Mt Roskill general electorate
Michael Philip WOOD (Labour) holds on, projected to win by 409 votes (1.12 pp) over Carlos CHEUNG (National).
Why. In the preliminary count Michael Philip WOOD leads by 262. About 6,416 electorate-vote specials are projected here. From 2023, Michael Philip WOOD’s multiplier is 0.945 and Carlos CHEUNG’s is 0.913. Specials are shared in proportion to preliminary share × multiplier, so Michael Philip WOOD gets 2,781 and Carlos CHEUNG 2,634: a net +147 for Michael Philip WOOD.
To tie, Carlos CHEUNG would have needed a multiplier of 1.060 (with everyone else’s unchanged); they got 0.913, which is 1.16× short.
Margin
1. Notional 2023 Mt Roskill
The 2026 Mt Roskill electorate is on 2025 boundaries. Its 2023 numbers are built from the 2023 electorates that overlap it. Each 2023 electorate’s votes are multiplied by its weight: the share of its 2023 Census electoral population that falls inside the 2026 Mt Roskill, worked out meshblock by meshblock. This assumes each 2023 electorate’s votes were spread evenly across its population. (% of 2026 Mt Roskill is the other way round: how much of the 2026 Mt Roskill’s population came from that 2023 electorate.) The party vote is needed too, for the step 2 fallback and the step 3 scale factor.
| 2023 electorate | % of 2026 Mt Roskill | Weight | Preliminary | Final | Weighted prelim | Weighted final | Sources |
|---|---|---|---|---|---|---|---|
| Mt Roskill | 72.8% | 77.9% | 26,247 | 35,312 | 20,448 | 27,510 | prelim · official · XML |
| New Lynn | 27.2% | 27.8% | 30,834 | 39,800 | 8,575 | 11,068 | prelim · official · XML |
| Notional Mt Roskill | 29,023 | 38,579 | |||||
Valid electorate votes. Each party’s candidate votes are weighted the same way. The 2023 preliminary pages are Wayback Machine captures.
2. Each party’s 2023 multiplier (electorate vote)
Special votes = final − preliminary. Multiplier = share of the specials ÷ share of the preliminary count. Above 1, special voters here liked them more than election-night voters did.
| 2023 candidate | Preliminary | Final | Specials | Prelim share | Special share | Multiplier |
|---|---|---|---|---|---|---|
| National Carlos CHEUNG, Paulo GARCIA | 12,777 | 16,617 | +3,840 | 44.02% | 40.19% | 0.913 |
| Labour Michael Philip WOOD, Deborah RUSSELL | 11,530 | 15,117 | +3,588 | 39.73% | 37.54% | 0.945 |
| Green Suveen SANIS WALGAMPOLA, Steve ABEL | 2,700 | 4,128 | +1,427 | 9.30% | 14.94% | 1.606 |
| ACT Rahul CHOPRA, Juan ALVAREZ DE LUGO | 1,739 | 2,337 | +598 | 5.99% | 6.26% | 1.045 |
| NZ Loyal Steve OLIVER | 232 | 319 | +87 | 0.80% | 0.91% | 1.144 |
| Freedoms NZ Phineas MANN | 45 | 60 | +15 | 0.16% | 0.16% | 1.012 |
| Total | 29,023 | 38,579 | +9,556 | |||
Candidates are matched to 2026 by party. Where several names are shown, the area took in parts of several 2023 electorates; names in lighter text are from electorates that make up less than 20% of the 2026 Mt Roskill’s population.
Party-vote multipliers, for candidates with no 2023 counterpart
Opportunity, Animal Justice had no 2023 candidate here, so their candidates use the party’s party-vote multiplier in this area, worked out the same way from the notional 2023 party vote:
| 2023 party (party vote) | Preliminary | Final | Specials | Prelim share | Special share | Multiplier |
|---|---|---|---|---|---|---|
| Opportunity | 558 | 847 | +289 | 1.90% | 2.78% | 1.468 |
| Animal Justice | 41 | 57 | +17 | 0.14% | 0.16% | 1.183 |
| All parties (party vote) | 29,444 | 39,809 | +10,364 | |||
A party new in 2026 has no 2023 party vote either, so it gets 1. A party that was part of an umbrella party in 2023 uses the umbrella’s multiplier.
3. How many special votes
In 2023, there were 9,556 specials here on 29,023 preliminary votes, a ratio of 32.9%. Every electorate’s ratio is multiplied by the same national factor, 0.646, so that the electorates’ party-vote specials add up to the national projection (fewer specials are expected in 2026). So: 30,150 preliminary × 32.9% × 0.646 = 6,416 projected electorate-vote specials.
4. Sharing out the 2026 specials
Each candidate’s weight is their 2026 preliminary share × their party’s 2023 multiplier. Their share of the 6,416 specials is their weight ÷ the total weight (0.9822).
| Candidate | Preliminary | Share | × multiplier | = weight | Special share | + specials | Projected | Share |
|---|---|---|---|---|---|---|---|---|
| Michael Philip WOOD Labour | 13,582 | 45.05% | 0.945 | 0.4257 | 43.34% | +2,781 | 16,363 | 44.75% |
| Carlos CHEUNG National | 13,320 | 44.18% | 0.913 | 0.4033 | 41.06% | +2,634 | 15,954 | 43.63% |
| Anjana IYER Green | 1,574 | 5.22% | 1.606 | 0.0838 | 8.53% | +548 | 2,122 | 5.80% |
| Jamie ATTENBOROUGH Opportunity | 760 | 2.52% | No 2023 candidate, uses party-vote multiplier | 0.0370 | 3.77% | +242 | 1,002 | 2.74% |
| Yang QU ACT | 763 | 2.53% | 1.045 | 0.0264 | 2.69% | +173 | 936 | 2.56% |
| Priyanka DEO Animal Justice | 151 | 0.50% | No 2023 candidate, uses party-vote multiplier | 0.0059 | 0.60% | +39 | 190 | 0.52% |
| Total | 30,150 | 0.9822 | +6,416 | 36,566 |
2026 source: preliminary count (XML).