Gallagher Index: This formula is used to measure the 'disproportionality' of the result, which is how far the percentage of seats each party wins is from the percentage of votes won by each party nationally. It is the square root of half of the sum of each party's vote share minus seat share.

D'Hondt: In this system, which was used for the 2026 election, seats are awarded to parties based on a quotent. In each round the party of the highest quotent wins, and the quotent is simply the number of votes divided by the number of seats won +1.

Sainte-Laguë: In this system, seats are also awarded to parties based on a quotent. In each round the party of the highest quotent wins, and the quotent is simply the number of votes divided by the number of seats won x 2, +1. The benefit compared to D'Hondt is a more proportional outcome which doesn't favour large parties. For more, see Dylan Difford's blog on the ERS.

Pure PR National: This is a base model to compare to where assuming no constituencies, parties are simply allocated seats equal to votes cast (rounded to 96).

AMS Reformed: In this system, a modified version of the old Additional Member System is used, whereby each voter has two ballots, one for a constituency (which for simplicity's sake is granted four FPTP members), and one for regions (which are granted eight members apportioned via D'Hondt). The major issue with calculating this is that voters are likely to vote differently in regional versus constituency ballots, which may be utilised to the advantage of two like-minded political parties, such as the SNP and Greens in Scotland this year. For this calculation, votes were assumed to be the same in each ballot. While the calculation is crude, it is a rough approximation of what the result could have looked like were the recent electoral reform not introduced.

STV - Hare and Droop: Voters are granted the ability to rank all candidates and parties. For this calculation we assume the same 16 constituencies are used with six members each, and a table below is used to apportion votes from one party to another. This is based on the Ipsos polling done 2nd-8th April 2026 with a sample size of 747, in which voters of all parties are asked "Regardless of how you currently intend to vote [...] how likely, if at all, would you be to consider voting for each of the following parties [...]?" While more accurate data on second and third preferences for voters would be beneficial to the accuracy of this calculation, local factors often play a large role in STV, which cannot be accurately modelled in this way. The figures used get us a lot of the way towards how an STV election would likely go in practice, and is similar to the results produced in Scottish councils.
The difference between Hare and Droop are the quotas used to apportion seats, Hare is a higher quota and more proportional, but often leaves the final seat(s) to the "largest remainder" of the smaller parties, but the Droop quota is smaller and favours larger parties. For more info on STV see this page by the ERS and for Hare v Droop quotas this blog by Dylan Difford.