I recommend taking some time to read what they publish rather than what you may have heard they publish. 538 consistently pushes for proper interpretation of st

I recommend taking some time to read what they publish rather than what you may have heard they publish. 538 consistently pushes for proper interpretation of stats and predictions—particularly uncertainty percentages, which is often not the case in other publications, even those that reproduce stats sourced from 538. In particular, the primary numbers 538 published for the 2016 election were chances of winning, which can easily be misinterpreted as percentage of the vote. As of Nov 8, they had Clinton having a 71% chance of winning (with Trump at 27%), with 302 electoral votes, and 48.5% of the popular vote. This isn't a landslide prediction, and the 71% chance of winning is much lower than what was reported in many other publications. https://projects.fivethirtyeight.com/2016-election-forecast/ https://projects.fivethirtyeight.com/2016-election-forecast/ In the submitted article, they're quite careful in their language in support of the title, which purposefully includes the operative word "may". This is in keeping with what I've come to expect from 538.