Latest national poll median date: October 20
Projections reflect recent polling graciously made publicly available by pollsters and media organizations. I am not a pollster, and derive no income from this blog.

Monday, October 12, 2015

Nanos: Liberals Up in Ontario

Two notes to start:
- Why do projections disagree on who's ahead? See this post for an explanation.
- It's the last day of advanced polls, from noon to 8pm local time! For those of you interested in strategic voting (polls suggest ~30% of you are), see my guide for the advanced polls. It was written Thursday night, and does not take into account polls released subsequently.

Today's Nanos further diverges from EKOS: the Liberals' national lead increases slightly to 6.8%, but more importantly, it is now at 12.5% in Ontario. So who's right? Are the Liberals up by double digits in ON (Innovative 12%, Forum 15%)? Or is ON a tight race (Angus Reid tie, EKOS has Tories ahead by 0.5%)? The answer to this question determines who's the true current seat count leader.

The projection, of course, comes somewhere in the middle: it has the Liberals up by 6.1% in ON, or 2.8% adjusted for turnout.

CON - 142, -1 (34.1%, -0.1%)
LIB - 109, +1 (33.0%, +0.2%)
NDP - 80 (22.6%, -0.1%)
BQ - 6 (4.8%, +0.1%)
GRN - 1 (4.6%)

The unadjusted projection sees the Liberals within 9 seats of the Tories:

CON - 128 (32.4%)
LIB - 119 (33.9%)
NDP - 82 (23.2%)
BQ - 8 (4.7%)
GRN - 1 (4.8%)

Depending on which projection you believe, the Liberals have a 15-40% chance of leading the seat count in the past few days.

There is no EKOS update this afternoon due to Thanksgiving. (Happy Turkey Day!) Usually, Global comes out with an Ipsos poll at 4pm ET/1pm PT on Monday, but it's unclear whether that'll happen today due to the holiday. If it does, I will update the projection this afternoon. If it doesn't, I'll write the second part of the series on understanding projections, which explains where I get the above probabilities from and why I think other websites, with the notable exception of The Globe's Election Forecast, get probabilities/seat ranges wrong.

10 comments:

Anonymous said...

I think at least for the 905 area around Toronto, there are many immigrants who are uncomfortable with bills the conservatives have passed like c-24, c-51, or supervisa etc as Liberals could point to them and tell them, "I told you so" when they chose to give the conservatives power in 2011.

So I was almost sure the conservatives would lose a lot of seats among immigrant communities but that was until the provincial liberals brought out the new sex-ed curricula. Now, am not sure as the argument switches to a values debate and would voters there choose values over their interests, one would have to wait and see. I know you sometimes compare Kathleen Wynne to Harper and incumbency but there are pronounced differences because Wynne is really likable and even though you may not like her policies among those in the middle, one can speak with her. I mean, she actually met with folks who disagreed with her on the sex ed curricula. Harper rammed through c-24 and c-51 and anyone who opposed him on these, he labeled them as appeasers which is ridiculous. He doesnt talk to anyone besides the 30% and whether this may/may not work remains to be seen.

Summit said...

"Wynne is really likeable?" Get off the drugs.

Anonymous said...

"Wynne is really likable and even though you may not like her policies among those in the middle, one can speak with her. I mean, she actually met with folks who disagreed with her on the sex ed curricula"

On what planet did this happen???? Likable???
The main complaint from parents of ontario is she pushed it ahead without consultation and it was written by a know and convicted pedophile...

Election Watcher said...

"written by a know [sic] and convicted pedophile"

See Myth #3 here:
http://www.huffingtonpost.ca/karyn-pickles/ontario-sex-education-curriculum-update_b_6746012.html

The comment section of this blog is not meant for propaganda and grossly exaggerated claims.

Anonymous said...

And now back to talking about polls and polling: Nanos must have had a terrible set of polls for the CPC and beyond belief high figures for the Liberals on Thursday last week and on Thanksgiving Sunday. My calculations have his daily 400 figures (i.e., not the rolling 3 day 1200 figures) thus from a week ago (Monday) to last night (Sunday) (the rows are in order: CPC, Liberal, NDP, Green and BQ):

126 128.4 126.4 124 114.4 116 115.6
140 137.2 134 136 139.2 140.4 142.8
92.4 92 96.8 100 99.6 100 97.2
18.4 18.4 18.4 16.8 21.6 20.4 19.2
19.6 20.4 22.4 21.2 22 20.8 22.8

That translates into daily polling figures for 400 respondents each day as follows:

29.3% 36.4% 29.2% 27.5% 29.2% 30.4% 27.2%
37.5% 28.2% 34.8% 39.0% 30.6% 35.7% 40.8%
20.8% 24.9% 26.9% 23.2% 24.6% 27.2% 21.1%
3.8% 3.9% 6.1% 2.6% 7.5% 5.2% 1.7%
6.0% 4.6% 6.3% 5.1% 5.2% 5.4% 6.6%

Now maybe someone thinks that the Liberals were around 40% and had a 13% lead nationally two out of the last 4 days of polling, but I don't. I think the explanation is more likely what pollster in the US have studied and found about Republican voters and the skew that develops on "family" oriented days when conservative "family values" voters simply don't answer the phone or talk to pollsters. Voters who tend more to be younger and single and not in a family gathering are more apt to talk to a pollster.

So I discount Nanos' Thanksgiving weekend and the days leading up to it.

Election Watcher said...

What assumptions did you make to get these numbers?

I'm curious about tomorrow's Nanos, as I think Friday might have been a bad sample for the Tories: that's when they dropped from 31 to 28.6. That day gets dropped in tomorrow morning's poll, so we'll see.

But you're right that holiday weekend telephone polls are problematic; the Tories were indeed implausibly low during Labour Day weekend.

That's not a big problem for my projection. Unlike ThreeHundredEight.com, I don't put that much weight on Nanos. The most recent Nanos only gets a weight of around 9%.

Anonymous said...

I kept running different sets of numbers and clever formulae till I figured out what worked about 2 or 3 weeks ago. I chose Sept 21-23 as the numbers for the major parties stayed relatively stable then (meaning that the 3 day averages more than likely reflected the 1 day samples across the 3 - 3 day averages reported). Once that was established, Excel does the rest, adjusting by the difference in raw vote as the 4th day drops off and the new day rolls on. I concede that some minor variations from what Nanos actually got on Sept 21 would adjust (in a minor way) the series that ensues. But I don't believe those to be significant when viewed over time and when the focus is on how large the daily variations must be to get the Nanos 3 day figures.
If you (or anyone else) has a better way of divining the starting point 1 day figures, I'd be indebted to you for explaining it.

Election Watcher said...

Probably the best way to go about it would be to take any two days, and try a whole bunch of pairs of numbers for each party. Then pick the pair that generates the smallest oscillations in the time series, and pray that the numbers add up to something like 99% each day. Sounds like it could be quite a bit of work though, so your method seems reasonable...

Habs24cups said...

As a Conservative voter your polling numbers are about the only good news I've come across in the last 5 days or so . It seems to me that based on your numbers if Prime Minister Harper can win 48-50 seats in Ontario & stay within 10 seats of the Liberals in Quebec he should be able to overtake Trudeau when the returns come in from out west .Your projection from BC is very encouraging . What do you base your BC numbers on ? Do you have internal polling data from individual ridings? And if so can you post it ..... Thanks

Election Watcher said...

These are NOT polling numbers. They are projections based on publicly available polls. My unadjusted numbers are consistent with other models based on uniform swing, while my adjusted numbers assume a Conservative turnout advantage (that is smaller than last time). If the Tories have the same turnout advantage as in 2011, they'd do even better; if the turnout is actually representative of the population this time, the outcome will be close to the unadjusted projection.

I have zero internal data as I am not involved with any political party. The Tories are benefiting from a near perfect vote split in BC right now. If the Liberals go up, they'll take Tory seats in the Vancouver area; if the NDP goes up, it'll take Tory seats elsewhere in BC.