Battlemap for Power Users

For sophisticated users, the Battlemap accuracy can be improved to take into account minor parties.

The normal assumptions built into the Battlemap are:

This framework works well when the minor parties stay at 8.0%, which was their level of support at the May 2005 General Election. However, their support can go up and down. This makes the Battlemap less accurate.

Sophisticated users can allow for this effect when using the Battlemap, by adjusting support levels for Conservative and Labour before looking up the Battlemap. The adjusted levels should be:

AdjCON% = CON% * 92% / (CON% + LAB% + LIB%)
AdjLAB% = LAB% * 92% / (CON% + LAB% + LIB%)

Example

As an example, suppose the support levels are CON 32%, LAB 34% and LIB 21%. The true predicted seats would be CON 214, LAB 345, and LIB 55. But the Battlemap, used naively, would suggest around CON 205, LAB 340, and LIB 70 seats. This is particularly inaccurate for the Liberal Democrats, as we are over estimating their support.

If we apply the adjustment we get

AdjCON% = 32% * 92% / 87% = 33.8%
AdjLAB% = 34% * 92% / 87% = 36.0%

Using these adjusted levels in the Battlemap gives a prediction of CON 213, LAB 348, and LIB 54, which is much more accurate.


Back to the Battlemap or return to Election home page.