1. Labour lost millions of voters between 97 and 2010. This is where most of the movement is.
2. Some of those lost voters went Labour to LD for leftie reasons so a lot of the LD -> Labour shift since 2010 is people who are *more leftie* than the average Labour vote.
Too much work, and not possible anyhow since the model was only invented in 2002. The model's autoregressive parameters would have been different for the earlier years, and I am not in a position to calculate them from scratch.
Then there is insufficient data to show whether the model is of any use in accurately predicting the future .
It's predicted every election correctly since 1945. But the current published model parameters include the 2010 result (previous versions were based on up to 2005, and 2001 respectively)
They use established statistical techniques of out-of-sample and one-step-ahead tests to see if the model "predicts" earlier elections correctly. It does.
It's a fairly simple model, based on current PM approval and "memory" of the previous two election results. Like sunspot activity, political fortunes follow cycles....
Too much work, and not possible anyhow since the model was only invented in 2002. The model's autoregressive parameters would have been different for the earlier years, and I am not in a position to calculate them from scratch.
Then there is insufficient data to show whether the model is of any use in accurately predicting the future .
It's predicted every election correctly since 1945. But the current published model parameters include the 2010 result (previous versions were based on up to 2005, and 2001 respectively)
They use established statistical techniques of out-of-sample and one-step-ahead tests to see if the model "predicts" earlier elections correctly. It does.
It's a fairly simple model, based on current PM approval and "memory" of the previous two election results. Like sunspot activity, political fortunes follow cycles....
But I want to see the evidence before I decide whether it is fact or fiction , A simple claim that it has predicted every GE correctly without the data to back the claim up is useless . We may as well consult sunspot activity . If the system has been in use since 1945 then the evidence should be there .
The polling on Lynton Crosby is actually very interesting, not for what it tells us about Lynton Crosby's visibility as for what it tells us about UKIP.
The knowledge that Conservatives, Labour supporters and Lib Dems have about Lynton Crosby is to all intents and purposes identical, not just in overall numbers but in every detail. They can be taken as showing a fair cross-section of how politically informed the general public is.
UKIP supporters are quite different. They are considerably less well-informed about who Lynton Crosby is. This is a telling indication that UKIP supporters are much less interested in political detail than the average.
Too much work, and not possible anyhow since the model was only invented in 2002. The model's autoregressive parameters would have been different for the earlier years, and I am not in a position to calculate them from scratch.
Then there is insufficient data to show whether the model is of any use in accurately predicting the future .
It's predicted every election correctly since 1945. But the current published model parameters include the 2010 result (previous versions were based on up to 2005, and 2001 respectively)
They use established statistical techniques of out-of-sample and one-step-ahead tests to see if the model "predicts" earlier elections correctly. It does.
It's a fairly simple model, based on current PM approval and "memory" of the previous two election results. Like sunspot activity, political fortunes follow cycles....
But I want to see the evidence before I decide whether it is fact or fiction , A simple claim that it has predicted every GE correctly without the data to back the claim up is useless . We may as well consult sunspot activity . If the system has been in use since 1945 then the evidence should be there .
Mark, their claims are correct and their academic papers back them up, and three months before the election they'll be solid again I imagine. But it's completely unrelated to what Rod is trying to do here, which of course he knows, that's why he hasn't gone back and done the calculations for two years out, and that's why I asked him for the 2008 "forecast"
Most of us could make a pretty good forecast of most GE results 3 months out without using quack formulae . There are some elections 1964 and Feb 1974 for example that an accurate forecast of the result 3 months before even to the largest party is purely luck .
It's from a peer reviewed scientific paper, has correctly predicted the past two elections (since it was published), and is actually now taught as part of a course in one of the top research universities in the world...
It's from a peer reviewed scientific paper, has correctly predicted the past two elections (since it was published), and is actually now taught as part of a course in one of the top research universities in the world...
My forecasts 3 months out from the last 2 GE's were also pretty much correct and IIRC more accurate than my eve of poll forecasts . Nevertheless I do not believe that any model could correctly have forecast the 1964 and Feb 1974 election results accurately without luck .
Perhaps you should address your advice to the originator of the stricture, yes?
I was addressing my "advice" to everyone who was/is throwing the observation around. You happened to be the most recent at the time of my commenting, so I apologise if by quoting you I made my comment appear partisan. That was not my intention.
If we do get a Lab/LD coalition after the next election I hope they make changing the voting system a priority. The question would be whether to hold a referendum or just do it through a parliamentary vote.
I'm getting worried again on England winning at lords ;-) to any pb cricket experts,what would be a good score for England to defend.
Aussies top order will come good in this series(they can't be that crap),pitch already turning but my nerves need easing from some of my fellow pb cricket experts ;-)
I'm getting worried again on England winning at lords ;-) to any pb cricket experts,what would be a good score for England to defend.
Aussies top order will come good in this series(they can't be that crap),pitch already turning but my nerves need easing from some of my fellow pb cricket experts ;-)
England could declare first thing in the morning and they'd have enough to win.
Perhaps you should address your advice to the originator of the stricture, yes?
I was addressing my "advice" to everyone who was/is throwing the observation around. You happened to be the most recent at the time of my commenting, so I apologise if by quoting you I made my comment appear partisan. That was not my intention.
I'm going to show my cricketing ignorance, but with Australia all over the shop today and obviously psychologically very wobbly, why didn't England enforce the follow-on?
It's from a peer reviewed scientific paper, has correctly predicted the past two elections (since it was published), and is actually now taught as part of a course in one of the top research universities in the world...
The way you are using it bears no relation to the model.
Yes, we understand, cracked record. The election is NOT YET.
Beyond that, you're not saying anything sensible...
I'm getting worried again on England winning at lords ;-) to any pb cricket experts,what would be a good score for England to defend
I saw a tweet earlier today saying that the highest 4th innings at Lords was 406 (by Australia)
Putting another 200 runs on the board will leave the Australians some hope. Another 250 with 2 days to go and there is only one winner.
The remaining England batsmen worry me,only bell we can count on(and he's got fail sometime) putting another 200 runs on could be well out of the reach of this lot,we could be seeing a repeat of trentbridge,just over 300 ;-)
I'm going to show my cricketing ignorance, but with Australia all over the shop today and obviously psychologically very wobbly, why didn't England enforce the follow-on?
Because wickets always deteriorate . It's almost never a good idea to enforce the follow-on.
I'm going to show my cricketing ignorance, but with Australia all over the shop today and obviously psychologically very wobbly, why didn't England enforce the follow-on?
Not showing ignorance at all; it's not only a good question but there's always some debate about using the follow-on as a tactic.
In England's case less so, because Cook is a cautious player and he tends to prefer batting another team out of the match instead of taking a risk and piling the pressure on the opposition batsmen. There may only be a handful of occasions where teams enforcing a follow-on have lost but they have acquired mythical status among captains.
In this case there are still three days left in the test match which is a huge amount of time. Teams that follow-on tend to do better in the second innings than they do in the first so it would only make sense to follow-on if there is a shortage of time. It's of the hottest days of the year so he's also giving his bowlers a bit of a rest. Getting Australia to bat in the fourth innings will suit England's bowlers more too, as the pitch will deteriorate and become more abrasive. This benefits England's best weapons - spin and reverse swing.
I'm going to show my cricketing ignorance, but with Australia all over the shop today and obviously psychologically very wobbly, why didn't England enforce the follow-on?
Questions like this are what makes Test cricket so interesting. It is the ultimate game-of-chess sport.
Cricket is a psychological-pressure sport. By not enforcing the follow-on Cook ensured that England get to play the rest of the Test without any serious pressure. Even if they make just 100 in this innings, they'll still be 340-odd in front and Australia will need to make a higher score than they've made so far in the series, batting last, to win. Extremely unlikely, as the pitch deteriorates and spins more.
If Australia had been inserted back in today, batted well, and gone into the lead sometime tomorrow afternoon/evening, with wickets in hand (not beyond the realms of possibility, because following-on always seems far less psychologically pressured than batting last) then England would've found themselves feeling pressured. Say Australia went 50 ahead, with a few wickets left, England would suddenly feel less happy about things. A position of seeming infallibility would feel more and more fallible with each passing run.
It is weird, but teams chasing low scores in the final innings, after being miles in front for large parts of the match, often collapse and lose.
The famous Australia/India Text from 2002ish, when Laxman made 281no when following-on, against an awesome Australian side, was a great case in point.
Comments
2. Some of those lost voters went Labour to LD for leftie reasons so a lot of the LD -> Labour shift since 2010 is people who are *more leftie* than the average Labour vote.
They use established statistical techniques of out-of-sample and one-step-ahead tests to see if the model "predicts" earlier elections correctly. It does.
It's a fairly simple model, based on current PM approval and "memory" of the previous two election results. Like sunspot activity, political fortunes follow cycles....
The knowledge that Conservatives, Labour supporters and Lib Dems have about Lynton Crosby is to all intents and purposes identical, not just in overall numbers but in every detail. They can be taken as showing a fair cross-section of how politically informed the general public is.
UKIP supporters are quite different. They are considerably less well-informed about who Lynton Crosby is. This is a telling indication that UKIP supporters are much less interested in political detail than the average.
You should move out of the 19th century and read this ;
http://www.ft.com/intl/cms/s/2/f1934ad6-ee7c-11e2-816e-00144feabdc0.html#axzz2ZWBmHyrD
You simply can't afford to do that if you're serious about providing a reliable service.
http://www.oddschecker.com/politics-and-election
Safari on Mac.
Just what the Pakistani body politic needs, a man cleared of electoral offences with a son cleared of trading fraud. I'm sure he'll clean up.
Once again, many thanks for the lay the draw tip on the 2nd test.
@MarkTyrrellUKIP
Labour MP loves Britain so much that he gives up British nationality for a job in another country
http://m.scotsman.com/news/politics/top-stories/mohammed-sarwar-to-take-pakistan-government-role-1-3007752 …
You seem to assume that all the "swing-back" will go to one partner and not the other.
Aussies top order will come good in this series(they can't be that crap),pitch already turning but my nerves need easing from some of my fellow pb cricket experts ;-)
Putting another 200 runs on the board will leave the Australians some hope. Another 250 with 2 days to go and there is only one winner.
Beyond that, you're not saying anything sensible...
https://twitter.com/hendopolis/status/358313089680166912/photo/1
In England's case less so, because Cook is a cautious player and he tends to prefer batting another team out of the match instead of taking a risk and piling the pressure on the opposition batsmen. There may only be a handful of occasions where teams enforcing a follow-on have lost but they have acquired mythical status among captains.
In this case there are still three days left in the test match which is a huge amount of time. Teams that follow-on tend to do better in the second innings than they do in the first so it would only make sense to follow-on if there is a shortage of time. It's of the hottest days of the year so he's also giving his bowlers a bit of a rest. Getting Australia to bat in the fourth innings will suit England's bowlers more too, as the pitch will deteriorate and become more abrasive. This benefits England's best weapons - spin and reverse swing.
Hope that helps.
And your prediction to help my nerves mr eagles ;-)
Mr eagles posted - England could declare first thing in the morning and they'd have enough to win.
It doesn't fill me with confidence ;-)
Cricket is a psychological-pressure sport. By not enforcing the follow-on Cook ensured that England get to play the rest of the Test without any serious pressure. Even if they make just 100 in this innings, they'll still be 340-odd in front and Australia will need to make a higher score than they've made so far in the series, batting last, to win. Extremely unlikely, as the pitch deteriorates and spins more.
If Australia had been inserted back in today, batted well, and gone into the lead sometime tomorrow afternoon/evening, with wickets in hand (not beyond the realms of possibility, because following-on always seems far less psychologically pressured than batting last) then England would've found themselves feeling pressured. Say Australia went 50 ahead, with a few wickets left, England would suddenly feel less happy about things. A position of seeming infallibility would feel more and more fallible with each passing run.
It is weird, but teams chasing low scores in the final innings, after being miles in front for large parts of the match, often collapse and lose.
The famous Australia/India Text from 2002ish, when Laxman made 281no when following-on, against an awesome Australian side, was a great case in point.