Canada 2011 Official Thread

[Teddy (IDS Legislator)]

You guys have the fortunate of knowing "Boragina" from the pundit's guide post

[Dan the Roman]

You missed the best part - EKOS's seat projection based on that poll.

Conservative 131 + Arthur, Liberal 62, NDP 100, Bloc 14

You must be logged in to read this quote.

EKOS queries the Green Party in their polls, which always inflate their votes and steals others. It also has a reputation for outlier polls, and Graves gets hit with being a "Liberal" pollster every now and then. It's not as if they haven't shown huge leads for the Conservatives before, though

Ekos for the last few weeks has been quite hackish. They have three times released mid-week "preview" polls with the Conservatives under-performing, usually in the 33-34%, and then release their final polls on Fridays, which almost always have the Tories back up to at least 36-37%.  Its clearly an effort to grab media attention and then backtrack so as to avoid being called on their numbers.

This looks par for course for an Ekos preview release. I am sure they will hedge their bets and we will get another Ekos before the end of the week with numbers like this:

Con 36
NDP 25
Liberals 24

[Teddy (IDS Legislator)]

Ekos has been polling for a very long time. They have a history of over-stating trends.

The Highest ever, and Lowest ever, for the Liberals, PC Party, Canadian Alliance, NDP, Reform Party, and Bloc Quebecois, have all come from Ekos at different times.

[Smid]

Teddy, it's got your final projection, presumably based on polling figures from the final week. How close were you if we plugged the actual results into your calculator?

[Teddy (IDS Legislator)]

Teddy, it's got your final projection, presumably based on polling figures from the final week. How close were you if we plugged the actual results into your calculator?

huh?

[Smid]

My assumption is that some of the seat variance between your prediction and the final result was due in part to a difference between the polls you used and the actual results on the day. If the on-the-day results were entered into your seat calculator, this probably would have improved your accuracy and therefore the accuracy of your model. Did you do any re-calculation following the election to show how many errors were actually caused by polling inaccuracies outside your control?

[Hatman 🍁]

There are so many issues with Teddy's map, I don't know where to start.

Vaudreuil-Soulanges voting Tory?
Saskatoon-Humboldt voting NDP before Rosetown-Biggar?
The Liberals picking up Churchill River?

[I spent the winter writing songs about getting better]

That law on election results publishing sounds even dumber than the laws some European countries have prohibiting publishing polls a certain amount of time before the election. I can kind of see the reason behind (I understand in the 80s West Coast candidates often complained about early calls for elections while the polls were still open in other states), but it's so ridiculously unenforceable today I'm amazed anyone takes it seriously.

[Smid]

I can understand broadcasting results on tv or radio, or publishing election results on general news websites to prevent people who might still be considering voting seeing results, unless they specifically seek those results out. I mean, I can understand supressing results in a general broadcast, so if you're watching tv news, you won't see the results, likewise radio. Also logging onto, say CNN's website wouldn't have a headline about the election results, however the Elections Canada website should have results, and if a media outlet registered a new web page specifically to report election results, that should be okay because anyone actively seeking election results is likely interested enough to either already have voted or to not be deterred from voting by simply looking up election results. 

[Teddy (IDS Legislator)]

There are so many issues with Teddy's map, I don't know where to start.

Vaudreuil-Soulanges voting Tory?
Saskatoon-Humboldt voting NDP before Rosetown-Biggar?
The Liberals picking up Churchill River?

The first is an error, the second two are correct on the math

[Teddy (IDS Legislator)]

It's a timing problem. People need to count the ballots, and such, and hence need to work. Work from the time the polls open to the time counting is complete. When it's 7am in Ontario, it's 6 in Winnipeg, 5 in Edmonton, and 4 in Vancouver. It makes it 8am in Halifax and 8:30 in St.John's.

People like the polls to close at 8pm. In Toronto, this would close the polls at 5pm in Vancouver, way to early.

8pm in Vancouver would be 11pm in Toronto, far too late.

9:30pm could be a compromise, and if Toronto, that would be 6:30pm in BC, which is a bit early, but could also be done. This would be 10:30pm in the Atlantic, but, that's a bit late.

If you had the polls close at a different time in the Atlantic, maybe, but then you'd need to like not-count the results, which gets too late for the workers, or just not release the results, which is possible.

[cinyc]

It's a timing problem. People need to count the ballots, and such, and hence need to work. Work from the time the polls open to the time counting is complete. When it's 7am in Ontario, it's 6 in Winnipeg, 5 in Edmonton, and 4 in Vancouver. It makes it 8am in Halifax and 8:30 in St.John's.

People like the polls to close at 8pm. In Toronto, this would close the polls at 5pm in Vancouver, way to early.

8pm in Vancouver would be 11pm in Toronto, far too late.

9:30pm could be a compromise, and if Toronto, that would be 6:30pm in BC, which is a bit early, but could also be done. This would be 10:30pm in the Atlantic, but, that's a bit late.

If you had the polls close at a different time in the Atlantic, maybe, but then you'd need to like not-count the results, which gets too late for the workers, or just not release the results, which is possible.

Polls in the Eastern, Central and Mountain Time Zones already simultaneously close at 9:30PM Eastern.  Those in the Pacific Time Zone close at 10:00PM Eastern.  Realistically, we're only talking about results from the Atlantic and Newfoundland Time Zones getting counted and released before the rest of the country.  Those areas close at 8:30PM local, which is 7 or 7:30PM Eastern.  And while those results may be interesting to us political junkies, they don't necessarily tell the rest of the country how the rest of the country will vote.  The Atlantic provinces are special in their unique ways.

So to me, the ban is totally unnecessary.  It's not like Ontario's results are going to be dumped at 7PM while polls are still open 3 hours longer in B.C.  Even if they were, the ban on putting info on Internet sites where folks have to actively seek out results is dumb.  The whole idea that those who want to know the results aren't going to vote because they know their party didn't fare well out East is extremely silly.

[Chancellor of the Duchy of Little Lever and Darcy Lever]

It's a timing problem. People need to count the ballots, and such, and hence need to work. Work from the time the polls open to the time counting is complete. When it's 7am in Ontario, it's 6 in Winnipeg, 5 in Edmonton, and 4 in Vancouver. It makes it 8am in Halifax and 8:30 in St.John's.

People like the polls to close at 8pm. In Toronto, this would close the polls at 5pm in Vancouver, way to early.

8pm in Vancouver would be 11pm in Toronto, far too late.

9:30pm could be a compromise, and if Toronto, that would be 6:30pm in BC, which is a bit early, but could also be done. This would be 10:30pm in the Atlantic, but, that's a bit late.

If you had the polls close at a different time in the Atlantic, maybe, but then you'd need to like not-count the results, which gets too late for the workers, or just not release the results, which is possible.

Not starting the count till all polls have closed seems like a good idea to me.  Fancy some British-style overnight counting?

[Teddy (IDS Legislator)]

Explain "British-style overnight counting"

[minionofmidas]

Counting the next morning, as some areas in Britain do.

[Verily]

There are so many issues with Teddy's map, I don't know where to start.

Vaudreuil-Soulanges voting Tory?
Saskatoon-Humboldt voting NDP before Rosetown-Biggar?
The Liberals picking up Churchill River?

The first is an error, the second two are correct on the math

Hmmm?

Saskatoon-Humboldt, 2008: CON: 53.8; NDP: 27.8
Saskatoon-Rosetown, 2008: CON: 45.4; NDP: 44.5

I fail to see how "the math" would have Humboldt fall first.

Same for Desnethe. 2008: CON: 46.7; LIB: 30.3

Are you really predicting a big swing from the Conservatives to the Liberals?

[Hatman 🍁]

There is a right wing independent running in Humboldt that may take votes away from the Tories.

[You kip if you want to...]

Quick question, is electoral reform a make-or-break policy for any NDP involvement in a coalition, as it is for the Liberals in the UK? Would a Liberal/NDP coalition with Bloc support mean likely electoral reform? I assume the Bloc and the Tories would be against a move to PR.

[Hatman 🍁]

Quick question, is electoral reform a make-or-break policy for any NDP involvement in a coalition, as it is for the Liberals in the UK? Would a Liberal/NDP coalition with Bloc support mean likely electoral reform? I assume the Bloc and the Tories would be against a move to PR.

I dont think such things have been discussed.

[Filuwaúrdjan]

The NDP support electoral reform - naturally, as it would benefit them - but they've never made it a signature issue.

[Insula Dei]

Wouldn't it be stupid for them to get rid of FPTP at a time where it would work in their favour if they are to replace the Liberals as the major leftwing party and could ultimately result in a NDP Majority government within a very reasonable amount of time? Yes, I'm being overtly optimistic here.

[Filuwaúrdjan]

Another EKOS poll, apparently. Looks all but identical: Con 34, NDP 28, Lib 24, BQ 6, Greenies 7 (figures rounded by me).

[cinyc]

Another EKOS poll, apparently. Looks all but identical: Con 34, NDP 28, Lib 24, BQ 6, Greenies 7 (figures rounded by me).

Was EKOS polling over the weekend?  Nanos didn't poll on Good Friday or Easter Monday (hence, no update this morning).  They probably shouldn't have polled Easter Sunday either - there are too many people away from home, potentially skewing results.

[Verily]

Wouldn't it be stupid for them to get rid of FPTP at a time where it would work in their favour if they are to replace the Liberals as the major leftwing party and could ultimately result in a NDP Majority government within a very reasonable amount of time? Yes, I'm being overtly optimistic here.

It would, tactically. The NDP is not really disadvantaged by FPTP except insofar as FPTP hurts smaller parties. Unlike the Lib Dems, the NDP is solidly on the left and thus not nearly as evenly distributed across the country. The Lib Dems could win more votes than Labour and still end up far back in seats, but the NDP would easily pass the Liberals in seats if they won more votes.

[You kip if you want to...]

Wouldn't it be stupid for them to get rid of FPTP at a time where it would work in their favour if they are to replace the Liberals as the major leftwing party and could ultimately result in a NDP Majority government within a very reasonable amount of time? Yes, I'm being overtly optimistic here.

It would, tactically. The NDP is not really disadvantaged by FPTP except insofar as FPTP hurts smaller parties. Unlike the Lib Dems, the NDP is solidly on the left and thus not nearly as evenly distributed across the country. The Lib Dems could win more votes than Labour and still end up far back in seats, but the NDP would easily pass the Liberals in seats if they won more votes.

I see. So, if these polls play out and the NDP finish second, it's likely that Jack Layton will become Leader of the Opposition, right?