In mid-May of this year, Florida politicos on both sides of the aisle turned their attention towards Duval County, where the race for Jacksonville Mayor was coming down to the wire. Despite the fact that Democrats made up 4.3% more of the registered voter base than Republicans in Jacksonville, Republican turnout exceeded Democratic turnout by 3.2% by the end of Election Day. With turnout numbers tilted in their favor, Republicans were hopeful that their nominee, Daniel Davis, would come out victorious. That didn’t happen. Instead, Democrat Donna Deegan won with 52.1% of the vote, besting Davis’s 47.9%. Despite strong Republican turnout – bolstered by the Florida GOP’s get-out-the-vote efforts – the Republican candidate had lost, and many on the right were left scratching their heads and asking, “What went wrong?” In this article, we’ll be answering that question through the lens of the Victory Formula, a model developed by our team for the purpose of understanding general election outcomes. If you’d like to take a deeper dive into this formula, you can read this article explaining how it works. But for our purposes today, all you need to know is that the formula takes in nine variables (three about voter registration, three about turnout, and three about support for each candidate), and produces the exact margin by which a Republican candidate wins (or loses) a head-to-head general election. Since the election in Jacksonville has already happened, we were lucky enough to have exact final numbers for voter registration and turnout, meaning all that’s left for us to understand is the layer we refer to as “support.” This layer has three simple variables: Republican loyalty, Democratic loyalty, and nonpartisan loyalty. Republican loyalty refers to the percentage of Republican turnout that resulted in votes for the Republican candidate. For instance, if 100 Republicans voted in the election, but only 75 of them voted for the Republican candidate, then Republican loyalty would be 75%. Democratic loyalty refers to the same concept, just on the Democratic side. And finally, nonpartisan loyalty simply refers to the percentage of the nonpartisan/third-party that went to the Republican candidate. None of these variables can be found in the election result data. Instead, we must use the Victory Formula to make inferences about their values, and from there, draw conclusions about why the race ended up the way it did. But there’s an issue: with three unknown variables, there are many different combinations that could result in Deegan’s 4.2-point victory. In fact, these potential “solutions” can be mapped out on a three-dimensional surface: