This is very symptomatic of polling data in Wales.
The election is just over 2 weeks away and this is only the 2nd poll we’ve had in the past month! FOR A NATIONAL ELECTION!!
And pollsters can’t get their numbers right for the d’hondt system!
We must change the England centric view of politics!
Plaid - real 37, published 33 Reform - real 29, published 29 Labour - real 14, published 17 Green - real 4, published 3 Con - real 11, none published LD - real 1, none published
The JLPartners / @telegraph.co.uk poll was also completely wrong in Wales.
I ran their numbers and even the 82 seats they did call are wrong.
Plaid are under, Labour are over, and it should be 11 Tories and 1 LD, but they have 14 seats uncalled.
Yeah, my guess is that they've forgotten the compensatory part. I'll game it out and see.
I'm making an interactive comparison too, do you have the JLP list seats per region?
Absolutely fantastic statement for the Telegraph to make in their article where not only did the provider of their MRP forget to remove Greens from constituencies and suggest Reform's strongest seat in Scotland would be Strathkelvin and Bearsden, *they didn't even do D'Hondt correctly*
@ballotbox.scot @markmcgeoghegan.bsky.social - You'll enjoy this.
If anyone can provide me with the detailed regional seats that JL Partners projected I would greatly appreciate it.
I'm trying to track how reliable MRPs are, it's going well for them.
I firmly believe MRPs in Scotland are rubbish mainly because the polling companies are not accounting for Scottish demographics.
That they don't even know how AMS works is an embaressing sign of it.
The other pollsters D'Hondt maths are fine.
I don't mean 'it was different from what happens when you run their numbers through our system' like I usually do.
I ran their own regional numbers and constituency winners through a D'Hondt calculator. They did it wrong!
They announced an outright SNP majority under these numbers! Which was weird because they had the SNP losing seats across Glasgow and the South.
How where they going to sweep so many list seats?
www.telegraph.co.uk/politics/202...
Seats under various MRP Projections More In Common, YouGov, FindOutNow, JL Partners SNP 57 67 67 57 Labour 16 15 14 21 Conservatives 12 7 12 13 Liberal Democrats 14 9 8 8 Greens 8 11 13 11 Reform 22 20 15 19
On the topic of MRPs being effectively useless, I think the Telegraph/JL Poll messed up their own D'Hondt calculations.
The real numbers are:
SNP: 57
Labour:21
Conservatives: 13
Liberal Democrats: 8
Greens: 11
Reform: 19
This is really basic stuff.
Lets take a simple example, the Greens are only running in 6 seats. Does anything think they'd pick two of their absolute worst places?
This MRP thinks so.
I don't expect them to win Shetland, but 1%? Come on.
🗳️ REGISTER TO VOTE 🗳️
Monday’s the deadline – so if you haven’t registered to vote for May’s local elections, do it here 👇
https://www.gov.uk/register-to-vote
Shetland: Greens on 1%
Argyll & Bute, Lib Dems on 14%, in third
Strathkelving & Bearsden, Lib Dems winning on 35%, triple their Argyll & Bute result.
Glasgow Southside, Greens on 7%, in 4th
I'd be curious to know how they dealt with prospective Green voters, because the SNP drop looks suspiciously Green shaped.
I also think some of the MRP seat projection is....unrealistic.
Our polling averages (when adjusting for the Greens) show how far from the norm this one is, in particular the gap between the list and constituency ballot is much less than the usual 10%.
This does not automatically mean its wrong!
SNP 57 + Green 11 = 68, a majority of 4
The SNP would need the Greens (formally or informally) to do anything at all. This is the lowest pro-Indy majority we've seen in any recent poll.
Central & Lothains West: 2 Lab, 1 Con, 1 Green, 3 Ref Edinburgh & Lothians East: 1 Con, 3 Green, 2 Ref Glasgow: 2 Lab, 2 Green, 3 Ref H&I: 2 SNP, 1 Lab, 2 Con, 1 Green, 1 Ref Mid Scotland & Fife: 2 Lab, 1 Con, 1 LD, 1 Green, 2 Ref South Scotland: 2 Lab, 1 Con, 1 LD, 1 Green, 2 Ref West Scotland: 2 Lab, 1 LD, 1 Green, 3 Ref
The SNP could lose so many seats to the Lib Dems that they'd get 2 list seats in Highlands & Islands.
Map of Southern Scotland, Conservatives and Reform winning seats in far South. Labour & Lib Dems in Edinburgh, Lib Dems in Highlands. Rest SNP.
Map of Northern Scotland, Lib Dems in Highlands and Islands. Reform in Banff & Buchan coast. Rest SNP.
This is the most colourful map we've had in a while because the SNP are well down here. The Tories, Reform, Lib Dems and Labour all win seats.
The Greens don't win seats on 1%, but do if they get a bit closer to 2%.
Table of above results
Our projection for the Holyrood election based on the latest @moreincommonuk.bsky.social poll:
🎗️ SNP — 57
➡️ Ref — 20
🌹 Lab — 17
🐤 LD — 13
🌳 Con — 11
🌱 GP — 11
Make your own on:
devolvedelections.co.uk/scotland/
This is the first poll with a sample from April, someone needs to let the polling companies know there's an election on the 7th of May.
Plaid win 58% of the time, Reform 42%.
There have been so few polls in Wales this year that its very hard to be sure of anything.
Plaid are narrowly favoured over Reform in our 1000x simulation but only just.
devolvedelections.co.uk/wales-projects
Polling trends Plaid: 29.6 Reform: 26.3 Labour: 16.1 Con: 10.3 Green: 9.5 Lib Dem: 5.8
This is the worst poll for Reform in a while, Plaid are also down a little. Labour and Lib Dems are flat, as the Greens and Tories tick upwards.
Plaid 33 + Labour 14 = 47. 2 seats short Green 6, LibDem 3.
Plaid and Labour would be close to a majority, either the Greens or Lib Dems would tip them over the edge.
This could be done on a case-by-case basis rather than a formal coaltion.
Map of Wales. Plaid in the West and Cardiff, Reform in the East. Only Casnewydd Islwyn has Labour ahead
Reform and Plaid lead across Wales, except Casnewydd Islwyn where it's almost a three way tie between those two and Labour.
Afan Ogwr Rhondda Reform Plaid Cymru Labour Reform Plaid Cymru Reform Bangor Conwy Môn Plaid Cymru Reform Plaid Cymru Plaid Cymru Reform Plaid Cymru Blaenau Gwent Caerffili Rhymni Plaid Cymru Reform Labour Plaid Cymru Reform Plaid Cymru Brycheiniog Tawe Nedd Reform Plaid Cymru Liberal Democrats Reform Labour Greens Caerdydd Ffynnon Taf Plaid Cymru Greens Reform Labour Liberal Democrats Conservatives Caerdydd Penarth Plaid Cymru Reform Greens Labour Plaid Cymru Reform Casnewydd Islwyn Labour Reform Plaid Cymru Conservatives Greens Labour Ceredigion Penfro Plaid Cymru Reform Plaid Cymru Liberal Democrats Plaid Cymru Conservatives Clwyd Reform Conservatives Labour Plaid Cymru Reform Reform Fflint Wrecsam Reform Plaid Cymru Reform Labour Conservatives Plaid Cymru Gwynedd Maldwyn Plaid Cymru Reform Plaid Cymru Plaid Cymru Reform Plaid Cymru Gŵyr Abertawe Reform Plaid Cymru Reform Labour Greens Reform Pen-y-Bont Bro Morgannwg Reform Plaid Cymru Labour Conservatives Reform Reform Pontypridd Cynon Merthyr Reform Plaid Cymru Labour Reform Plaid Cymru Reform Sir Fynwy Torfaen Reform Conservatives Labour Reform Plaid Cymru Greens Sir Gaerfyrddin Plaid Cymru Plaid Cymru Reform Plaid Cymru Labour Plaid Cymru
Reform are narrowly taking a lot of 6th seats, even a slight georgraphic variation could see a few tumble or a couple more gains.
Eluned Morgan is one of those who could survive, she's behind by 0.5% for the last seat.
We have a button for this on the site: devolvedelections.co.uk/wales/
Table of seat numbers
Our projection on the latest JL Partner's Poll for next Month's Senedd Election
🌼 PC — 33
➡️ Ref — 33
🌹 Lab — 14
🌳 Con — 7
🌱 GP — 6
🐤 LD — 3
Make your own at devolvedelections.co.uk/wales/
Judge them by error from real results, that's what their aiming for.
If you wish to enter other numbers please feel free to do so, our site will work with any numbers you can imagine.
devolvedelections.co.uk/northern-ire...
LucidTalk's track record is very good and it is not our job to manually change pollsters numbers (except in the unusual case of the Scottish Greens)
Pollsters weight based on likely voters not demographics since some demographics don't vote as often as others.
That could help them, we'll factor in some candidate effects closer to the time but we don't know who's running yet.