- Oliver Munoz
Polivision brings data visualisation to one of the most fundamental aspects of U.S. political science – the electoral college map. To win a presidential election, a candidate must win 270 ‘electoral votes’. With each individual state worth a defined number of electoral votes, candidates win the election by winning enough states, worth enough electoral votes, to reach 270.
There are existing web tools that allow users to build a path to 270 electoral votes by manually ‘flipping’ states on the electoral map from one party to another. With clean and beautiful UI, Polivision takes this idea further by embedding previous election result data, and exit poll data, into the image of the map. This enables users to interact with the map both on mobile and desktop devices, powerfully and realistically, with intuitive sliders to manipulate key voting behaviours such as swings, and increases or decreases in turnout.
Polivision enables users to pursue two key questions: Firstly, what would it take for outcome X to be achieved? Secondly, what is the effect of input Y? The inclusion of exit poll data allows users to ask these questions when focussing on specific demographics (for example, what would be the effect if 5.5% of men switched their vote?). We want users to mix the data around and visualise what would happen if the didn’t (or did) vote; this is huge to understand the importance of voting.
Data visualitation is very complex to simplify for the average user. Similar tools out there are trying to replicate what POLIVISION has been doing from day one, but with so many layers of complexity in a tool like this, they fall short. Though POLIVISION’s target audience is politics aficionados, anyone can pick up the tool and quickly understand how to use it. To make a great user experience, the team had to deeply understand how politics work, but more importantly how the users think and how they might use this tool to predict the elections.
POLIVISION’s design is fully user-centred, addresses some of the major pain points including simplified data visualisation and manipulation of complex layers of information to come up with a clear result: Who could win the 2020 US elections. The platform was also built and thought in a way that can be scaled up and implemented to other elections (mid-term elections) and it’s not limited to the US, can also be modified to other countries (Australia).
Data as at 12 September | Launched 28 August 2020. Sessions: 5,114 / Pageviews: 10,249 / Sessions by top 3 locations: USA (58.51%), Australia (22.49%), Canada (6.94%)