The Lion's Pride Volume 11 (Winter 2019) | Page 35
demographic information, voting history—was unstructured” (Patterson,
2016). As Wasserman noted, the data was already there; someone just
needed to collate it. This is where the math comes in. Every campaign
nowadays has computer models. These models are used to predict the
outcomes of elections based on inputting data which the campaign has
collected or purchased. While they are called models, they are, in a
sense, just long algorithms. Most current models are built like static
machines which take in data, then run math programs to compare said
data against large amounts of controlling variables. Campaign models
use data like location or if the person voted in the last election to help
narrow down targets for advertising (or more direct methods) to
influence the vote (Patterson, 2018). This tactic will be even more
valuable when AI begins to be incorporated into the models. AI, or
artificial intelligence, will allow people to create programs which, in
turn, start making contextual connections on its own, with little human
intervention. The current way data is used requires a human to be
standing by cranking the dials till the expected outcome pops out. AI
will lessen the reliance on humans to ensure a model's accuracy.
FiveThirtyEight is a leader in the news plus data analysis market.
They are owned by ABC news and use data imputed into math models to
predict outcomes of a wide variety of events. They release many of their
models on the code sharing website GitHub. Their models are based on
building data sets like a CSV and running a series of algorithms which