Opinion Poll Bonanza!

[Filuwaúrdjan]

The MORI poll you've got up is "all certain to vote". Their normal poll is: Lab 38, Con 33, LD 23. So if we put both MORI polls together ('cos I don't trust either on their own)...

Lab: 38 to 35
Con: 39 to 33
LD: 23 to 21

The NOP poll is:

Lab: 36
Con: 33
LD: 21

[Filuwaúrdjan]

Most changes are well within the MoE and generally fit into the current pattern. Having said that there does seem to have been a slight shift towards the Tories (at the expense of both parties) but they still aren't polling where they were mid term and even the MORI certain poll gives Labour a lead in seats with uniform swing.

My guess for a majority is somewhere between 80 and 60

[Filuwaúrdjan]

Roughly how much of a nation-wide margin does Tories need to win the election? It's quite much isn't it?

To win an overall majority of one it's something like 8pts (might be 9... not entirely sure)...
To give an example of how hard it is for the Tories to win a majority if Labour hasn't collapsed, in 1992 the Tories won a tiny majority but lead the PV by a wide margin. If very few votes had switched in a handful of constituencies, Labour could have won and been behind in the PV by something like 7pts...

Reason for this is increased regional polarisation, something that began with Harold Wilson's narrow win in '64 (when Labour failed to win a load of traditional margins in the South, but won previously safe Tory seats in the North), and the fact that turnout in middle class areas is (nowadays) always higher than more working class areas.
The results look distorted and to a certain extent they are, but not as much as it looks so to speak.

[Filuwaúrdjan]

Yes and no. Yes in that PV and Seat numbers are never, ever close to each other, no in that each vote counts more, results are more local and there's less chance of dodgy backroom coalition deals.

Depends how you look at it really. Personally I prefer the Aussie system