Our user-defined prediction tool allows users to manually enter tactical voting parameters, letting you control the extent of tactical voting for your prediction. This feature works on the basis of Electoral Calculus' tactical voting model.
The model's parameters are based on a polling we have conducted, usually prior to each general election. We asked voters how they would vote supposing that their seat was a close contest between various pairs of parties. These A/B tests tell us how voters might react to various electoral possibilities.
If the seat has three (or more) likely contenders, then tactical voting will not be applied.
Full details of the polling and questions asked are found here.
Suppose a seat is a close contest between Labour and Reform. We asked voters of other parties how they would vote in this situation. Their answers are given in the matrix below:
| First Pref | CON | LAB | LIB | Reform | Green | YP | OTH |
|---|---|---|---|---|---|---|---|
| CON | 37% | 11% | - | 52% | - | - | 1% |
| LIB | - | 65% | 26% | 9% | - | - | 0% |
| Green | - | 66% | - | 3% | 30% | - | 0% |
| YP | - | 59% | - | 2% | - | 39% | - |
| OTH | - | 14% | - | 56% | - | - | 30% |
For Conservative supporters, the first row of the table says that 37pc of them would stick with the Conservatives and not vote tactically. Eleven per cent would vote tactically for Labour to keep Reform out, and 52pc would vote Reform to keep Labour out. A residual 1pc would vote for other parties.
For Liberal Democrat supporters, the second row of the table says 65pc would tactically vote Labour, 26pc would stay with the Lib Dems and 9pc would vote for Reform.
The remaining rows describe how supporters of the Green Party, Your Party and other parties respectively would behave in this situation.
We assume that Labour and Reform supporters would vote for their first preference party in this case.
Therefore, in seats Labour-Reform seats, our model will allocate 11pc of Conservative support to Labour, and 52pc to Reform.
Each party's supporters will engage in tactical voting to different levels, and in different directions. In our 2025, the two main types of tactical voting were partisan (left-wingers vote for other left-wing parties, and right-wingers for other right-wing parties) and also anti-Reform tactical voting.
Tactical voting patterns in Wales and Scotland, and at other elections (see 2024), are different.
If you enter 100% for a given party, our tactical voting model will be applied in full effect. This means that in, say, Labour-Reform seats the full amount of tactical voting that our model predicts will be applied to the other parties.
If you enter 50% for a party, the model will be applied to 50% effect. This means that the tactical voting model will still be applied, but with 50% less tactical voting. In numerical terms, Conservative supporters' tactical voting under these conditions would look like this:
If you enter 0% for a party, it assumes no tactical voting, so none of their supporters' votes will be allocated to other parties.
Tactical voting input parameters apply to each party individually. This allows you to set different levels of tactical voting for different parties. For example, if you think that lots of Labour, Liberal Democrat and Green supporters will vote tactically, but Conservative and Reform supporters won't vote tactically, you could enter something like the following:
| CON | LAB | LIB | Reform | Green | |
|---|---|---|---|---|---|
| Tactical Fraction | % | % | % | % | % |
Voters who vote tactically do not all pick the same destination for their tactical vote. It depends on which parties are competitive in the seat.
The trends of tactical preferences measured in 2025 are estimated as follows: