to his models?


Wow, they ran their simulation 10,000,00010,000,000 times? That’s more than a jillion!


Yes and no. The 538 model has undergone a lot of changes, mostly around disseminating statistics to the general audience. That said, 538 still seems very naive when it comes to voter suppression tactics, vote hacking (ie. redshift), etc…


Statistics. Probability. And… random selection… by a chicken playing… “bingo.”


Was 538 actually wrong in 2016, though? I seem to remember Trump having something like a 20% chance of winning, toward the end of it. One out of five isn’t impossible, people. Things that are only 20% likely happen ALL. THE. TIME. I think our biggest issue is that people don’t understand statistics and probabilities all that well, as they think 80% is the same as 100%. They are not the same values.


Not at all, 538 was actually pretty damn accurate. Remember, Clinton won by 3m votes.


This topic was automatically closed after 5 days. New replies are no longer allowed.