I notice that the CBC's Canada web page headline currently is: "NDP gaining support, poll suggests". What the CBC is trying to hide here is that to the degree the vote splits on the left, the odds are the Conservatives will do better than as indicated by Ve. If this situation holds, the Conservatives will win a majority. It must be because of all those gaffes we've heard about, and the horrors of not pandering to the artismists, eh what?

We can't get complacent. As evidenced by this article it would appear those crafty liberals have come up with a bold new tactic which I must say I would never have foreseen -- groveling and pleading for mercy:
http://start.shaw.ca/start/enCA/News/NationalNewsArticle.htm?src=n092715A.xml

MY reading of the polls is about as basic as it gets, take a CBC poll for instance, chop about 20% off the frontrunner then give about 60% of that to second place, on almost any subject, and you will be pretty close.

The CBC does not conduct polls in federal elections, therefore your
technique is useless here, even if it has no value in other cases.

Thanks for your efforts in producing the tables and graphs, Vitruvius. I am not yet sold on the idea of the Conservatives getting a majority, but I agree with your point that further splitting on the left will lower the threshold by which that majority is made. Now, on the other hand, this cautious approach of mine does have the benefit of making surprises somewhat pleasant.

I was just about to write a nastygram asking for an explanation and then thought, I bet he explains this somewhere ;)

Cool idea, but could the graphs be a tad larger? By the time you get seven graphs on a single plot it can be hard to see the forrest fer da trees.

As long as one does not assume that green line refers to the Green Party (like I did at first glance), there is no need to call an ambulance.

Barely more than two weeks to go.

Wednesday the debates happen. It seems highly unlikely that SD will score a win there and SH should do fine. Between then and now Jack may have some heavy slogging with the new focus on his sympathies for the truthers and he could take some nasty hits on it in the debate. Should he weather them and come off as half-sane (as opposed to the regularly anti-american socialist insane) he could navigate the ND to a second place spot in the polls and at the polls on the 14th.

There is already talk that the LP may start throttling back on spending and letting the party go down (not likely to happen - there are too many of the caucus higher-ups - iggy being one of them - who may lose their seat if they pull the plug) and that is a sign that they are already deep into the red. Their campaign is about to fall off the radar save the network newscast coverage. The collapse within the 416 will see all of the "star" liberal candidates staying close to home trying to save their own skin rather than the more marginal 905 seats.

The BQ has followed the arts issue down the rathole of trivial issues. Look for it to be a non-issue by the time the french debate ends. And then the support for the BQ will bleed away to the CP - not the NDs as they have very little organizational strength on the ground in QC - putting the CP firmly in Majority territory by the dawn of the 14th.

You're welcome, Brent. I agree with you on your cautious, or conservative approach, that's why I said "If this situation holds...". I don't do predictions; I'm just interested in studying the relationships between Ve, its history in previous elections since 1962, and the cross-pollster trends in this election's tracking polls.

AtlanticJim: The graph image is a link, pick it for a larger view.

Louise: It's a graph of Ve which is computed from:

        ( C - L ) × ( C + L ) ÷ 100

It is not a graph of any raw poll data for any party.

Did i piss someone off?

Just post the link, Glenn, not the run-on ill-formated data.

I am well aware of CBC the point thank you, I was referring to almost anything the CBC pushes in reference to polls they quote and my method is anything but ‘scientific” but it works for me.

OK

Just post the link, Glenn, not the run-on ill-formatted data.

Couldn't you just have said Hey Glen just post a link saves space ,thanks, instead you come out with a put down. You come across as a condescending know it all and that will drive nubees to the site away. That doesn’t help Kate at all.

Point taken, Western Canadian, sorry about that.

Thanks you for the acknowledgment Vit.

I must also say I throughly enjoy everything you bring to SDA, Thanks. Keep up the good work. Now I had better get to work, don't want to miss my tee time.

Vitruvius:

I don't do predictions

There is some wisdom in this. Yet, at the dual risk of being wrong and of introducing a somewhat off-topic (but related) idea into this thread, I will raise an idea. Let us suppose for this exercise that the Conservatives achieve a mild majority, say, 165 seats.

The idea is to ask the question: what will be left of the Liberal Party outside Toronto, Montreal, and Vancouver after this election is complete? Let me first acknowledge that to ignore the combined political clout of Canada's 3 largest cities (approximately 25% of the nation's population) is not my intention: I mean to simply identity the vote split along its partitions.

The reason for this partitioning exercise hearkens back to the early days of Reform. Those of us who were early members were often disheartened by the oft-repeated claims that Reform was never a national party because it could not gain seats east of the Ontario-Quebec border and could not gain many seats east of the Manitoba-Ontario border. An equivalent claim on the Liberal party (if my assumption is correct) is that the Liberal party is not a national party because it cannot gain significant seats outside Canada's 3 largest metropolitan areas.

Now, I am under no illusion that the MSM would make such a claim even if the data supported such a claim. Yet, it is a theme that I would be pleased to see discussed here at SDA. I realize that I am slightly ahead of the curve and it is not wise to count one's chickens before they hatch. Perhaps, we could return to this idea once the election is over - assuming, of course, that the election results supports my thesis.

I have to go run some errands with my wife. I will look for your (plural: Vitruvius, et.al.) responses later. Thanks.

I've updated my Vitruvius's Experimental Election Predictor essay to include the last week's data, and I added a few paragraphs to provide some explanation of the impetus behind Ve. For those who are interested in this experiment, the new paragraphs read as follows:

The idea behind Ve is straightforward. If there were only two parties, C and L, then the ( C + L ) ÷ 100 term is equal to one, and Ve simplifies to: ( C - L ). That's a pretty obvious measure of who's going to win (there are no minority governments if there are only two parties), as long as Ve isn't too close to zero (in which case per-riding splits become dominant).

With more than two parties, the certainty of Ve decreases, so the × ( C + L ) ÷ 100 term reduces the value of Ve toward zero, according to the percentage of votes not going to C or L, in which case the "close to zero" argument in the previous paragraph kicks in. I should also note here that it may be the case that in the remaining two weeks before the current election we will see the NDP poll higher than the Liberals in some polls. In those cases Ve will be calculated with the L variable equal to the NDP value. Just think of L as "second place", not Liberal.

The real value in reducing the per-party poll results into a single Ve value is that then allows us to plot the Ve values, against time, for multiple pollsters, and still have the graph come out simple enough to parse. It's the gross trends that Ve exposes best, and those are best studied across multiple pollsters.

The interesting thing about Ve in Canadian federal elections is that if we look at the actual election results since 1962 (as shown in the graphs and tables later in this essay), then, at least historically speaking, if Ve is greater than 8 we have a Conservative majority, if it's between 0 and 8 we have a Conservative minority, if it's between 0 and -4 it's indeterminate (one Conservative minority, one Liberal minority), if it's between -4 and -8 we have a Liberal minority, and if it's less than -8 we have a Liberal majority.

Brent Weston - that's an interesting concept - a political party focused only in the three major cities.

I think that is the reality of the Liberal Party which has over the past several decades moved into an ideology, as I keep ranting, based around class. A Canada that is made up of two classes: the Rulers and the Ruled.

The Rulers consist of the professionals, who are primarily funded by government: civil service, judiciary, legal, health care, education. The Liberals increased the role of government in all these areas - excluding private roles in health care, in education, etc. The Liberals also moved into and aligned themselves with financial institutions.

This large role of government in the economic, educational, health care, systems of Canada..set up a class of people embedded in that govt. And these roles are authoritative roles; the rest of us are dependent on their professional expertise since private education, health, etc, are not allowed or promoted.

These people are the Big City Liberals. They view the rest of us as..well..irrelevant. They tell us what to do; we are 'beer and popcorn'.

Then, these three big cities are also the focus of most of our immigration. The Liberals deliberately set up immigration within Canada as a Voting Bloc. Newcomers are not encouraged to integrate but to remain isolated and financially dependent on Liberal funding to maintain their isolation (cultural centres, schools, welfare).
They vote to retain that Liberal largesse.

Then, the economy of Canada has been, as set up by the decades of Liberal rule, an economy that is dependent on the US consumer and on foreign investment. Canada's taxes are so high that we can't develop an investor class to build our own industries. And, we don't compete in the world market. That means that we don't develop a robust small and medium size business class.

The Conservative party focuses on this small and medium size business class - the entrepreneur who will take risks (the professionals and immigrants don't take risks). With the emergence of the West, and even, the growing population in small towns, there are more members in this class. The Liberals have totally ignored this part of the economy.

So- the Liberals are the party of the big city govt funded professionals. And balkanized immigrants. They aren't the party that enables an emerging dynamic middle class.

And what about the Bloc? That's another of my rants; I maintain that it has no right to sit in the federal parliament. No right. What if each and every province had their own Default Party? Why should Quebec have its Default Party?

Topic? What topic? Why do I even bother reading those cyan
words above the Post a Comment widgets shown below?

vitruvius - calm down. Sometimes one has to 'allow things to happen'..and that includes thoughts. Allow your thread to generate thoughts about...other things.

After all, there is little that one can say about your graphs - as data about the Relation Between C and L - other than to acknowledge their reality.

Then, given this reality, and no-one is disputing this, one might possibly move on to consider WHAT is happening to set up such a C-L Relationship - and that's where one explores cities/rural; class; etc. And WHY it is happening.

What's wrong with those further quests?

The topic here is Ve Update; the topic is not On the
Philosophy of the Structure of Canada's Ruling Systems.

I was thinking of replying to Vitruvious direclty.

This is a work of inspired genius. Does it hold water? That remains to be seen.

Vitruvius,

"I should also note here that it may be the case that in the remaining two weeks before the current election we will see the NDP poll higher than the Liberals in some polls. In those cases Ve will be calculated with the L variable equal to the NDP value. Just think of L as "second place", not Liberal."

I think you'll get wrong results if you do that. I get the following average % of seats per percent of vote (since 1962):

Liberals - 1.16
Conservative/Reform/PC's - 1.07 (1.09 PCs only ... 1.02 Conservative/reform etc)
NDP - .53
Bloc - 1.38

A liberal % is worth twice as many seats as a NDP % ... probably has something to do with voter concentration.

The interesting thing about Ve in Canadian federal elections is that if we look at the actual election results since 1962 (as shown in the graphs and tables later in this essay), then, at least historically speaking, if Ve is greater than 8 we have a Conservative majority, if it's between 8 and 0 we have a Conservative minority, if it's between 0 and -4 it's indeterminate (one Conservative minority, one Liberal minority), if it's between -4 and -8 we have a Liberal minority, and if it's less than -8 we have a Liberal majority.

There are two interesting things about this. The first is the lack of symmetry about these "breakout" levels. The second is one of the data points violates the rules.

Let us consider the symmetry issue. We do not have enough data points to know the levels with any degree of certainty. For instance, there are no data points greater than 0 and less than or equal to 4.0 . Therefore I suggest we can force the symmetry for now until we have more data. That is to say: if it's between 8 and 0 we have a Conservative minority becomes if it's between 8 and 4 we have a Conservative minority and if it's between 0 and -4 it's indeterminate becomes if it's between 4 and -4 it's indeterminate. We really have no data that forces us to choose the first pair of rules over the second pair (or the second pair over the first) but choosing the first pair introduces the symmetry problem before we have even determined that we have to solve it.

The second issue is more problematic. The election of 1974 yielded 141 LPC seats in a Parliament of 264 seats - a clear majority. That majority with its corresponding Ve of -6.7 violates the following rule: if it's between -4 and -8 we have a Liberal minority. Hmmmm.

ET:
Thanks for your thoughts. In consideration of house rules, I will postpone follow up until another thread.

I appreciate the anomaly at -6.7, Brent, still,
that's a long way from somewhere around +9.

Well, I went and copied the wrong Ve value. The -6.7 is from a different data point. It should be -6.1 . The principle remains unchanged, however.

As you hint, Vitruvius, you can work out the bugs later; the main point of the current Ve is suggesting Conservative majority if the polls accurately reflect the real poll on election day.

Agreed.