Interactive dashboards modelling six scenarios across 64 English councils, published alongside Towards Data Science articles.
Election coverage tends to report point predictions with false confidence. The interesting question is not who wins, it is how much of the outcome is driven by assumptions rather than signal.
We modelled six scenarios (S0 to S5) across 64 active English authorities and showed the calibrated uncertainty band around each stated assumption, rather than a single forecast. The honest finding: even the strongest shock we modelled moves the result by only about 13 percent of the median uncertainty band.
Two interactive dashboards with a multi-panel navigator, published alongside companion articles on Towards Data Science and built on official ONS and government election data. The model is locked and the source files are hashed, so anyone can reproduce the results.
A public, reproducible piece of civic data analysis released ahead of the May 2026 UK local elections. Explore the live dashboards above.
We scope the problem properly first, then build production systems that companies rely on. Book a call and tell us what you need.