Negeri Sembilan 2026 forecast: BN+PN win in every simulation

Barisan Nasional and Perikatan Nasional, standing as one ticket, win the Negeri Sembilan election in every one of our 10,000 simulations. The central projection gives the alliance 26 of the 36 seats against 10 for Pakatan Harapan, with a 90 percent range of 23 to 30. Bersatu, contesting alone for the first time, wins nothing in any scenario we can defend with data.

The certainty at the top hides the honest part underneath. Which 23 to 30, and which of PH's seats survive, turns on a number nobody has measured in this state for years. That is what the rest of this piece is about.

How the forecast works

There is no public polling for this election, so the simulation is built from measured results instead. We drew 10,000 scenarios for how each community votes, with the ranges anchored to real elections: the state's own 2022 parliamentary vote, where essentially the whole non-Malay electorate sat with PH, and Johor three weeks ago, where BN clawed back roughly one Chinese vote in seven and a bigger slice of the Indian vote. Each scenario also varies how much of its old vote Bersatu keeps, how far urban seats tilt, and adds seat-level randomness sized on the surprises Johor actually produced. Every input comes from ElectionData.MY's open archive.

Which seats are actually close?

Twenty-four of the 36 seats fall the same way in at least 99 of every 100 runs. The whole Malay belt is locked for the alliance; Lobak and Mambau are locked for PH. The election lives in the middle of the table below. Three are true coin flips: Repah at 44 percent for PH, Nilai at 35, and Sri Tanjung at 61, the seat where four Indian candidates are chasing the same 27 percent Indian roll. The bottom two, Chennah and Rantau, look safe on the numbers but we flag them for the star incumbents the model cannot see.

Those bottom two rows deserve a health warning. The simulation knows nothing about individual candidates except in Rantau, where we handed the alliance eight extra points for Tok Mat because his 2023 result beat every demographic expectation by that much, which is why it now reads 0 percent for PH. Chennah's 4 percent is blind to the fact that the candidate is Anthony Loke, the DAP secretary-general, in the seat he won with 61 percent in 2023. A personal pull the size of Tok Mat's would drag Chennah back into a real contest, and we would not bet against him having one.

Why does Bersatu win nothing?

The arithmetic is unforgiving. To win its best seat, Bersatu must keep at least 78 percent of the vote PN polled in 2023. In Johor three weeks ago, the side that walked away from a consolidation kept 27 percent. Even when our simulation lets Bersatu retain 45 percent, well above anything observed, its seat count stays at zero across all 10,000 runs. Its real effect on this election is different: in a handful of Seremban-belt seats, every Malay vote Bersatu takes comes out of the alliance's column, and that quietly helps PH hold its firewall.

What could make this wrong?

The single widest input is the Indian vote. Negeri Sembilan has the peninsula's highest Indian voter share and nobody has measured it standing alone since 2018, when about a quarter of it was still with BN. Our scenarios let PH's share of it run anywhere from 25 to 95 percent, and that one dial moves seats like Jeram Padang, Sri Tanjung and Lukut by itself. Candidate effects beyond Rantau are ignored, turnout is assumed to look like the past, and there is no polling to correct any of it. If the forecast breaks, the first place it will show is Jeram Padang, a third Indian by roll, where an Indian PN candidate already came within seven points in 2023.

On the night of 1 August, watch three things. Repah, Nilai and Sri Tanjung tell you who is winning the coin flips. Jeram Padang and Chennah tell you whether Johor's swing travelled north. And the five Seremban-belt seats, from Rahang to Lukut, are PH's firewall: if any of them falls, the alliance is heading past 30. We will grade this forecast in public after the result, the same way we graded our Johor predictions.