Alec van Helsdingen

I've finally gotten around to extrapolating the 2023 NZ census data out to the new electorate boundaries - and added selected stats to my electorate profiles. Pretty happy with how they've come out! #nzpol

In 3 seats (Banks Peninsula, Hutt South and Waitakere) National are more likely than not to lose. But if that didn't materialize, they gained seats (most likely in Wigram) or they did slightly worse on the party vote an overhang is very possible.

National currently has only 5 list MPs. Their party vote is down about 6.5% or 8 seats. So they'd win about 3 overhangs if the electorates stayed the same as in 2023.

This is a prediction for the next election. My prediction is that there is a ~ 50% chance that National wins more electorates than seats its party vote entitles it to.

more details: abmvanhelsdingen.github.io/nz.html

#nzpol

We have TPM in the lead in each of their six electorates individually. But they are more likely than not to lose at least one.

National and TPM have about a 50% and 80% chance respectively of winning overhang seats.

The coalition are favourites for re-election, though not as strongly as in December.

New Zealand Election Forecast

Labour 34.3 (+0.2)
National 31.4 (+0.1)
NZF 9.6 (+0.4)
Greens 9.3
ACT 7.7 (-1.2)
TPM 2.9 (+0.9)

+/- versus December

further details at: abmvanhelsdingen.github.io/nederland.html

Netherlands Election Forecast

D66 27 (+1)
VVD 22
GL-PvdA 21 (+1)
PVV 19 (-7)
CDA 16 (-2)
JA21 12 (+3)
FvD 11 (+4)
Others 22

+/- versus October 2025 election

NZ election on Saturday 7 November 2026. #nzpol

GitHub - ABMvanHelsdingen/Migration-Map Contribute to ABMvanHelsdingen/Migration-Map development by creating an account on GitHub.

Here's the GitHub repo: github.com/ABMvanHelsdi...

I'll put it on GitHub later this week.

The model for migration doesn't account for historical migration patterns, different population densities etc - it is purely based on geography. Some points are a few km offshore due to low-res coastlines used to determine whether a point is over land.

The GIF was made with gganimate in R.

Note there are only 512 points for generations 10 and further back to save computation time- these are a sample of all my ancestors.

This GIF is what I obtained from this simulation exercise. From 32 points mostly in small pockets of the Netherlands, the UK and Ireland, by 20 generations ago they are almost evenly spread across those three countries, and many are further afield.

#genealogy #rstats

So I used the distances between child and parent birthplaces in my known pedigree back to the 9th generation, and modelled these as a Zero-Inflated Gamma distribution. And with that I could simulate unknown birthplaces in my pedigree.

I was curious about how geographically spread out my ancestors were going back in time. I know where each of my 32 ggg-grandparents were born, but of course many lines peter out further back.

Nice 🌆

Whether it’s Pete Buttigieg learning to speak eight languages or Donald Trump successfully identifying a giraffe, there’s plenty of intellectual accomplishments on both sides of the aisle.

Much more likely they are moving to Labour, but simultaneously Labour are losing votes to National, and National losing votes to ACT and NZF.

And perhaps some of them no longer plan to vote for anyone.

More details, including charts and maps at: abmvanhelsdingen.github.io/nz.html

#nzpol

All three government parties gained over the last month, the opposition lost support, and the possibility of an overhang for National is now being considered by the model.

Thus we now make the current government strong favourites for re-election.

We forecast that for an election held now, National has a 70% chance of winning an overhang (more electorates than seats their party vote entitles them to). They have lost ~7% of the party vote since 2023, but only look like losing a handful of seats.

We make Labour favourites to gain Hutt South, Banks Peninsula, Te Tai Tokerau and Waitākere.

But they lose one seat (Ōhāriu) from the boundary changes and are in danger of losing Wigram.

For the first time, the forecast includes predictions for each electorate. This would not have been possible without the work of @overhang-aonz.bsky.social calculating the 2023 results if the new electorate boundaries had been in effect.

New Zealand Election Forecast
Labour 34.1 (-0.2)
National 31.3 (+0.8)
Greens 9.3 (-0.3)
NZF 9.2 (+0.3)
ACT 8.9 (+0.8)
TPM 2.0 (-1.6)

+/- versus November

#nzpol

Now out via the newsletter: the final exhaustive (and exhausting) analysis of the effects of the 2025 boundary review on electorate contests.
theoverhangaonz.substack.com/p/the-overha...

Plot with error bands (90% credible intervals)
#nzpol