Put every party on a single left–right line, weight each one by how many people say they'll vote for it, and you get one number for where the electorate's mass sits. Right now that number is +1.85 — right of centre, and moving further out.
Where a circle sits along the line is the party's political position. Negative is left, positive is right, zero is dead centre. Labour at −2 is mildly left; Reform at +8.5 is a long way right.
The area of each circle is how much of the vote that party is polling. Big circle, lots of voters at that position. Restore Britain is far right but small; Labour is nearer the middle but carries roughly seven times the weight.
It marks the balance point of the whole electorate — each party's position pulled by its size. It is not a party and nobody has to be standing there. It just says which way the country's weight leans.
The strip under the axis translates numbers into words: anything from −1.5 to +1.5 counts as centre, and the further out you go the stronger the label. Boundaries are conventions, not facts.
| Party | Score | Band | Polling | 2024 | Pull |
|---|---|---|---|---|---|
| Polling is a composite of the PolitPro trend, the PollCheck seven-poll average and the BritPolls five-poll average, all as of 6 August 2026, with Restore Britain taken from the YouGov and PollCheck trackers. Pull is score × share — how hard that party tugs the average. | |||||
The three figures above answer different questions, and it is easy to read one as though it were another. Here is what each is actually measuring.
Imagine the axis as a plank and each party as a weight placed at its position, heavier the more votes it has. The centre of gravity is where you would put the fulcrum to balance it.
A number near zero means the country's weight is evenly matched either side of centre. Positive means more weight sits to the right — either because more people vote right, or because the right-hand parties sit further out. Both count.
How far the balance point has travelled since the last general election, on the same axis with the same placements. Positive is rightward.
For scale: 0.64 is roughly a fifth of the distance between Labour and the Liberal Democrats. Real, but not a landslide of opinion — most of it comes from Reform and Restore Britain growing rather than voters individually changing their minds.
The average distance between a voter and the balance point — a measure of how polarised the country is, not which way it leans.
A low number means most voters cluster near one another. A high number means the electorate has pulled apart into camps far from each other. Britain's has risen sharply since 2024, which is a different and arguably bigger story than the rightward drift.
An average is not a majority. Line every voter up along the axis and the person standing at the halfway point — the median voter — sits at −1, in Liberal Democrat territory, left of centre. The average sits at +1.85, well to their right.
The gap exists because an average is sensitive to distance and a midpoint is not. The left of the axis is crowded but close to the middle: Labour at −2, the Lib Dems at −1, the nationalists at −4 and −5. The right is three parties, and two of them sit at +8.5 and +9.5 carrying more than a quarter of the vote between them. One party a long way out pulls harder than three standing near the centre.
Which is how the left-of-centre parties can hold 51.5% of the vote against 46.7% for the right, and the balance point still lands on the right-hand side of zero.
Against the 2024 general election the balance point has shifted right by 0.64 points. Real, but modest beside the churn underneath it. Labour has shed around ten points since 2024, and only part of that went rightward; a comparable share went to the Greens, in the opposite direction.
Those two flows partly cancel in the average and compound in the spread. Spread is up from 4.77 to 5.56, a 17% rise, and Restore Britain's arrival at +9.5 accounts for a good part of it. The electorate has not so much travelled rightward as pulled apart, hollowing out the ground between −2 and +5 where British elections used to be decided.
Each party gets a position on a −10 to +10 axis. The centre of gravity is the average of those positions weighted by polling share: multiply each score by its share, add them up, divide by the total placed share. Spread is the weighted standard deviation around that average. Roughly 2% of the vote sits with parties too small or too varied to place and is excluded from both.
The placements are the contestable part, and they are mine. One axis has to hold economics, immigration, social liberalism and the constitution all at once, which it cannot really do. Reasonable people would put the Lib Dems on either side of zero, would argue whether the Conservatives belong at +5.5 or nearer +7, and would dispute how much daylight there is between Reform at +8.5 and Restore Britain at +9.5 — the two agree on most things and differ mainly on how far to push them. Move Reform to +7 and the centre of gravity falls to about +1.49, cutting the measured shift since 2024 by more than half. The ranking is sturdier than the arithmetic; treat the second decimal place as decoration.
Polling is a composite of three published aggregates rather than any single one, because they currently disagree by more than usual — during the first week of August the trackers had Labour and Reform between one and two points apart in either direction. Restore Britain, registered as a party in February 2026, has polled between 3% and 4% through the summer. The next general election is not due until 2029.
A centre of gravity says nothing about who governs. First past the post decides that separately, and rarely in proportion.
Open the UK simulator