Booth-Wise Past Results: The Dataset That Shows Where Fights Are Won

Ask most campaigns about their constituency's last result and they can quote the margin. Ask them which forty booths decided that margin and the room goes quiet. Booth-wise past results are the constituency's history at the only resolution that matters — the polling-station level where every future contact plan, route and resource decision is made. The data is public, free and persistently under-read, which makes it the highest-return analysis work a campaign does before the first door is knocked.

The work is part arithmetic, part hygiene. Returns are assembled at booth level across the last three to four elections, cleaned against the current electoral roll — booths split, merge and renumber between cycles, and an unparsed renumbering silently corrupts every trend. Candidate votes are normalised against electors to give share, not just volume, and turnout is attached to each line because a booth's abstention rate is as strategic as its preference. Done properly, this is a fortnight of careful desk work producing the seat's single most reusable asset.

Then the reading begins, and it is where the value concentrates. Booths sort into families: the loyal bases, the swing clusters, the low-turnout strongholds of the opposition that a turnout strategy can suppress or convert, the booths where one candidate's personal vote outran their party's. Each family gets a different plan — defend, persuade, mobilise, ignore — and the map's shape dictates the whole ground operation: where coordinators are thick, where events go, where the budget for last-mile contact concentrates.

The cross-reads multiply the insight. Results against the demographics map show which communities moved and why. Results against infrastructure — a new road, a college — separate development's real electoral yield from its assumed yield. Shifts between consecutive elections mark the seats' rising and falling forces more honestly than any punditry. This is the base layer the socio-economic profile sits on: history first, structure second, strategy third.

The discipline extends through the campaign. The war room's tracking sheets keep the same booth keys, so field data — door survey rounds, contact rates, early-vote intelligence — aggregates into comparable trend lines. By the final fortnight the campaign is reading its own booth-wise picture, not the state's mood. Elections are decided in three-hundred-part arithmetic, and the campaign that parsed its parts early is the one that spends the endgame where the seats are actually won.