2025 House Of Reps Postcount: Coalition vs Teals (Goldstein, Bradfield, Kooyong etc)

On this page:

Bradfield (IND ahead, extremely close, headed for recount) - coverage on new thread

Goldstein (Liberal gain from IND)

Kooyong (IND retain)

Click here for link to Reps postcount hub and tallyboard page.  

-------------------------------------

This is another grouped seat thread to cover those seats where there is a competitive straight 2CP contest (realignment not required) between the Coalition and any teal (or teal-adjacent if I add more) independent.  There are currently three of these I am watching: Bradfield, Goldstein and Kooyong.  On election night it appeared Monique Ryan had retained Kooyong, Zoe Daniel was ahead and likely winning in Goldstein and Nicolette Boele was projecting ahead in Bradfield, but the early postal counts have had some savage commentary on that.  Updates will follow as the count unfolds over coming days.  It's also worth noting that while I refer below to teals doing well on absents compared to other indies, this tends to be by default because the Labor and Green primaries are high (the teal primary tends to be low).  This may mean that performances on absents are less strong this time around (see comments).

Bradfield (Lib vs IND 2.5%)

Sunday: In Bradfield the live count is at 81.2%.  Nicolette Boele (IND) leads Gisele Kapterian (Lib) by 905 votes, but I project this to drop to about 796 when a missing 2CP for the booth St Ives North is added.  The first 8000 postals cost Boele 1385 off her lead.  There may be about another 7000 to come, but it's highly likely that the remainder will be significantly less hostile, because of the early postals/late postals dynamic.  Could they still be hostile enough to wipe out the expected lead?  They could, but that's no guarantee.  (The postals counted so far are 10 points worse for Boele than day votes; last time this gap finished up at 6).  In 2022 Boele did well on absents (independents often don't) but not so well on declaration prepolls; possibly these two categories will disadvantage her slightly combined.  At the moment this seat seems to be tracking to a very close finish.  Would I rather have this lead in the hand than be chasing?  Yeah I probably would ... but only just.  My standard booth health check did not catch any errors in this seat.

Monday: The live count is at 83.6%.  The St Ives North booth knocked 89 off Boele's lead and another c. 2500 postals took another 400, though there was some slowing in the damage rate.  (Postal split is now 6167-4384).  Boele leads by 416, it's hard not to see Kapterian taking some sort of lead on the remaining postals, the question being if it's enough.  Possibly I would now take Kapterian's position. 

Tuesday: And I would definitely now take Kapterian's position - unknown to me the St Ives PPVC booth had only partially reported, on fully reporting it has improved Kapterian's position by 440 votes which looks a lot like game over to me.  

Wednesday 8:40:  More postals broke 539-414 so that's a further slowing of the rate to 56.5% but Kapterian is now ahead by 178.  

Thursday 10:00:  Now the late postals are really starting to do their thing with the batch in yesterday breaking only 495-468 to Kapterian, who leads by 215.   There are 889 awaiting processing plus however many still arrive which might be a few hundred, but by this stage postals are no longer reliably doing much for Kapterian and might indeed stop breaking to her at all.  Because postals have started breaking so weakly I have put this one back to likely Liberal win.  There is some potential for absents to behave oddly because of the redistribution - for instance it could be that Boele does well in absents in booths that were in Bradfield last time but no longer are.  There are also still several booths (some large) awaiting rechecking.  

Thursday 12:20: Absents will be added today.

Thursday 3:30 Absents haven't been added yet but the lead has come down to 194 presumably on rechecking.

Friday 6:00 I seem to have somehow overwritten yesterday's update in which Kapterian gained 50 votes on the first batch of absents, mainly because of a very low primary for Boele but also because of a weak preference flow to Boele in them.  Kapterian has also gained 6 on out of division prepolls and now leads by 250.  These have a similar issue - low primary vote and weak preference flow.  Every booth of any serious size has been rechecked so it's not clear where Boele can still gain from (well, late postals may help a bit but there won't be enough of those!)  It again looks very likely that Kapterian will win the seat.  

Friday 7:45 Boele did make a gain of 41 off the remaining postals on hand and now trails by 209.  There won't be many postals next week so she needs a big turnaround in one of the other categories. 

Saturday 1:15 Boele gained seven on rechecks so far today and now trails by 202.  Six booths are still to be rechecked.

2:40 More absents have been added boosting Kapterian's lead by 12 but this appears to have been more than cancelled out by rechecks as the gap is now 193.  Three booths awaiting rechecking - one is the already noted St Ives PPVC which is relatively large and might be Boele's last chance.

Sunday: Kapterian now leads by 211 (not sure what the gain was on here).  There may still be about 2700 votes to come, Boele would need some miraculous batches of absents or prepolls.  

Monday 4:50: Various networks have called this for Kapterian today; I have not absolutely called it but I think Boele's chance is very low.  Today she got a 53.9% split off a batch of absents but she would need to roughly keep doing that over all the remaining votes to win - the gap is now 180.  The automatic recount margin of 100 is worth keeping an eye on here though the chance of any seat on a margin of 75+ falling over on a recount is low.  

Monday 5:17: Boele gains 1 on out of electorate prepolls which is not enough as it takes another c. 300 votes out of the count, she will now probably need over 54-46 on the rest.  

Tuesday: And this is why I did not call it ... Boele has had a spectacular batch of out of electorate prepolls (62.4% 2CP) and now trails by only 59. There's still over 1000 of those to go ... Bradfield could be headed for a recount!  (I estimate remaining formal votes at 1040 out of div prepolls, 330 absents, 280 postals arrived plus a few hundred still to be received.  Anything can happen here.)

Wednesday: Most of the remaining votes counted and Kapterian's lead is up to 80 with several hundred to go. A much more comfortable position than yesterday and might even yet avoid a recount. On the other hand if it comes back a few dozen a recount could still flip it.

Thursday 11:15:  Not a good sign for Kapterian's hopes of avoiding a recount as her lead drops to 70 with 533 to go (closer to 500 after informals) plus any postals that arrive tomorrow.  

Thursday 3:00:  Margin now down to 43 with about 400 to go plus any further postals (I may have missed 105 outstanding prepolls earlier).  Recount now very likely.

Thursday 4:25: SMH has reported no action today.  Also suggests that the distribution of preferences will go ahead ahead of a full recount (if needed).  

Monday 19th 12:40:  Lead is 50 with 453 awaiting processing (mostly dec prepolls followed by postals).  Kapterian would need 55.6% if all were admitted and formal, but probably more like higher 50s, to avoid an automatic recount - still might happen.  

Monday 3:05: Sensation!  Boele gets a huge break on the last lot of postals and jumps to a 19 vote lead.

Monday 3:40: And now 40 votes.  I am advised there are no votes remaining, the 59 showing are partial admissions where the voter's Senate vote is accepted but not their Reps vote.  There will now be the distribution of preferences, followed (all but inevitably) by the automatic recount if the margin is still under 100.  I am writing a fresh post to follow this.  

Coverage continues on new thread

Goldstein (IND vs Lib 3.9%)

Sunday: Amid the trainwreck of the Coalition's results a bright spot on Sunday has been Tim Wilson surging towards recovering his seat from Zoe Daniel.  Goldstein in 2022 had a large postal count that flowed strongly to Wilson producing a postcount shift of 1.63% 2CP from the ordinary count (including within division prepolls) to the final results.  This year Daniel has 51.73 in the ordinary count, which makes it surprising that the ABC was projecting it to her at all last night; maybe some prepolls had not reported at the time.  Daniel's position is clearly fragile on that basis alone but Goldstein is another Victorian seat with an increased postal count; I would expect around 21 K opposed to 17 K in 2022.  While the 2022 postals were bad enough for Daniel (9.5% worse on 2CP than the ordinary votes) the first half or so of the 2025 postals have been a brutal 16.7% worse.  That will tail off as is the usual pattern with postals but with Daniel only 95 votes ahead it is hard to see the next few thousand postals not putting Wilson well into the lead to stay there.  

My standard booth health check detected a possible error in one booth - Malvern Goldstein PPVC - where the preference flow to Daniel appears oddly weak at 44% to Daniel; every other booth is at least mid-60s.  If this is an error it would only improve her position by 100 or so votes, however it's not the usual transposition error if so; it could be 50 votes in the wrong pile for instance.

Tuesday: I may have accidentally overwritten a comment yesterday when the lead narrowed to 50 or so votes but anyway Wilson is now 387 ahead.  There are still many thousand postals to come and while the damage might slow (the gap has come down to 16.3%) I am not at this stage seeing any reason to doubt that Tim Wilson will recover his seat.  

Saturday: Ben Raue spotted an apparent error in flow in the Hampton polling place and on checking it it appears to me that the 2PP numbers have been reversed in data entry on the second entry - if correct this is a very rare error at AEC level.  Assuming it is an error and is fixed it cuts the margin by 170, but the current margin is 1257, there is still no reason to think the seat could be saved (especially as it is the only one.)

Monday night: Closing in the Goldstein count today to a 963 vote margin from over 1300 has attracted comments on social media but a large chunk of the closing was simply the Hampton error noted above being fixed.  Daniel did gain by 1024-892 on absents, a 53.44% split.  That's because absents tend to be bad for Liberals and good for ALP and Greens, which offsets them being also bad on primary votes for Daniel.  Absents being 1.58 points better for Daniel than ordinaries is in fact weaker than 2022 when they were 2.68 points better for Daniel, so all this has been taken into account.  The absents are about half done and most of the rest is out of division prepolls, which are typically also bad for independents on primaries but less bad for Liberals; on 2022 patterns the out of division prepolls will be slightly better for Wilson than ordinaries (this is what's happening in Kooyong) and will do close to nothing, But even if they were as good for Daniel as absents, there shouldn't be more than about 7000 votes left and Daniel needs those to split twice as strongly as absents.  So no, nothing to see here at all.

Tuesday: Daniel has in fact done remarkably well on the first 2000 absents, the primary vote for Wilson in these is low and the break is 54-46. However she now still trails by 798 with what I expect to be just under 5000 to come, and needs a 58-42 split so even that won't do it. Some more social media nonsense about this today, premised on people not realising that not everybody votes.

Tuesday 4:30: And now Daniel almost does get the asking rate on the next 1000 absents, with a 57.2-42.8 split, so the margin comes down to 660.  On the rate of absents and out of division prepolls so far Daniel is projected to gain a further 242.  Provisionals did nothing much in 2022 (which seems unusual as normally they skew left), but there are fewer than 200 of them anyway.  Remaining postals won't help Wilson, may even help Daniel a bit.  I estimate 991 absents and 1773 out of division prepolls to go.  

Wednesday: Again, remarkably, Daniel almost gets the required rate on out of division prepolls (I don't have the exact split but looks like 57-43 off 900 or so.) Would now need close to 60-40 off remaining votes, a higher share of which are postals.  

Wednesday 3:30: About 500 postals were counted and this being the end of the postal count it is, they split to Daniel but only very weakly (she gained another 20 and now trails by 508).  There are now 1035 absents awaiting counting (expect 991 formal), 195 provisionals (expect 186 formal), potentially 875 prepolls (expect 823 formal if all the outstanding few hundred exist), and 78 postals plus however many more trickle back (say maybe 200).  So say about 2278 to go and Daniel needing 61%.

Wednesday 5:40 The constant (and pretty amazing really) narrowing in this count is producing much interest - Daniel had another very strong batch of absents but I still estimate the asking rate at just over 61% (and this is after a revision to how I calculate rejection rates that has slightly increased my estimate of what is left).  Just over 58% to take it to a recount.  

Thursday 10:25 Provisionals (excepting 21 held back) have been counted and of course after doing nothing in 2022 they also broke heavily to Daniel (68-35), again just below the required rate, and the gap is 368.  However only another 20 postals were added to votes awaiting processing, making me suspect the number still to return is probably below 100, in which case Daniel would need at least 62.6% to tie or 58.7% for a recount.  There has been some amusement with leaked internal messaging circulating online of Wilson scrambling for scrutineers to make sure he wins the victory he already claimed (yes the same one where Daniel says we should respect voters by waiting for the seat to be clear after not doing that on election night).  

Thursday 11:20 Down and down and down - like a T20 cricket game where the chasing team needs 14 an over to win and gets 12, then needs 18 to win and gets 16, then needs 24 to win and gets 22 and ends up needing nine off the final ball ... I don't have the exact break but Daniel has closed by another 66 votes to 302 off about 280 absents, which again is at almost the required rate, but absents are her best major category so far.  Another three postals have made it into the count; there are now 951 votes awaiting counting but a few percent will be informal, plus whatever postals return tomorrow.  

Thursday 1:25 The margin is out to 305 and the number of absents awaiting processing has gone down by 18 (not sure what all that's about), also the number of out of division prepolls awaiting possible return has been increased by 12.  (Edit: another 8 prepolls now disallowed.) It seems I overwrote an earlier revised estimate of the getting rate but it is irrelevant now as more votes have been counted.

Thursday 3:07 A weak break to Daniel on a few hundred out of division prepolls gaining only a further 12 votes while the number of remaining such votes has gone down by 422 despite some more being added to envelopes issued.  Daniel now needs probably 68% of what is to come (around 62% for recount) which won't happen.  

Friday 12:30 There was a small adjustment possibly on formality rulings with the margin coming down to 289 then a throw of out of 250 out of electorate prepolls with Daniel gaining a further 31 (about a 56-44 split) - again very good but not what is needed.  The current margin is 258 with potentially 570 votes outstanding (not all of which will be formal) plus today's postals.  Even if there are 650 to go (I'm allowing 100 postals today which is probably generous) then Daniel needs 70% to tie or 62% for an automatic recount.  If Daniel misses the automatic recount margin narrowly that would improve her chances of a discretional recount but unless she can show systematic errors in the count that wouldn't overturn a margin of over 100.  Given the quality of Daniel's scrutineering (at least according to Wilson!) I suspect there is not much left to claim.    I wish to stress here that the leaked group chat messages from Wilson are NOT a valid reason for a recount - he was simply seeking scrutineers to do their jobs and the AEC decides what is a formal vote with scrutineers from all sides able to have input into the decision.

1:40:  254 now, another 4 off with little change in votes remaining.  

8:00 I've had a recurring issue with overwriting updates in this coverage - this afternoon the margin dropped to 208 and now to 206 but the number of votes left is now just 332 (not all formal) plus a probable few dozen postals.  Daniel would now need about 65% to make the automatic recount margin.  

Tuesday: There was no counting on Monday because of a wait for a very small number of remaining votes to be transported.  The preliminary count should finish today.  Nicolette Boele's surge in Bradfield where she picked up 90 votes off the last few hundred shows some potential that Daniel can still get inside the automatic recount margin though she would not necessarily stay there after the distribution of preferences.  Bradfield also highlights that some number of the remaining votes may not actually be countable (in Bradfield's case 59 were partially valid declarations acceptable for Senate purposes only).  

Tuesday 20th 1 pm: Margin is 203 with 308 shown as potentially remaining.  

3pm: More minor adjustments, 204 with 316 showing.  

4:10 I am hearing that Wilson has won the initial count by 124.  The distribution of preferences will be next with Daniel needing to gain 25 for an automatic recount (even at the outer edge of an automatic recount there is no reason to suspect the outcome would change).  

6:40 When I first saw the site updated the margin was 128, it is now 131.  The distribution of preferences has commenced and errors being corrected in booths will cause the margin to bounce around.  I won't be regularly commenting on these changes unless it gets close to the automatic recount margin.  

Thursday 22nd 6 pm: A piece in the SMH references Daniel considering asking for a recount and referencing some points made by Simon Jackman.  One of these is population growth since the under-100-votes threshhold was introduced in the wake of the McEwen 2007 challenge.  In terms of the numbers of either formal or total votes in Goldstein, a 99-vote maximum in 2007 converts to a 134-vote maximum in 2025 (or arguably 135 or 136 depending on rounding); as I write 134 is the margin. Also it is not clear to me that the threshhold is intended to be population-sensitive; Michael Maley has said that from memory it was inspired by the potential for a bundle of 50 votes to be misplaced.  The 2008 Henderson report that recommended the 100 vote threshhold noted that usually in the past to that point margins over 100 had been rejected and margins above 100 accepted, with one exception either way. There have been three previous rejections in the 101-120 vote range.  

Jackman refers to an issue at the Hampton prepoll booth where votes were transferred from Daniel to the Greens; I can confirm that this booth's numbers have changed since the early stages of the count via a c. 500-vote correction but this looks to be obviously a case of a 500 vote parcel in the wrong pile as the Greens were originally on 2.7% (cf 6% in 2022) when the Hampton day booths have slight swings to them.  I am certain Daniel's scrutineers would have been all over it in rechecking if the votes were actually hers.  The 2CP margin change in this booth was 144 votes in Wilson's favour.  Jackman has also argued that the count had been "a bit bumpy" in prepoll centres but bumpy prepoll counts are entirely normal given the size of the booths and the speed they are counted at on the night.  None of this looks to me like convincing reason for a recount.   I note also that the AEC does have the power to recount only a part of the count (eg a specific booth).  

There was also an issue in figures for the Hampton day booth (noted on thread above) but this was fixed.  

It should also be noted that the argument that a population increase justifies a proportional increase in the recount threshhold is incorrect.  Errors that are captured in recounts are frequently random in nature, producing a degree of drift back and forth in the recount process.  As the sample size of votes increases, the chance of a total error that exceeds a certain percentage of the sample does not remain the same, it decreases.   

Friday 23rd 3:30: Wilson now 167 ahead, further weakening the already weak case for a recount,  That looks like there was a substantial correction somewhere but I haven't been watching this one closely to say.  

Friday 4:15:  And now a bigger correction, looks like 48 Wilson preferences in the Daniel pile at Oakleigh PPVC, Wilson leads by 264.  

Friday 5:45: Would you believe 86 Wilson preferences in the Daniel pile at Hampton PPVC, Wilson leads by 436? Maybe Wilson's scrutineers really were asleep!

Friday 10:40: 444 now.  I don't know if that's the end of the distribution yet but all the large PPVCs are done making significant errors far less likely.  

Saturday 24th 11:30: I posted this update earlier but it did not go through for some reason.  There's been a further big correction, this time on postals and going the other way with Daniel gaining a massive 137 from Wilson, counterbalanced by Wilson gaining 22 from Daniel on absents.  Wilson now leads by 220. It is unusual to see so many substantial corrections this late in a close count.  Because postals do not show as a booth with an update time it is unclear whether all postals have been checked yet.  

1:50 And another one but by now they are booth corrections so only small - in the Beaumaris Central booth Daniel loses 19 to Wilson and the gap is 260.  It would not be surprising if Daniel uses the large number of corrections in the preference distribution to call for a recount on the grounds that the primaries could have similar errors but I am not sure the AEC will accept that; they may argue that the candidates were entitled to have scrutineers present who should have caught such errors.

2:40 260 is the margin at the end of the distribution of preferences and a recount request has been submitted. The district returning officer will make a decision, a rejection can be appealed at a higher level.  I understand that the fluctuations observed in the distribution of preferences have been cited as part of the reason for the request.  The AEC's options include granting a partial recount (eg of the leading candidate's primary votes and informals).  

Overnight 24th/25th: just adding a note that as best I can tell none of the several significant corrections in Goldstein have involved incorrect primary votes.  All have involved votes misattributed to 2CP, either in the form of votes being physically in the incorrect pile or in the form of numbers of votes being incorrectly recorded.  

Monday 26th 12:30: The AEC has granted a partial recount, which will take about four days starting Wednesday, and which will be of primary votes for the leaders and informals.   The AEC statement confirms that the previous issues with the count were data entry errors during fresh scrutiny (rechecking).  Specifically the issues included both one instance where the provisional total from the first count had not been updated to the new total, and cases where numbers were incorrectly entered.

Kooyong (IND vs Lib 3.5%)

Sunday: This is a similar story to Goldstein except that Monique Ryan starts in a better position.  Kooyong in 2022 also had a 1.63% postcount shift but Ryan has an ordinary vote 2CP of 52.30%.  In 2022 over 19000 postals were 8.3% worse than the ordinary votes.  In 2025 the Kooyong postal count might reach about 21K with 13K still to come; the first 8K were 14.8% worse.  It remains to be seen how strongly for Amelia Hamer the remaining postals will be once the late postal left shift kicks in but if they were, say, 8% worse then that would still leave Ryan with a lead of a few hundred from her current 1900.  Even 10% worse would be survivable, but continuing at the current rate would not. Ryan should also make modest gains on prepolls and absents combined.  At present I think that she is still pretty well placed to hang on.  My standard booth health check did not catch any errors in this seat.

Monday: A second lot of postals has been added breaking 1230-747 to Hamer, which is almost exactly the same break as the first lot (4934-2959).  Ryan's lead is now 1408; if the remaining postals are now 10% worse than the booth votes then that would make her position very dicey (needing to gain a few hundred off absents, out-of-electorate prepolls and provisionals, and the lack of a letup as yet increases the chance that they will be that or worse.  So I am considerably less confident re Ryan holding on compared to yesterday, but we will see how further lots of postals go.  

Monday 8:40: A third lot of postals has gone in breaking "only" 1194-778 (60.5% to Hamer); I think it would be optimistic to say that's the start of the slowing as it's within margin of error of the previous two. These were a mere 12.8% worse than the booth votes.  Ryan's lead is down to 995 with possibly 9000-9500 postals to go.  (Note that not all postals issued are ever returned) 

There's been interest in what is going on with these postals.  Some possible factors include Orthodox Jewish votes (significant in both Kooyong and Goldstein), better Liberal targeting and also a reversion of postals to a more conservative voter set after 2022, when some postal voters were more left-leaning than usual because of COVID.  

Wednesday 4:50: More postals have gone in in two batches since my last update, and I believe the rate slowed in the latest batch, but the postals in total are still 13.4 points above the previous batch (current postal tally 9633-6136).  The proportion of postals that has come in is high with 83.7% of those issued already back, so there is going to be an unusually high postal return rate to compound Ryan's problems.  There could easily be 6000 formal postals left and even if the rate on the remainder is only 54 or 55% to Hamer that will put Hamer into some sort of lead and Ryan will have to overturn it on absents and prepolls (which is not guaranteed).  However if the postals are weakening faster than that (I believe today's final batch was only 55) then Hamer might not even get enough to lead.   Ryan's current lead is 366.

Wednesday 6:40: Major development - a rechecking correction in the giant Kew PPVC booth has increased Ryan's lead by a priceless 210 votes (she now leads by 590 after other minor changes).  The booth is so large that my booth health check can't catch an error that proportionally small, but given how close the count is it makes a big difference.  That makes it very difficult for Hamer as it will be challenging to erase Ryan's lead on the remaining postals.  

Wednesday 7:40: Again! Hamer has dropped 100 votes on rechecking in the Ashburton PPVC.  

Thursday 4:45:  Another lot of postals today broke 1030-932 (52.5 to Hamer).  Barring a large correction in remaining rechecking (and there are some sizeable prepolls still unchecked) Ryan looks safe now.  

Thursday 8:30 And even safer after a second lot broke 1034-928 to Ryan.  I've talked a lot through this count about how early postals can be very right wing but by the end of the count they can slacken off so much that they are even sometimes trending left, and here's a case in point.  

Friday 9 pm In what seems to be the usual teal dynamic, Ryan's doing very badly on absent primaries but being pulled back up on absent preferences by Labor and the Greens, who do well.  She lost 2000 absents today 984-942 but that gets rid of 2000 absents so that's fine.