Methodology

This forecast is built using a constituency-level statistical model designed to estimate the current electoral landscape of the UK. The model begins with the official results of the 2024 General Election for every parliamentary constituency in Britain. These constituency-level results provide the baseline from which all projections are generated. 

​

To estimate how voting intention has evolved since the election, the model incorporates publicly available polling from multiple organisations. Data is aggregated across a range of pollsters to reduce the impact of sampling variation and short-term fluctuations in public opinion. To ensure the forecast remains responsive to the latest political developments, polling is weighted by recency. Data sources include: YouGov, Opinium, Find Out Now, More in Common, BMG Research, Lord Ashcroft Polls, Norstat, Survation, and Ipsos.

​

Polling averages are used to calculate changes in public support relative to the 2024 election. These vote-share changes are then applied to constituency-level election results to estimate how each seat would perform under current political conditions.

 

 Recognising the distinct political dynamics of the UK’s constituent nations, the model has separate polling frameworks for Scotland and Wales. Constituencies in these nations are therefore influenced by Scotland and Wales specific polling trends. The model also applies weighted regional adjustments to English constituencies to reflect geographical differences in party support.

​

Where a constituency has held a parliamentary by-election since the 2024 General Election, the model adds the by-election result into the constituency baseline using a weighted approach. Individual constituency projections are subsequently aggregated to estimate seat totals across the House of Commons and identify the likely parliamentary balance. Forecasts are updated regularly as new polling data becomes available.