|
|
Post by greenhert on Jun 16, 2024 8:15:50 GMT
Can someone explain why the Survation MRP has Reform winning this? I am truly bewildered. Survation has lots of weird results, e.g. the Lib Dems gaining N Dorset rather than the much more winnable Mid Dorset & Poole N snd there are lots of other examples. Pinch of salt time I think. Survation polls fail to sufficiently take into account the importance of tactical voting and the impact of smaller parties and Independents, and special cases like this. |
|
|
|
Post by shadsy on Jun 16, 2024 11:36:22 GMT
This is an interesting betting contest.
Latest odds:
4/7 Cons
5/2 LD
6/1 Lab
20/1 Reform.
I've had a small bet on Labour.
|
|
carlton43
Reform Party
Posts: 50,900 
Member is Online
|
Post by carlton43 on Jun 16, 2024 11:40:15 GMT
This is an interesting betting contest. Latest odds: 4/7 Cons 5/2 LD 6/1 Lab 20/1 Reform. I've had a small bet on Labour. You can probably afford it. |
|
graham
Non-Aligned
Posts: 1,345
|
Post by graham on Jun 16, 2024 12:04:57 GMT
Survation has lots of weird results, e.g. the Lib Dems gaining N Dorset rather than the much more winnable Mid Dorset & Poole N snd there are lots of other examples. Pinch of salt time I think. Survation polls fail to sufficiently take into account the importance of tactical voting and the impact of smaller parties and Independents, and special cases like this. Though MRP polls are also likely to influence tactical voting. |
|
steve
Non-Aligned
Posts: 547
|
Post by steve on Jun 16, 2024 12:41:31 GMT
This is an interesting betting contest. Latest odds: 4/7 Cons 5/2 LD 6/1 Lab 20/1 Reform. I've had a small bet on Labour. You've shifted the odds slightly with the Cons drifting out to 11/18 on. |
|
haroldthepolitician
Labour
'You can be just what you want to be, Just as long as you don't try and compete with me' - Jarvis
Posts: 216
|
Post by haroldthepolitician on Jun 17, 2024 11:20:40 GMT
Survation has lots of weird results, e.g. the Lib Dems gaining N Dorset rather than the much more winnable Mid Dorset & Poole N snd there are lots of other examples. Pinch of salt time I think. MRPs are never good at predicting every individual seat, only at predicting the overall seat total. I don't particularly agree with this, it's more a case that alot of MRP projections right now are incredibly lazy and just done for the publicity of it being an MRP. The only MRP projections I genuinely trust are Yougovs at the moment as they have a consistent track record both at being relatively accurate on overall outcome, but also being consistently accurate on individual seats. (Yougov had some strong predictions in 2019, like Canterbury staying/swing towards labour.) |
|
|
|
Post by batman on Jun 17, 2024 11:36:31 GMT
I wouldn't go as far as saying I trust YouGov's MRPs, but they are more plausible than those of other companies.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 17, 2024 11:46:05 GMT
MRPs are never good at predicting every individual seat, only at predicting the overall seat total. I don't particularly agree with this, it's more a case that alot of MRP projections right now are incredibly lazy and just done for the publicity of it being an MRP. The only MRP projections I genuinely trust are Yougovs at the moment as they have a consistent track record both at being relatively accurate on overall outcome, but also being consistently accurate on individual seats. (Yougov had some strong predictions in 2019, like Canterbury staying/swing towards labour.) In the case of Canterbury, it was a race between the two main parties (so there is a large swath of data to explain what sort of person votes for each) in a constituency that had existed for several elections and where the swing can easily be attributed to the demographics. This is a perfect case of where MRP can work very well.
Here we have a constituency where most of it has come from a constituency where one of the two main options does not exist this time around, and where that option/its voters cannot easily be understood by a statistical model. Then the largest contribution from outside of that constituency (Pinhoe) has a completely different voting history to the rest. The Cranbrook/Clyst Honiton/West Clyst area in the constituency is rapidly growing and it is difficult to understand how those areas will vote (e.g. for an English town of its size, Cranbrook is quite a bit of an outlier for things like irreligion and average age). There are some constituencies where a good model can predict the outcome better than an educated punter (e.g. Canterbury in 2019). There are some where an educated punter can far outperform a statistical model (e.g. Islington North). Here is an example of a constituency where we can only speculate without ground knowledge/specific polling.
|
|
haroldthepolitician
Labour
'You can be just what you want to be, Just as long as you don't try and compete with me' - Jarvis
Posts: 216
|
Post by haroldthepolitician on Jun 17, 2024 17:41:16 GMT
I don't particularly agree with this, it's more a case that alot of MRP projections right now are incredibly lazy and just done for the publicity of it being an MRP. The only MRP projections I genuinely trust are Yougovs at the moment as they have a consistent track record both at being relatively accurate on overall outcome, but also being consistently accurate on individual seats. (Yougov had some strong predictions in 2019, like Canterbury staying/swing towards labour.) In the case of Canterbury, it was a race between the two main parties (so there is a large swath of data to explain what sort of person votes for each) in a constituency that had existed for several elections and where the swing can easily be attributed to the demographics. This is a perfect case of where MRP can work very well.
Here we have a constituency where most of it has come from a constituency where one of the two main options does not exist this time around, and where that option/its voters cannot easily be understood by a statistical model. Then the largest contribution from outside of that constituency (Pinhoe) has a completely different voting history to the rest. The Cranbrook/Clyst Honiton/West Clyst area in the constituency is rapidly growing and it is difficult to understand how those areas will vote (e.g. for an English town of its size, Cranbrook is quite a bit of an outlier for things like irreligion and average age). There are some constituencies where a good model can predict the outcome better than an educated punter (e.g. Canterbury in 2019). There are some where an educated punter can far outperform a statistical model (e.g. Islington North). Here is an example of a constituency where we can only speculate without ground knowledge/specific polling.
Oh yeah, it's simply not possible to model strength of individual candidates without specific polling - though I meant to say in most cases (where an independents support is negligible and their isnt a unusually strong party candidate) Yougov have typically done a good job at predicting voteshares in seats to a reasonable margin of error |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 26, 2024 7:52:29 GMT
Both Labour and the Lib Dems seem to be going hard here. The fact that neither are relenting will obviously be helpful for the Tories, who probably won't be able to take this seat for granted in future elections even when they're doing relatively well, given new developments in and around the Exeter part.
In any case it's difficult to see a large winning vote share here. Maybe something like 30-30-20-15 for the top 4 parties, with Reform in 4th whilst the top 3 is an exercise for the reader.
|
|
nyx
Non-Aligned
Posts: 1,034
|
Post by nyx on Jun 27, 2024 1:34:46 GMT
Both Labour and the Lib Dems seem to be going hard here. The fact that neither are relenting will obviously be helpful for the Tories, who probably won't be able to take this seat for granted in future elections even when they're doing relatively well, given new developments in and around the Exeter part. In any case it's difficult to see a large winning vote share here. Maybe something like 30-30-20-15 for the top 4 parties, with Reform in 4th whilst the top 3 is an exercise for the reader. Any evidence of a meaningful Labour campaign? I'd thought the Labour Party was focusing on Plymouth and relatively speaking ignoring this. |
|
|
|
Post by froome on Jun 27, 2024 7:12:48 GMT
Both Labour and the Lib Dems seem to be going hard here. The fact that neither are relenting will obviously be helpful for the Tories, who probably won't be able to take this seat for granted in future elections even when they're doing relatively well, given new developments in and around the Exeter part. In any case it's difficult to see a large winning vote share here. Maybe something like 30-30-20-15 for the top 4 parties, with Reform in 4th whilst the top 3 is an exercise for the reader. Any evidence of a meaningful Labour campaign? I'd thought the Labour Party was focusing on Plymouth and relatively speaking ignoring this. A constituency which has Exmouth as its main town and includes the environs of Exeter really ought to be fertile territory for Labour. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 27, 2024 7:38:06 GMT
Any evidence of a meaningful Labour campaign? I'd thought the Labour Party was focusing on Plymouth and relatively speaking ignoring this. A constituency which has Exmouth as its main town and includes the environs of Exeter really ought to be fertile territory for Labour. r.e. the previous comment: Ben Bradshaw's twitter reveals plenty of recent activity in the constituency with a sizeable number of activists.
I completely agree. I suspect if it weren't for the influence of Claire Wright, the Lib Dems may have wound down their campaign a little here. And it will have become a lot more true with new housing in Cranbrook and on the outskirts of Exeter: these are place where most of the residents are young, and a high proportion will be commuters who interact with colleagues in a firmly Labour city each day, only living outside of it as they have been priced out.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jun 28, 2024 15:02:38 GMT
I've had a Facebook advert from the Conservatives here (I don't live in the constituency but I'm also not surprised to get adverts for it). The only thing that was mentioned was that the Conservative candidate lives in the constituency, whilst the Labour candidate lives in Exeter and the Lib Dem and Reform candidates live in Honiton and Sidmouth. No mention of policies or potential PMs. Make of that what you will.
|
|
|
|
Post by monksfield on Jun 30, 2024 7:33:16 GMT
This is one of the most interesting betting propositions of the lot, very hard to read. The outcome could be a genuine three way marginal, snd because of Wright some of the MRAPs have been giving this to Reform who I can’t see being in contention.
I’ve gone for Labour in the poll.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jul 1, 2024 13:43:43 GMT
I've had a Facebook advert from the Conservatives here (I don't live in the constituency but I'm also not surprised to get adverts for it). The only thing that was mentioned was that the Conservative candidate lives in the constituency, whilst the Labour candidate lives in Exeter and the Lib Dem and Reform candidates live in Honiton and Sidmouth. No mention of policies or potential PMs. Make of that what you will. ...and now a second, from the Conservative candidate's Facebook page rather than the local party's page. Blue, but no mention of party outside of the imprint. Mentions the candidate's Marine background and three education related pledges. Nothing clearly aimed as a hit to other parties' policies. I suspect the Conservatives are not feeling too comfortable about this one, but not obvious whether it's team yellow or team red they're worried about. |
|
haroldthepolitician
Labour
'You can be just what you want to be, Just as long as you don't try and compete with me' - Jarvis
Posts: 216
|
Post by haroldthepolitician on Jul 1, 2024 19:07:56 GMT
 A very unpredictable seat indeed (Graphed above is the last month of MRP and other projections.) |
|
Sg1
Conservative
Posts: 1,084
|
Post by Sg1 on Jul 1, 2024 21:29:23 GMT
As things stand as of the time of my comment, the percentages in the constituency poll could plausibly be similar to the percentages on Thursday.
|
|
CatholicLeft
Labour
2032 posts until I was "accidentally" deleted.
Posts: 6,718
Member is Online
|
Post by CatholicLeft on Jul 4, 2024 17:56:45 GMT
The Reform Candidate was, in August of last year, convicted of violently kicking an elderly deaf dog (that subsequently died) - he claimed he was protecting his family and own dog and that the other dog had bad recall - well, being deaf, it would have. He was fined £2,000, banned from keeping an animal and sentenced to 200 hours community work.
Nice fella.
|
|
|
|
Post by observer on Jul 4, 2024 17:59:25 GMT
The Reform Candidate apparently was previously convicted of kicking an elderly deaf dog so badly, that it died - he claimed he was protecting his family and own dog and that the other dog had bad recall - well, being deaf, it would have. He was fined £2,000, banned from keeping an animal and sentenced to 200 hours community work. No Labour candidates ever shown bad character? |
|
