Ontario 2011 (6th October)

[adma]

I'm concerned. How can a 6 point gain only result in 3 more seats. (The article says the NDP is down in Parkdale-High Park, yeah right!).

I'm suspicious of P-HP, too.  However, an only-3-seat pickup *could* be plausible if the gains are largely "bottom-feeding" in little-or-no-hope seats.  (Sort of like 2003, where the NDP lost official party status while gaining share, thanks to dead cat bounces in the 905 and wherever else.)

[Hash]

Going canvassing, it's always amusing to see how many people are idiots ready to be brainwashed into believing anything and then thinking they're absolutely correct and that everybody else is wrong. They talk about stuff like energy, HST or healthcare but have no clue why they hate it.

[Linus Van Pelt]

You must be logged in to read this quote.

Again, everyone's doubts about the riding polls are correct, but still, these "questionable" results are quite consistent with a pattern that shouldn't be surprising just observing the campaign and thinking about the odd overall numbers, that the parties are basically trading class/education groups.

[MaxQue]

I never understood what the problem with HST is.

[mileslunn]

I think the overall seat projection is not totally out of whack, but I question some of the individual seats.  In the case of the Kitchener ridings, it is probably true that Kitchener-Waterloo is the most Liberal leaning if you had a generic candidate, but I would be quite shocked if they actually picked this up.  I suspect the Liberals would have won this in 2003 and 2007 had Elizabeth Witmer not been candidate.  In addition, I think McGuinty's decision not to attend the Northern Ontario debates will help the NDP win in some of these ridings.  Another aspect is who does the undecided break.  At least federally since 2004, it seems the undecided has generally broken in favour of the governing party which is good news for the Liberals.  By the same token Tory supporters are more firm in their support and more likely to show up on E-day thus much like the federal election, NB election, and Toronto mayoral election it is possible the polls are underestimating Tory support.  It will be interesting to see what the polls after the debates are.  I think considering most want stability, I believe that a majority is more likely than a minority, but which party it will be is tough to say.

[Hatman 🍁]

Here's the overlay of the forum research survey into GTA ridings:

http://www.thestar.com/staticcontent/1058980

I must say, while some make sense, Im a little flabbergasted about York South Weston for example. But, that's good news for Ottawa Centre, as if some of the obvious NDP ridings in Toronto aren't NDP in this poll, then OC becomes more likely as one of those few pick ups

[Holmes]

A lot of close races... and Liberals come on top in every one? I'll just wait for the official results. Riding polling is too crazy and unreliable. I'd like to know what they have for Timmins-James Bay though. If it's close between Liberals and NDP, in the garbage it goes.

[mileslunn]

Interesting how much lower the PCs are provincially than what they got in many of the ridings in the last federal election.  Is this because Tim Hudak is not catching on, or Michael Ignatieff was far more disliked than McGuinty, or perhaps is a just a case of many sticking with the devil they know rather than the one they don't.  Off course we will have to see what the actual results are on election day as most polls going into the final days of the federal election put the federal Tories at 38-41% which is only slightly higher than what the PCs are at now provincially.  No poll after Good Friday put the Tories even within the margin of error of what they actually got in Ontario (44.4%).

[RogueBeaver]

Interesting how much lower the PCs are provincially than what they got in many of the ridings in the last federal election.  Is this because Tim Hudak is not catching on, or Michael Ignatieff was far more disliked than McGuinty, or perhaps is a just a case of many sticking with the devil they know rather than the one they don't.  Off course we will have to see what the actual results are on election day as most polls going into the final days of the federal election put the federal Tories at 38-41% which is only slightly higher than what the PCs are at now provincially.  No poll after Good Friday put the Tories even within the margin of error of what they actually got in Ontario (44.4%).

I think the lesson from May is that the polls will only clue in to the final trend at the last minute, and even then decently underestimate the winner's margin. OLP's ground game is much better than the LPC's, they have a seasoned veteran as their leader, and the opposition is making a series of unforced errors. McGuinty is as well- copying the Iggy strategy of saying the NDP is too left for one thing.

[DL]

I think you mean that McGuinty is copying the Ignatieff strategy of saying the NDP is too rightwing - with his absurd accusation that because both the NDP and PCs voted against the last Liberal budget (for opposite reasons) they are somehow identical! Of course the Libs then turn around and claim the NDP is too leftwing...I guess they can't make up their minds.

BTW: One more thing to mention about the Forum poll. Even though they surveyed something like 400 people in each riding. In each riding they only asked people how they would vote on the generic party vote. They did not read the names of the local candidates in each riding - and in many cases that can make a big difference.

[Hatman 🍁]

Here are some more riding results: http://www.bayshorebroadcasting.ca/news_item.php?NewsID=39064

Those races seem to make sense.

If anyone else can find anything, it would be much appreciated.

[Hatman 🍁]

A new Ontario projection! (trying to make sense of the Forum Research poll)

http://canadianelectionatlas.blogspot.com/2011/09/ontario-election-2011-prediction_26.html

[Teddy (IDS Legislator)]

Riding polling fails because it is based on telephone exchanges, and ridings cross those. Ridings even cross postal codes, with one code in particular being in Guildwood and Scar Centre. I even argued with a political person on the phone telling them that I KNOW I am in Guildwood, because the rail line is the border, and they trying to tell me the opposite because of my postal code. That is why Parkdale, and the Kitcheners, etc, are showing up wrong. The poll is correct, but placing the voters in the proper riding is not.

[Hatman 🍁]

Riding polling fails because it is based on telephone exchanges, and ridings cross those. Ridings even cross postal codes, with one code in particular being in Guildwood and Scar Centre. I even argued with a political person on the phone telling them that I KNOW I am in Guildwood, because the rail line is the border, and they trying to tell me the opposite because of my postal code. That is why Parkdale, and the Kitcheners, etc, are showing up wrong. The poll is correct, but placing the voters in the proper riding is not.

Good point Teddy. Maybe a lot of the Parkdale people were counted in Davenport (same with York South-Weston)

[lilTommy]

With all the problems with these and other polls (numbers too small to get a real idea, or no reagional breakdowns, etc) they can be used to the advantage of some campaigns.
So they say the NDP is trailing in YSW, PHP (really?) at any rate, that can bolster volunteers; the last thing you want it to be over confident and think you have it in the bag. In Toronto Danforth, i've gotten 3 calls already to volunteer (which i have, and since i already voted might do eday too) so they are not taking a safe riding for granted. I suppose these poor numbers (seats counts really) can be used to bolster our campaigns, get people motivated to work for every riding.
This has also lead to speculation, should we even have this much polling? or polling during an election at all? Look at Manitoba? has there even been a poll since June? would that be better for the parties to just run a campaign and let the chips fall on election day? (ok sorry if you all want to kill me cause i love the polling everyone is doing!)

[Filuwaúrdjan]

Constituency polling is a mugs game; as we saw just a few months ago. But, presumably, these are breakdowns from a massive poll, rather than specific riding polls? If so, then that would make them near useless.

[Teddy (IDS Legislator)]

301 has been the standard for polling for some time over here. Dont ask me why. A poll of this size is large enough to put 301 voters in every riding - so I presume it is indeed individual riding based polls. For the reasons I outlined, though, I wouldn't trust the poll to be 100% accurate.

[DL]

Yes, but we have no idea if the results were demographically weighted in each individual riding (as opposed to at the province wide level) and we know that no local candidate names were used - just generic party vote.

[Hatman 🍁]

With all the problems with these and other polls (numbers too small to get a real idea, or no reagional breakdowns, etc) they can be used to the advantage of some campaigns.
So they say the NDP is trailing in YSW, PHP (really?) at any rate, that can bolster volunteers; the last thing you want it to be over confident and think you have it in the bag. In Toronto Danforth, i've gotten 3 calls already to volunteer (which i have, and since i already voted might do eday too) so they are not taking a safe riding for granted. I suppose these poor numbers (seats counts really) can be used to bolster our campaigns, get people motivated to work for every riding.
This has also lead to speculation, should we even have this much polling? or polling during an election at all? Look at Manitoba? has there even been a poll since June? would that be better for the parties to just run a campaign and let the chips fall on election day? (ok sorry if you all want to kill me cause i love the polling everyone is doing!)

Without polling, the NDP does not form opposition I think.

Constituency polling is a mugs game; as we saw just a few months ago. But, presumably, these are breakdowns from a massive poll, rather than specific riding polls? If so, then that would make them near useless.

Breakdowns from the massive poll. But as we saw from the GTA map, most of the ridings make sense. It's a few odds one that make you scratch the head. But then again, even actual election results will make you scratch your head from time to time.

[lilTommy]

Without polling, the NDP does not form opposition I think.

Huh? can you explain that a little more

[Insula Dei]

Without polling, the NDP does not form opposition I think.

Huh? can you explain that a little more

Opinion polls play an important part in the narrative of an election campaign and can help to creat an upwards trend for a party. I suppose that's what Hatman's getting at.

[lilTommy]

Just as i talk about gettign ride of polling during elections (which i rather like so i was being devils advocate)...
Two new polls:

Ekos- Lib 34.9%, PC 31.4%, NDP 24.7%, Green 6.3%

http://www.ekos.com/admin/articles/FG-2011-09-27.pdf

Abacus- PC 37%, Lib 33%, NDP 23%, Green 6%

http://abacusdata.ca/wp-content/uploads/2011/09/Ballot-Ontario-Sept-26-2011.pdf

[Insula Dei]

Ontario might well be heading for the best possible result (a Liberal minority government) on these figures

[Hash]

If Abacus, the trashy pollster for the ultra-trasy far-right SUN rags is showing a 4pt edge for the PC, then it's good news

From my internal sources, the Liberals have a 6pt lead or so in Ottawa-Orleans and we're not on the party's watchlist

[Antonio the Sixth]

Just as i talk about gettign ride of polling during elections (which i rather like so i was being devils advocate)...
Two new polls:

Ekos- Lib 34.9%, PC 31.4%, NDP 24.7%, Green 6.3%

http://www.ekos.com/admin/articles/FG-2011-09-27.pdf

Abacus- PC 37%, Lib 33%, NDP 23%, Green 6%

http://abacusdata.ca/wp-content/uploads/2011/09/Ballot-Ontario-Sept-26-2011.pdf

Is such an enormous variation frequent in Canadian/Ontarian polls ?