Which electorates to watch? In some of them, the spread of the party vote will tell you heaps about how the whole night is going. In others, high-profile local candidates will confound the big trends.

Who said it’s just about counting the numbers. Sure, the relative size of the urban swing to Labour and the rural swing to National in the party votes will determine who gets to form the government. But in many electorates high-profile candidates are going to confound those trends. And in some electorates, the relative sizes of the party vote will provide big clues to how the whole election is going. The one to watch? Whanganui.

Before we get to the seats, a note on our approach. These predictions look to the RNZ poll of polls for a baseline. Incorporating polls from the last nine days, RNZ has National on 45.1% (55 seats), Labour on 37.2% (46 seats), the Greens on 7.2% (9 seats) and NZ First on 6.6% (8 seats). Assuming Act wins 1 seat and the Maori Party 1 or 2, this result would mean National and Labour/Greens could both form a government, but only with the support of NZ First. Act and the Maori Party could join in, but they wouldn’t determine the outcome.

If the Greens and NZ First both get 5% or more on the night, National and Labour will be looking to control around 47% or more of the vote on the night, in order to form a government. If either the Greens or NZ First fails to make the 5% threshold, that golden victory target will drop to about 45%. If they both fail to make it back into parliament, it will be somewhere in the low 40s. See here for more on how to count the votes and here on what will be happening on the night.

But when you’re sitting there watching the results come in, don’t forget about the electorates: there’ll be some telling action on that front. Here are some of the ones to watch, and why. The assumptions are that the election will deliver the same result as the RNZ poll of polls.

Auckland Central

Education minister (and party leader in waiting – you read it here first) Nikki Kaye won this seat by only 600 in 2014, the smallest lead in the country. But her powerful personal support should turn into a comfortable majority this time. Still, it’s an urban liberal seat and the party vote is worth watching. National took 45% of it last time, against Labour and the Greens which both won around 22%. Both should push close to meeting National in the middle, and if one of them beats National straight up it will be a rough night for the governing party.

Christchurch Central

It was much the same in Christchurch Central: 44% to National against 26% for Labour, with the Greens on 16%. But the expected urban swing to Labour should see National’s Nicky Wagner (majority 2450) lose the seat, and the party vote should close right up too. It used to be safe Labour, after all.

Hutt South

Labour’s Trevor Mallard held on last time with a 709 majority, but he’s gone to the list. His replacement Ginny Andersen has been campaigning very hard, but then so has list MP Chris Bishop. He’s very popular and may well take the seat, but the party vote should swing strongly to Labour (last time: 45% to 28%). If Andersen wins and/or if Labour eliminates the party vote gap, it could be looking good to form a government.

Ilam

Ilam is Gerry Brownlee’s castle. Except there’s an upstart firing flaming arrows at the ramparts. Raf Manji was the top-polling councillor in the Christchurch City Council election last year, and is standing in Ilam as an independent. He’s fed up with both the main parties, and he didn’t want to join the troops for the Greens or Top either. He’s a former international banker, economically centre-right and environmentally radical, and if he does make it into parliament he will offer creative energy and a highly competent governance skillset to whichever side forms the government.

Will he win? He says there’ll be only a couple of thousand votes in it. If the good folk of Fendalton are in the mood for shaking things up a bit without necessarily going so far as to change the government, he could.

Maungakiekie

Last time the Greens took 9% of the party vote in this seat, but this year they’ve put in Chlöe Swarbrick to try to boost that much higher. She’ll be hoping for a party vote of 20%+. However, the consequence may be that she also picks up a lot of electorate votes, in the process denying Labour’s Priyanca Radhakrishnan the chance to win the seat. For more on this see here.

As with Hutt South: if Radhakrishnan wins the seat and/or Labour wins the party vote, the government could be changing.

Northland

Winston Peters won Northland in a 2015 by-election with a majority of 4441 and this is the first time he’s defended the seat in a general election. He might not win: National has a strong candidate, National has run an excellent rurally focused campaign, there have been no “build more bridges” fiascos and Labour’s Willow-Jean Prime has campaigned strongly (they basically stayed out of the by-election). She will suck up a lot of the anti-government vote.

Still, Peters has stumped all round the electorate, campaigning strongly too.

It’s too hard to call. But one of the few consistent trends in the polls has been the steady decline of NZ First. If that continues, both Peters and the party will be in trouble. It won’t be a case of “lose the party vote but saved by the electorate”, or vice versa. Either he’ll win the seat and the party will crest 5%, or they’ll both slip out.

Ōhāriu

Peter Dunne’s personal magnetism gave him a majority of 710 last time (just beating Ginny Andersen, now standing in Hutt South). If that margin isn’t comfortably eliminated this time by Labour’s Greg O’Connor, National will win the election.

As for the party vote, it went nearly 50% to National. That should be chopped in half.

Tukituki

Former Hastings mayor Lawrence Yule is standing in this seat. Despite being mayor during the Havelock North water contamination scandal, he was voted back into office in the council election last year, so locals obviously don’t hold him accountable. He should hold the seat. Still, he’s a new electorate candidate and Labour has the experienced and high-profile Anna Lorck standing against him. All things being equal, she should dent the 50% support Yule’s predecessor Craig Foss won in 2014.

But all things are not equal. Labour’s water tax will have frightened this electorate, full of wineries and orchards, and it’s more likely there’s a stamped to National.

Waimakariri

It’s the same story here. This semi-rural, urban fringe Canterbury electorate used to belong to Labour and it held up pretty well under Clayton Cosgrove. But he’s retired, water issues are in play, and while Labour has a strong replacement candidate in Dan Rosewarne, National’s Matt Doocey should expect a big increase on his 20914 majority of 2506. The party vote will go the same way.

But if those things are not happening, this is one of the seats that will tell you something big is up for Labour.

Wairarapa

Wairarapa was headed for an upset until the urban/rural divide took hold, with both Labour’s Kieran McAnulty and New Zealand First’s Ron Mark presenting as more dynamic and engaged politicians than the incumbent National MP, winery owner Alistair Scott. Scott doesn’t even live in the electorate.

This is another seat like Tukituki and Waimakariri: safe for National, along with the party vote. And like those other rural seats, if there isn’t actually a clear trend to National, Labour can take heart for the overall result.

Whanganui

The rural appeal should extend into the provincial cities too. But will it – or more to the point, how strongly will it?

Whanganui could be the best bellwether seat of all. There is no incumbent MP. It’s a seat both main parties have held in the past. The city has all the social and economic issues that should make it ripe picking for Labour. And, contradicting that, it has all the attachments to the rural economy (the dairy economy) that should push it even further into the arms of National.

National’s Chester Borrows won it last time with a majority of 4505, which was 13% of the total electorate vote. Labour won 14,144 votes. If it wins 2250 votes from National, it will take the seat. That would be a 16% increase on what it got in 2014. Compare that to the current polling, which suggests the party vote for Labour is going to increase from 25.1% in 2014 to 37.2% this time. That’s almost 50% more.

Labour is standing Steph Lewis, a mediator and the daughter of a prison officer. Her party profile says her parents were saving to buy a farm but “issues with access to the farm meant the money they saved to build a home went on lawyers’ fees instead”. National is standing Harete Hipango, a lawyer.

Strip out all the arguments about dairy and water and taxes and everything else. Labour should win this seat. The Whanganuis of the country should be part of its natural constituency.

But you can’t strip out all those things. Labour probably will win 50% more votes that last time. But how many of those extra votes will come in the provincial cities? Whangarei, Hamilton, Tauranga, New Plymouth, Palmerston North, Napier and Hastings, Nelson, Timaru, Greymouth, Invercargill. And Whanganui.

Labour should win the party vote in the big cities and National should do the same in the rural seats. But it’s in the provincial cities where the election will be won and lost. Watch Whanganui. It might not fall, but if it clearly moves to Labour, Jacinda Ardern should become prime minister. If it stays as it is or moves to National, it’ll be welcome back Bill.

The Māori seats

The seven Māori seats are very important: for a good run down on each of them see here.

On balance there may not be any change to the parties that win each seat, but the party vote will be critical. The Māori Party won just 1.32% of the overall vote in 2014 and desperately wants to do better. Anyone can vote for the Māori Party, of course, but in practice few people not on the Māori roll do so. So the party has to do better than last time in the party vote in the Māori seats. But even in Waiariki it won only 22% against Labour’s 44%. Te Tai Hauāuru gave it 17% and the rest just 10-12%.

In its favour: the tactical alliance that sees the Mana Party standing in only one seat. Mana, as Internet Mana, won 1.42% of the overall party vote in 2014, which is a little more than the Māori Party itself. If the Māori Party held its own support and picked up most of the Mana support, it would be in line for 3 seats.

Against: The Māori Party vote has collapsed to just 0.3%. Even with those Mana votes, it will win one to three electorates, but is most unlikely to be eligible for any more on the strength of its party vote.

Watch the Māori Party vote. If they win more than one seat without winning more than 1.3% or so of the party vote, they will create an overhang. The size of parliament will increase, making it a little bit harder for everyone to form a government.

simon@thespinoff.co.nz @simonbwilson