Ontario 2018 election

[King of Kensington]

Star snapshots DVW.

I think KW prevails because she'll be able to appeal to enough of the Blair Morrisons and Log McQuaigs.

[Pragmatic Conservative]

Pollara

PC 37% (+5)
NDP 37% (-6)
Liberal 20% (+3)
Green 5% (-)
Other 1% (-1)

https://www.pollara.com/wp-content/uploads/2017/12/Pollara_ONElxn2018-MethoDataBrf_May30-Jun2.pdf

[ON Progressive]

Pollara

PC 37% (+5)
NDP 37% (-6)
Liberal 20% (+3)
Green 5% (-)
Other 1% (-1)

https://www.pollara.com/wp-content/uploads/2017/12/Pollara_ONElxn2018-MethoDataBrf_May30-Jun2.pdf

I don't really know if this should be compared to their other polls. Their methedology has completely changed. It was all online last poll, and is a mixture of phone/online in this poll (1080 phone calls, 367 online surveys).

[EPG]

Are people more likely to refrain from telling pollsters their true intention if they are backing Doug Ford? I think so and I suspect PCs will do fine.

[adma]

Ugh. How many of those Liberal voters would've switched to the NDP had they not voted ahead of time? There was a lot of this in the 2011 federal election. Costed the NDP at least one seat in Quebec (Westmount) from what I recall.

One of the many reasons I would prefer people not vote ahead of time.

Though to be fair:  had the results shifted by a point and a half or so in three Toronto ridings in 2015, you could switch the party labels there.  (And advance polls made the difference in a number of Quebec NDP "saves" that year, too.)

[DL]

Are people more likely to refrain from telling pollsters their true intention if they are backing Doug Ford? I think so and I suspect PCs will do fine.

That’s an old theory . Nowadays rightwingers are very aggressively proud to support pigs like Ford

[King of Kensington]

Quito Maggi is promising a surprise result in Agincourt, is suggesting the Liberals still have strength in Humber River-Black Creek and that things have shifted in St. Paul's as well.

[Pragmatic Conservative]

Ugh. How many of those Liberal voters would've switched to the NDP had they not voted ahead of time? There was a lot of this in the 2011 federal election. Costed the NDP at least one seat in Quebec (Westmount) from what I recall.

One of the many reasons I would prefer people not vote ahead of time.

Though to be fair:  had the results shifted by a point and a half or so in three Toronto ridings in 2015, you could switch the party labels there.  (And advance polls made the difference in a number of Quebec NDP "saves" that year, too.)

How acurate traditionally are these polls of early voting anyway?
Abacus seems to have the Liberals on average 3% higher then other pollsters so its  possible their advanced voting is overestimating the Liberals.

I also imagine as advanced voting is a small portion of the electorate and only a small portion of their sample claimed to have voted early those early voting numbers would come with a huge margin of error.

[Hatman 🍁]

Quito Maggi is promising a surprise result in Agincourt

Knowing him, this probably means the Tories are ahead which would not actually be a surprise at all.

[King of Kensington]

Indeed, Agincourt is probably the first seat in Toronto to fall to the PCs (perhaps excluding Etobicoke North).

[Filuwaúrdjan]

Are people more likely to refrain from telling pollsters their true intention if they are backing Doug Ford? I think so and I suspect PCs will do fine.

That’s an old theory . Nowadays rightwingers are very aggressively proud to support pigs like Ford

Ford is hardly a break for the PCs anyway - Harris led them into three elections and o/c won two. Not everything is about or like Trump...

[Filuwaúrdjan]

How acurate traditionally are these polls of early voting anyway?

Polling subsamples are not particularly accurate in any country (and are not polls themselves and shouldn't be treated as such) and polling in Canada is amongst the least accurate in the industrialised world and has been for many decades. So to answer your question, what does two plus two equal?

[Filuwaúrdjan]

Pollara

PC 37% (+5)
NDP 37% (-6)
Liberal 20% (+3)
Green 5% (-)
Other 1% (-1)

https://www.pollara.com/wp-content/uploads/2017/12/Pollara_ONElxn2018-MethoDataBrf_May30-Jun2.pdf

I don't really know if this should be compared to their other polls. Their methedology has completely changed. It was all online last poll, and is a mixture of phone/online in this poll (1080 phone calls, 367 online surveys).

It's unusual for herding to be this blatant, lmao.

[ON Progressive]

How acurate traditionally are these polls of early voting anyway?

Polling subsamples are not particularly accurate in any country (and are not polls themselves and shouldn't be treated as such) and polling in Canada is amongst the least accurate in the industrialised world and has been for many decades. So to answer your question, what does two plus two equal?

Is Canadian polling actually that bad? They were pretty good in 2015.

[Hatman 🍁]

How acurate traditionally are these polls of early voting anyway?

Polling subsamples are not particularly accurate in any country (and are not polls themselves and shouldn't be treated as such) and polling in Canada is amongst the least accurate in the industrialised world and has been for many decades. So to answer your question, what does two plus two equal?

Is Canadian polling actually that bad? They were pretty good in 2015.

The problem isn't insomuch the polls themselves, but the volatility of the electorate, I think.

[mileslunn]

Final Abacus poll is out and shows no change at all from last week

NDP 37%
PCs 33%
OLP 23%

But they also find that when given a binary choice 60% prefer a Horwath NDP government vs 40% who prefer a Ford PC government and among Liberal voters that ratio rises to 80/20

Draw your own conclusions about how Wynne’s concession affects things

Very interesting stuff. The early vote is surprisingly strong for the Liberals too (34 PC, 30 NDP, 27 Liberal)

Ugh. How many of those Liberal voters would've switched to the NDP had they not voted ahead of time? There was a lot of this in the 2011 federal election. Costed the NDP at least one seat in Quebec (Westmount) from what I recall.

One of the many reasons I would prefer people not vote ahead of time.

True advanced polls are overused, but I find its mostly people who are diehard anyways and unlikely to change.  It's basically a way to ensure their vote is locked in case something comes up that prevents them from voting.  Also you need it as there are those with travel plans and its important those out of the province or country can vote too.

[King of Kensington]

Burlington is interesting.  I guess not being very Fordian, having a bit of an urban core and closeness to Hamilton = some NDP potential?

[mileslunn]

Are people more likely to refrain from telling pollsters their true intention if they are backing Doug Ford? I think so and I suspect PCs will do fine.

That’s an old theory . Nowadays rightwingers are very aggressively proud to support pigs like Ford

Depends what types.  If you look at the last US election, they nailed the percentage of non-college educated whites who voted for Trump, but underestimated the number of college educated whites who voted for Trump.  I think amongst uneducated voters, many are proud to support Ford, but amongst the more educated you might get some shy, but probably only makes a difference or a point of two.  You might also get some shy Liberals while as silly as this seems, I think you could get some shy NDP supporters amongst the wealthy since generally if you are making six figure salary it is seen as taboo to vote NDP just as it is seen as taboo to vote Ford if university educated, but in the end largely a wash.

[mileslunn]

How acurate traditionally are these polls of early voting anyway?

Polling subsamples are not particularly accurate in any country (and are not polls themselves and shouldn't be treated as such) and polling in Canada is amongst the least accurate in the industrialised world and has been for many decades. So to answer your question, what does two plus two equal?

Usually pretty good, but about 5 years ago they had some misses, BC 2013 being the most extreme example of a big miss.  But since 2015 they've generally been pretty good and it looks like a lot of them retooled to correct for the mistakes they made earlier.

Is Canadian polling actually that bad? They were pretty good in 2015.

[King of Kensington]

OK I see Guildwood and Agincourt now.  So Mitzie Hunter still has some support.  This is a tough call for the "strategic voter."

[adma]

How acurate traditionally are these polls of early voting anyway?
Abacus seems to have the Liberals on average 3% higher then other pollsters so its  possible their advanced voting is overestimating the Liberals.

I also imagine as advanced voting is a small portion of the electorate and only a small portion of their sample claimed to have voted early those early voting numbers would come with a huge margin of error.

OTOH, there is a tradition of right-of-centre overperformance and NDP underperformance in advance polls.  When NDP overperforms, it's either when the demos most prone to advance voting tend leftward (as in many cultural-class urban ridings) or the campaign momentum shifts dramatically away t/w the end of the campaign (as in 2015 federally).

Though given "Ford populism", it could well be that the PCs are, well, underperforming their usual advance-poll overperformance.

Keep in mind that if it were all about the advance polls, 1990 would have seen something like a 3-way seat split.

[Filuwaúrdjan]

The problem isn't insomuch the polls themselves, but the volatility of the electorate, I think.

Yeah, party loyalty and party identification are both unusually low in Canada, which creates very difficult conditions for polling firms. It's not that polling misfires totally all the time or even that regularly, but that even more than in most places you should always affix 'roughly' or 'approximately' before any figures.

[Filuwaúrdjan]

OTOH, there is a tradition of right-of-centre overperformance and NDP underperformance in advance polls.  When NDP overperforms, it's either when the demos most prone to advance voting tend leftward (as in many cultural-class urban ridings) or the campaign momentum shifts dramatically away t/w the end of the campaign (as in 2015 federally).

Though given "Ford populism", it could well be that the PCs are, well, underperforming their usual advance-poll overperformance.

Keep in mind that if it were all about the advance polls, 1990 would have seen something like a 3-way seat split.

Do they skew elderly as e.g. postal votes do in most places?

[Linus Van Pelt]

Presumably most countries don't have pollster CEO's who sh*tpost on Twitter like Quito Maggi or Darrell Bricker, at least.

[PeteB]

OK I see Guildwood and Agincourt now.  So Mitzie Hunter still has some support.  This is a tough call for the "strategic voter."

Not really. Unlike some other Scarborough ridings, in Guildwood the ABF (anyone but Ford) voter should support the Liberals.