2015 Federal Election Prediction

[cybercoma]

That poll is an anomaly from the worst predictor of the last election. We'll have to wait and see if it's a trend, but I wouldn't put much stock in it yet.

[On Guard for Thee]

You're not a Conservative. That's exactly the kind of impression I'd expect you to come away with.

You certainly don't have to be of any particulat political stripe to come away with that impression. Ducking and dodging is ...ducking and dodging. Doesn't matter what color tie they have on.

[waldo]

That poll is an anomaly from the worst predictor of the last election. We'll have to wait and see if it's a trend, but I wouldn't put much stock in it yet.

outlierliar... stating the obvious. The shift is wholly predicated upon a ginormous change in Ontario... from what had been a fairly yearly constant aggregated polling difference of ~5% points favouring the Liberals over the CPC, to a now 17% point difference favouring the CPC... a "swing" of 22 points in Ontario. Clearly, this poll so excited Simple, he created a whole new thread... title emphasizing that "voters are starting to pay attention"! Clearly; especially in Ontario, hey Simple!

[waldo]

L OL

Election predictions 2 years out is just plain dumb.

it's now 15 months max... but nothing to stop Harper from pulling the pin earlier, particularly if, for example, the CPC Senate/Robocall Scandals begin to 'heat up'

[On Guard for Thee]

He'll go to the polls before he goes to court. So would I actually where I him.

[On Guard for Thee]

I really wanna hear what Nigel Wright has to say now he's got the RCMP off his back.

[Bryan]

That poll is an anomaly from the worst predictor of the last election. We'll have to wait and see if it's a trend, but I wouldn't put much stock in it yet.

Angus Reid was the most accurate:

Once Again, Angus Reid Provides Best Forecast of Canadian Election

[TORONTO – May 3, 2011] – Angus Reid Public Opinion, the public affairs practice of global research and technology firm Vision Critical®, has provided the most accurate forecast of a Canadian federal election for the second time in a row.

In 2008, Angus Reid’s final prediction of the Canadian federal ballot was the most precise. This year, the firm has made history by once again providing the most accurate forecast, with a calculation for the five main contending federal parties that differed from the actual vote returns by five percentage points.

https://www.visioncritical.com/sites/default/files/pdf/2011.05.03_Results_CAN.pdf

[waldo]

oh my Brian! Angus Reid telling you, in your bolded size increased lettering, that Angus Reid provided the "most accurate assessment". When you factor in the respective margins of error, Nanos Research had the most accurate prediction... Angus Reid failed to predict the NDP number within it's margin of error.

[Bryan]

oh my Waldo! Angus Reid included not only their data, but everyone else's on the last page of the press release. Nanos came in third.

[waldo]

oh my Waldo! Angus Reid included not only their data, but everyone else's on the last page of the press release. Nanos came in third.

try to keep up... again, as I stated, factor the pp margin of error. Again, when you do that, Nanos Research can lay claim to the best prediction. Or are you (and Angus Reid) stating their polls margin of error means nothing?

[Bryan]

You try to keep up…AR did better than Nanos, PERIOD.

AR had NDP at the exact number that Nanos did, it's their CPC numbers that were more accurate.

Margin of error is irrelevant, the ACTUAL numbers are what matters. If you're closer to the final vote, you're more accurate, FULL STOP.

[waldo]

You try to keep up…AR did better than Nanos, PERIOD.

AR had NDP at the exact number that Nanos did, it's their CPC numbers that were more accurate.

Margin of error is irrelevant, the ACTUAL numbers are what matters. If you're closer to the final vote, you're more accurate, FULL STOP.

look at the data from your own linked AR article reference! Notwithstanding all the numbers have been rounded, neither of your claims are accurate; neither of your claims that "AR had NDP at the exact number that Nanos did, it's their CPC numbers that were more accurate". Gee, I hope no one follows your lead and starts accusing you of dishonesty/lying.

if a poll can't meet it's expressed margin of error for each of it's respective party predictions, well.... FULL STOP! It seems you're quite willing to arrive at a final differential based on a tally of each respective party prediction number... but you're not willing to carry that over and also factor the polls expressed margin of error for each respective party prediction number. Go Figure!. How self-serving of you.... FULL STOP!

on edit: PERIOD!

Edited April 24, 2014 by waldo

[cybercoma]

Margin of error is irrelevant

Oh come on, Bryan. LOL

[Bryan]

You're both being really stupid.

Worrying about margin of error over whether the numbers are actually accurate is like saying "yeah, we lost, but we covered the spread." Margin of error is a useful metric for tracking where you went wrong. If your numbers are literally more accurate than the other guys', he did NOT predict better just because he covered his MOE.

[cybercoma]

Oh man. We're being really stupid? Keep it up, Bryan. Your posts are comedic gold right now. We're looking at predictions: margin of error is meaningless. haha. It's too much, man.

[WWWTT]

You're both being really stupid.

Worrying about margin of error over whether the numbers are actually accurate is like saying "yeah, we lost, but we covered the spread." Margin of error is a useful metric for tracking where you went wrong. If your numbers are literally more accurate than the other guys', he did NOT predict better just because he covered his MOE.

Man I find it funny that you clearly do not see that these polls only prove that there is a "swing vote" out there that jumps ship on the drop of a hat!

We saw this swing vote having an impact that gave Harper his majority in the first place!

WWWTT

[bleeding heart]

I still maintain that these polls are marginally telling at best.

I remember well the similar debates revolving around the last election...and as far as I remember, very, very few of us got it right. (Probably no one got the NDP success correct).

I'm not on a high horse here...I get the interest in the topic, and am even willing to play along.

But at bottom, we're just conjecturing about the outcome of True Detective or Breaking Bad before the final episodes were in.

[cybercoma]

The final polls before the election saw the NDP swing, particularly Nanos. What is certain from polling numbers in Canada is that campaigns matter. A lot of people decide or even change their minds during campaigns.

[The_Squid]

Anyone that thinks error margins are irrelevant really doesn't understand statistics!

[Keepitsimple]

Man I find it funny that you clearly do not see that these polls only prove that there is a "swing vote" out there that jumps ship on the drop of a hat!

We saw this swing vote having an impact that gave Harper his majority in the first place!

WWWTT

Excellent point......but for the past year it was looking like that swing vote was pretty firmly entrenched with the Liberals......but this poll (just one poll, I acknowledge) suggests that the swing vote is clearly back in play. The significance is that Trudeau now has to "win back" those votes by doing something. That means his handlers have to let him speak - and that presents big, big trouble. Voters are starting to pay attention.

[WWWTT]

I still maintain that these polls are marginally telling at best.

I remember well the similar debates revolving around the last election...and as far as I remember, very, very few of us got it right. (Probably no one got the NDP success correct).

I'm not on a high horse here...I get the interest in the topic, and am even willing to play along.

But at bottom, we're just conjecturing about the outcome of True Detective or Breaking Bad before the final episodes were in.

Man you have bad memory! Or is it selective memory? Take your pick.

The final polls before the election saw the NDP swing, particularly Nanos. What is certain from polling numbers in Canada is that campaigns matter. A lot of people decide or even change their minds during campaigns.

Very true! It was very clear that the NDP were going to have a real strong showing! And in Quebec! Layton's approval numbers were very strong!

It was actually the conservatives who made the biggest leap from the polls since no poll out there was giving them a majority, funny how conservatives forget that isn't it!

WWWTT

[waldo]

Excellent point......but for the past year it was looking like that swing vote was pretty firmly entrenched with the Liberals......but this poll (just one poll, I acknowledge) suggests that the swing vote is clearly back in play. The significance is that Trudeau now has to "win back" those votes by doing something. That means his handlers have to let him speak - and that presents big, big trouble. Voters are starting to pay attention.

just one poll... you acknowledge? Well, alrightee! Again, that poll presents a 22 point change in Ontario - that's quite the Ontario "swing vote"! Who knew?

[WWWTT]

Excellent point......but for the past year it was looking like that swing vote was pretty firmly entrenched with the Liberals......but this poll (just one poll, I acknowledge) suggests that the swing vote is clearly back in play. The significance is that Trudeau now has to "win back" those votes by doing something. That means his handlers have to let him speak - and that presents big, big trouble. Voters are starting to pay attention.

Actually that's not the usual conservative approach.

The conservatives use attack adds as their main focus in elections.

Payback is going to hurt come this election!

WWWTT

[waldo]

was that 'Orange Crush' made up of typical swing voters... or a rather one-time anomoly? Cause like, who crushed the crush?

[cybercoma]

Quebec's Orange Crush was probably an anomaly. I can't see them supporting Mulcair like they did Layton. And in any case, the problem for the Liberals was that Layton brutally destroyed Ignatieff in the campaign by pointing out how much time he was away from Parliament. It doesn't matter how much you F up in Parliament, people don't want to get the impression that you're not doing anything there. Layton's narrative was also helped by the Tories campaign saying Ignatieff doesn't give a crap about Canada. Well, not showing up to work proved it. So people went for the alternative and that was the NDP. Quebec this time will swing to Trudeau, where they've traditionally supported the Grits as their non-separatist party of choice. Stephen Harper seems completely out of touch with Québec politics. He'll continue to take a beating there and it doesn't matter. He has proven for the first time in history that you don't need Québec to win. With gerrymandering and the new seat counts, he'll ensure it.