8:11pm – I haven’t bothered to pay too much attention, but with most booths reported now the Nationals primary is down to just under 65%. An easy win, so I won’t be updating any further.
6:40pm – With ten booths reporting, the Nationals are on 71% of the primary vote, with the Shooters second on just under 15%. Easy win for the Nationals.
6:00pm – Polls have just closed for the Northern Tablelands state by-election in New South Wales. It seems likely that the Nationals will retain this seat with ease. If you’d like to read up on the by-election, you can check out my guide.
Nothing yet. I’ll be calling this one very quickly though.
Polling places:
Armidale: Armidale, Armidale EVC, Armidale South, Armidale West, Ben Venue
Glen Innes: Glen Innes, Glen Innes EVC
Guyra: Guyra, Guyra EVC
Inverell: Inverell, Inverell EVC, Inverell East
Moree: Moree, Moree EVC, Moree East, Moree North
Uralla: Uralla, Uralla EVC
Small rural towns: Ashford, Ashley, Black Mountain, Boggabilla, Boomi, Bundarra, Dangarsleigh, Deepwater, Delungra, Ebor, Emmaville, Gilgai, Gravesend, Kentucky, Mungindi, Newling, North Star, Pallamallawa, Rocky River, Ross Hill, Roxy, Tingha, Toomelah, Walcha, Warialda, Woolbrook, Yetman
The first results have come through from Ebor Public School. I would just call it right now for the Nationals but I’ll wait for a few more booths just to be fair.
Results are now in from North Star Public. The Nationals are still well ahead. One more booth and I’ll call it even though I could theoretically call it now.
From the current numbers it looks like it will end up as a National vs SFF 2CP count, although that is a moot point as the Nationals will easily win anyway.
Nth Star Comm. Hall – 98 out of 116 formal votes for the NAT candidate. That’s a crazy first preference count.
I wonder what the highest first preference count (by % of the total formal) for a single candidate in a venue has been. I would guess no votes for a particular candidate would happen now and again – as it did in this venue.
Results are now in from more polling places: Boomi, Gravesend, Woolbrook and Yetman. The Nationals are well ahead. Gravesend is the first booth to have a distribution of preferences, with the Nationals leading the Greens with 95.9% of the vote. The Nationals have won.
@G look at Nowendoc at the 2022 federal election.
Anyway, I’m calling this, NATIONALS RETAIN.
Lot of smaller venues starting to pop up. NAT candidate on 76.84% of first preference (583 ballot papers).
National party loss
Just joking
6:35pm – NAT candidate at 70.29% of first preference votes. Still only small venues reporting in.
Toomelah, a small Aboriginal community that seems to have always voted heavily for Labor in contrast to the rest of the seat which votes heavily for the Nationals, appears to have the Nationals and the SFF tied as the leading party there, both at five votes. The Greens and Fisher are both at four votes each while Ledger is at three votes.
BREAKING: Boomi Public School has voted 100% for the Nationals on TCP. The Greens have no primaries or preferences in that booth.
@NP Wow! Has this ever happened before? From memory in 2019 there was a booth in Riverina that votes 100% for the Nationals.
I basically called it at 6:21pm when the first booth results came in, 21 minutes after polls closed. But then I properly called it at 6:30pm, nine minutes after the first booth results came in. I think that’s a record for me calling an election on a live count.
Oh yep the booth was at Tallimba. 45 votes for Nationals on Primary vote, Labor recorded ZERO at that booth.
@James correct. That booth was Tallimba. 100% of TPP votes went to the Nationals, with only three candidates getting first preference votes there: the Nationals had 93.75% (45 votes), the United Australia Party had 4.17% (two votes) and the Greens had 2.08% (one vote). Yes, you read that correctly: in that booth Greens preferences flowed to the Nationals BEFORE Labor, something I’ve never seen before. But of course the most likely scenario is that the one Greens vote was from a donkey voter.
Toomelah appears to be the closest booth, with the Nationals and the Greens tied at six votes each (50% each). Though of course, it’s a very small polling place in a remote Aboriginal community and this is a by-election, so the already low voter turnout is extremely low, probably under 20%. The whole electorate’s turnout will be somewhere between 45% and 70%.
What’s happened so far:
✅ The Nationals easily hold this seat with a massive primary vote lead and a massive TCP lead
✅ Earliest live election call I’ve ever made (a record that will most likely stay for a long time)
✅ The Nationals get over 65% primary and nobody else even comes close to getting 20% of the primary vote
✅ Labor vote split between every candidate
✅ The Nationals win every single booth
✅ Armidale booths take ages to come in
✅ The Wikipedia editor doing the live results takes a break when it’s clear the Nationals have won a landslide yet again
And a wildcard that’s not a huge shock but still somewhat surprising:
✅ The Nationals win a polling place with 100% of the vote there
I should’ve made a bingo list!
NP – Are you making a play to replace Antony Green when he steps aside around the time following the next federal election?
To give some context about Boomi, the town that voted 100% Nationals on TCP, here’s some background info:
Boomi is a small rural town in the Moree Plains Shire LGA. It’s in the federal seat of Parkes and the state seat of Northern Tablelands. Boomi is located 94km northeast of Moree, and as of the 2021 census had a population of 207, with 54.5% being male. Indigenous people make up 7.7% of the population.
At the 2023 state election, the TCP vote for the Nationals against the Greens was 98.59% (70/71 votes), which is massive (if you’re wondering how I found this out, I used an online tool from the NSWEC where you can see the TCP or TPP between any pair of candidates in a seat and results per polling place). The actual TPP vote there between the Nationals and Labor was 94.59% for the Nationals, which is also massive.
@G we’ll see…
7:28pm – Most polling day venues reporting in. NAT candidate on 65.08% of first preference votes.
Finally got some votes in from Armidale. At Armidale City Public School, the Nationals have 321 votes. The Greens are somewhat close in second with 200, Duncan Fisher has 45 votes, the SFF have 44 votes and Natasha Ledger has 35 votes.
The Greens vote in Armidale is higher than other regional/rural cities and towns because Armidale is a university town (comparable to college towns/cities in America such as Hanover in New Hampshire or university towns/cities in the UK such as Oxford in England).
Labor voters in Armidale itself have mostly gone to the Greens or the Nationals, while in rural areas they’ve gone to the SFF or the Nationals.
In contrast though, Ben Venue Public School (also in Armidale though in the suburb of Ben Venue not the city centre) has 608 Nationals votes, 292 Greens votes, 136 SFF votes, 78 Duncan votes and 77 Ledger votes.
7:40pm – 4 polling day venues to come. First postals counted – NAT candidate first preference 494 out of 665.
NAT candidate 64.77% of first preference count.
Tipiing a nats win
Armidale Town Hall reporting in with 2175 votes. NAT candidate 1008 first preference.
3 polling day venues and all early voting venues to come.
@John “tipping a Nats win?” It’s over. The Nationals have 65% primaries. Only three election day booths haven’t got primary vote counts yet: Inverell Public School, Moree Masonic Centre and Mungindi Hall.
To be clear, Ben Venue PS and Drummond Memorial PS are both in Armidale, in addition to Armidale City PS, Armidale Town Hall and the Armidale EVC.
Mungindi is now the last election day polling place. Still nothing from there.
Boomi has voted 100% Nationals on TCP. Besides Tallimba at the 2019 federal election, have any other booths ever voted 100% for one party?
73 less 65…..8% primary vote swing… since there is no 2pp….. this can be spun as a loss of 8% for the coalition
I think the Nationals have performed relatively well in comparison to the previous (2013) by election. Adam Marshall was a former Mayor with a fairly significant profile. This time the new Nationals candidate Brendan Moylan has a much lower profile (non-political background) yet he still managed to poll 2% better compared to Adam Marshall.
Then again, Adam Marshall was running against another Mayor (Jim Maher of Armidale) in 2013 so in that regards the Nationals at this by election should have been able to poll over 70% primary vote against no name figures.
Writing on the wall for tweed
Another unsurprising landslide
@MQ if you use the online tool the NSWEC provided for the last election you’ll see that the swing against the Nationals and to the Greens on TCP was 3.9%. But it was still 85.4% which is massive and it’s 1.6% better than the 83.8% TPP vote they got in 2023 (against Labor).
Even though it’s a small booth in a rural area the Nationals will be happy about winning 100% of the TCP in Boomi.
Congratulations Brendan Moylan for your massive record-breaking win and I wish all the best for Adam Marshall (word on the street is he wants to succeed Barnaby Joyce in New England).
With only one significant venue yet to report in (Armidale – Early Voting Centre), NAT candidate sitting on 67.93% of the first preference vote.
Of minor note is the EC guesstimated the “second” candidate for the TCP incorrectly. It ended up being the SFF candidate over the GRN candidate.
More broadly, it might be time to remove the TCP Counts from election nights and implement a “likely elected candidate” model, which kicks in after you’ve counted the actual votes and probably after election day. In this specific scenario, one candidate received 50% +1 of the first preference votes, so you would not need to do this particular count.
Will all due respect I don’t think that’s a good idea. The whole point of the 2CP is to be a short cut so we can find out who is likely to win on the night. If you have to wait for the votes to be counted, there’ll be seats we don’t have a result in for weeks even when the top two is clear.
There’s no harm in having the rare miscalculation in the top two, they won’t even have to wait for the count to finish to do a new 2CP (although they may do so considering how decisive it is), but in the vast majority of cases when the 2CP is clear you’d be losing valuable information.
I’m not talking about no TCP Count or any significant delay. There’s no reason we have to find out the likely winner on the night; it can wait 1-2 days. We will still have first preferences, trends, historic TCP, etc.
The whole thing goes unnoticed when you get it right but when you get it wrong, it’s an issue. What if you have a random combination of a LOT of close seats and quite a number of very close 3 leading candidates – you might not know on the night anyway. Speculation is fun for election tragics but wouldn’t it be better to have a slight pause in the process and then have it done correctly across the board? Far more rigorous.
The current process still relies a human (or Antony Green) to guesstimate, so why not have a better foundation.
The TCP is an indicative count only. In the absolute purest scenario, you probably would not run it – you’d go straight to the full preference count and people would just have to wait. No harm would be done and the current goverment would continue to govern.
Back on topic though.
You have a system and a direction. You do the full preference count at the end as usual.
Once you’ve done all or the very vast majority of the first preference counts then you can clearly, each and every time, get an accurate “likely elected candidate” (LEC) by following one of three options.
Option 1 – If the leading candidate has 50% + 1 of the first preference votes, that person is the LEC. No TCP Count is undertaken as it’s simply not necessary. OR
Option 2 – If there are two clear leading candidates (as determined however) then 2CP Count.
Option 3 – If based on a determined % variable if there are 3 very close candidates leading on first preference, then a 3CP is undertaken.
EC’s by their very nature aren’t organisations who like to guesstimate and this removes it.
This also saves a bunch of hassle for the respective EC and their election day staff, when you can move the whole LEC count process to post election day with rested teams of staff counting.
This is such a bad idea I am struggling to express it politely.
You’re trying to fix a non-existent problem and in the process completely breaking the part of the system that works well.
In the vast majority of seats, the 2CP does a good job of telling us who is going to win on election night. If they don’t, you know that it is genuinely very close. When you experience an election without a 2CP (such as a mayoral election or a Tas LC count) there’s often much more uncertainty as to whether a seat is actually going to be called.
People have a legitimate interest in knowing the outcome as quickly as is reasonably possible and the 2CP does that well.
I absolutely do not agree that it’s an issue when the 2CP gets it wrong. It would be better if they got it right, but it’s really not a big deal.
Yeah if we start to have a lot more seats with complex three-way contests we could potentially switch to running 3CPs in some seats, but we’re not there yet.
People who call elections aren’t “guesstimating” – the 2CP is rigorous data. If you didn’t have it, we’d have to rely much more on guesswork.
Just to clarify, this “likely elected candidate” process happens at a central counting venue following election day where you have all (relatively) the ballot papers in the one venue and you know (or only have a modest amount of counting still remaining) which candidate/s have the most votes for the whole electoral area.
If the electoral area does require a count to determine the LEC, you just “bang out” each venue, one after the other until the LEC is resolved.
A better timing option might actually be after the check count. This would pick up any initial counting or transcription errors in the paperwork (which happens) and would also be helpful in very close 3CP counts.
Something like, all available single member electorate ballot papers to be initial counted and then check counted by the end of Monday. LEC process starts from the Tuesday morning. Depending on variables, the LEC could start from the Wednesday (as you could want a little more time to complete your check counts).
Just think… much more election TV.
Ben –
“This is such a bad idea I am struggling to express it politely.”
Respectfully disagree. Without going into specifics, I’m well versed on many elements of this field of work.
“You’re trying to fix a non-existent problem and in the process completely breaking the part of the system that works well.”
It works well when there are always two clear leading candidates AND they are correctly chosen ones by the person deciding. If incorrectly chosen, why should thousands of election day staff be undertaking a count which is going to be ignored shortly after its’ completion and which then has to be restarted by more staff? And to reiterate, I’m just suggesting delaying the TCP Count, not getting rid of it.
“In the vast majority of seats, the 2CP does a good job of telling us who is going to win on election night. If they don’t, you know that it is genuinely very close. When you experience an election without a 2CP (such as a mayoral election or a Tas LC count) there’s often much more uncertainty as to whether a seat is actually going to be called.”
“Calling a seat” is a media driven process, not an electoral process. Why would you not want a process that does it correctly (or correctly enough) in every single scenario? What benefit, knowledge or value did THIS TCP Count at THIS by-election provide? (That would not be provided by the full preference count).
“People have a legitimate interest in knowing the outcome as quickly as is reasonably possible and the 2CP does that well.”
Which people? What legitimacy? How does a delay of 2-3 days (of a tenative outcome) affect the operation of any element of society? You’re still going to know the likely sentiment, first preference swings and maybe probable broad outcome. “As quickly as is reasonably possible” is how long? I love to know election results as soon as possible but that doesn’t mean it’s best for the system operating the electoral process.
“I absolutely do not agree that it’s an issue when the 2CP gets it wrong. It would be better if they got it right, but it’s really not a big deal.”
So why have an electoral process where it’s ok if part of it’s wrong? Do it properly, right, once, the first time.
“Yeah if we start to have a lot more seats with complex three-way contests we could potentially switch to running 3CPs in some seats, but we’re not there yet.”
When will this be? Who will decide? Who will train the thousands of election day staff to do another process? Who will decide if the election day venue is to undertake a 2CP or 3CP Count and what if it’s wrong?
“People who call elections aren’t “guesstimating” – the 2CP is rigorous data. If you didn’t have it, we’d have to rely much more on guesswork.”
In that case, the chosen TCP candidates should be correct all the time or as almost close to all the time as is rigorously possible. It didn’t work for this by-election. And it’s nothing about calling elections. It’s about delaying the proces slightly to get the best result – you still get your likely winnning candidate, you just have to wait a little longer. The TCP data is rigorous if it’s for the voters’ two most preferred candidates.
The TCP count is “Two Candidate Preferred”. The two candidates who voters most prefer. Most prefer in actual ballot paper reality, not are likely going to most prefer. You can very likely work this out in many cases but you can’t actually know until the voter has put pencil to paper and you have the two highest vote getters in front of you.
So why not acknoweldge all this and work out a more cohesive, single, replicable in all situations system around the reality of the actual vote rather than have multiple outs, variants and workarounds when the long established default gets wobbly.
“why should thousands of election day staff be undertaking a count which is going to be ignored shortly after its’ completion and which then has to be restarted by more staff”
Because this issue is relatively rare and the process works well the rest of the time. There is the option to suspend the 2CP count if it is clear that it won’t be correct.
“I’m just suggesting delaying the TCP Count, not getting rid of it.”
But the value of the 2CP is in the immediacy of the data. Delaying it massively reduces the information we have.
“What benefit, knowledge or value did THIS TCP Count at THIS by-election provide?”
You can’t know ahead of time if a 2CP count will prove to be useful or not, but by the time you know that you need one, it’s too late to start. Hence why we do it everywhere.
Again I don’t understand the argument that this needs to be 100% correct or is not worth doing. It works really well most of the time.
I’m becoming increasingly convinced this is some kind of troll as it’s hard to believe anyone could actually believe something so silly.
Interestingly, in this seat in 2023 state election Labor only got 10% of the primary in what was a great result statewide, There was also no prominent independent. Labor actually won this seat during the Wranslide and in 1991 they got 31.6% of the primary vote. Could it be the norm that in the future that even in good elections that Labor gets single digits and maybe it maybe a NAT V SFF 2CP or even NAT Vs GRN 2CP. The will be a student vote in Armidale so the Greens could outpoll Labor in this seat as well even if it nowhere near enough to win this seat.
@Nimalan I think eventually it could be Nationals vs Greens or Nationals vs SFF. But I still don’t think Labor or the Greens will win any booths.
At the 2022 Federal election, Labor won the 2pp at four of the Armidale booths. Overall the town would have voted for the Nats but it is still possible.
As Nimalan points out the seat was held by Labor during the Wran years. In those days, the seat was on the Tablelands only. With strong independents in the region – Richard Torbay and Tony Windsor – Labor took the “the enemy of my enemy is my friend’ approach and basically outsourced their efforts. In the long term, they are basically left with nothing. They have done similar things right across inland NSW and also in Northern Victoria though the latter has never been friendly Labor territory. I assume local branches – whatever there is of them – must be nothing more than husks – how they get anybody out on election day is a mystery – or they don’t. It partly explains why the Labor primary votes are in the low 30s and they are so dependent on preference flows.
Agree Nether Portal, i still dont think Labor or Greens will any booths, nor will Greens campaign here so it is more symbolic if the Greens outpoll Labor in a seat they dont both target. Another seat which i think Labor may get outpolled by SFF is Cootamundra it is really only Gundagai that is Labor friendly.
Thats it an interesting point Redistributed, maybe that will happen in the Teal seats as well if the Libs eventually win them back and maybe in a decade even in good election Labor will only poll in the low to mid Teens and the Greens will outpoll Labor. Even back in 2016 Labor fell below 20% in Kooyong and the Greens nearly out polled them. Peta Credlin often claims the Teal seats are “transition” and that voting Teal is just a stepping stone until Labor eventually wins the Teal seats. However, maybe the Teals will just finish Labor off just like as happened in Northern Tablelands, Northern Victoria and maybe even Gippsland East.