It has been over a month since the 2018 midterm elections, and even though the final seat tally in the House is still undecided (thanks to possible election fraud in North Carolina district 9), it seems appropriate to reflect on the model I built for CNN, including successes, mistakes, and lessons learned. The Model To refresh on the details of how the model works, I recommend going to my model talk blog post, the methodology…
Election 2018 Battlegrounds …another Q&A with Harry Enten
Election Day is approaching! It is always interesting to look at the trends leading up to an election, like which demographics or geographic areas seem to be the most important. So, with just a few days left in this election cycle, let’s take some guesses about what we will be talking about post-election. This is an extension of the Q&A published by CNN on 30 October. As of this writing, our model has 22 toss-up…
Model Talk: Q&A with Harry Enten
Over the last few months, I have had the opportunity to work with Harry Enten at CNN (formerly Fivethirtyeight) to create “The Forecast” — CNN’s 2018 midterm election model. It has been an awesome experience, and I intend to write a longer reflection after the elections. For now, I would like to share some of the finer details about the model that we created, since I know that there are some people out there who,…