Election Night Analytics: Turnout, Vote Shares, Swing, Seat Projections & Disproportionality

What you'll build

What you will be able to do

Each lab ends with a concrete result that you build yourself and the checker confirms, cell by cell.

• Turn the raw count sheets of 12 districts into valid votes, turnout per district, a turnout ranking and the overall turnout of the whole region – without the classic average-of-percentages mistake.
• For each district you find the winner and the runner-up with INDEX/MATCH, the winner's vote share and the majority as a share of the valid votes.
• Compare this election with the last one: each party's share now, its change in percentage points, the two-party swing between Lantern and Harbour in every district and for the whole region.
• Before the first result arrives you build the desk's target list: the margin in every seat, the swing that would flip it, a marginality ranking and how many seats each party loses at a test swing.
• Apply the national change from a fictional poll to every district, find the projected winner of each seat and turn the winners into projected seat totals and net gains.
• At 23:40 seven of twelve districts have reported. You measure how much of the electorate is in, show why the raw share so far is misleading, and project Lantern's final share from the change in the reported districts.
• Compare each party's share of the votes with its share of the 12 seats, express the gap in points and summarise the whole result in one disproportionality index.
• Finish the night's dashboard: seats per party, the gains of each party, the district with the biggest swing and a seat chart for the front page.