For my final projection, click here.
Just some thoughts.
I still think the NDP will outperform expectations, even my own projection, but I have 0 way of proving that with the math.
Additionally, I am hopeful that we (people who do election projections) can find ways to pool resources to help us have access to, for example, riding polls.
I've also decided on a system to help me personally afford these.
Views are always important, as is data and resources. A PEI election will get less views on the blog than a Quebec election will, and, there will be less resources (IE coverage) of a PEI election to begin with. As such, in the 6 "smaller" provinces, even if riding polls do exist, using them is of less use.
In terms of the "big 4" provinces, while I do quite like BC and Alberta politics, the fact is simply that with limited resources (IE my money) I have to make tough decisions, and as such, I will not be getting any riding polls for either of these provinces even if they are available.
Quebec, while always entertaining, is French. Language. Is. Important. As such, with limited understanding, and with a simply "different" political culture, such polls would not be of great use to me.
This leaves us with two. Ontario, and Canada. Traditionally, they each hold elections every 4 years, but even if minority governments half that, this is still only one such election a year. I can afford $60 for these, once a year.
As such, I think this will help solve that issue.
Thursday, June 7, 2018
Wednesday, June 6, 2018
Final Projection: Ontario
I present to you, for this election, my final projection
This has honestly probably been the single most agonized over decision with regard to a final projection that I've had to make. All the data tells me that people want a majority PC government, lead by Andrea Horwath.
There's actually a lot less I have to say on this than I otherwise thought. People don't want Doug Ford, but they do want the PC Party. The problem is that this sort of thing has occurred before, and, generally, what happens is people vote for the party and not for the leader. This is not always the case, and the leader can drive the vote, but I'm not seeing that being enough here.
The other problem is that there is a lot of evidence that many people don't trust the PC Party, and, they do generally trust the NDP with another chance to govern.
In short what we are seeing happen is just enough Liberals going NDP to give Doug Ford a government, without enough Liberals staying with that party to block such a government. The NDP would need either 5 more points from the Liberals for their own Majority, or, 5 less points from the Liberals for a minority (as, many of the Liberal seats are ones that would be won against the Tories, not the NDP)
That being said, I am unconvinced that the PC lead is as strong as many people think. Ford is weak. Additionally, most of the polls showing the PC party far ahead have very uneven weightings for age. In short; I am unconvinced the people these polls are reaching are actually representative of the voting population.
Despite that, I can not ignore the overwhelming evidence that the PC party is doing well, and is likely to win tomorrow evening.
However, again, the evidence that the NDP will outperform expectations. When they ask people how they are going to vote, the same number of people say they are voting PC as say Ford is the best premier. Less people are undecided on the best premier than on the vote. To explain, look at these fictionalized numbers I've provided as an example.
1000 people polled
300 are voting PC
250 NDP
150 Liberal
300 are undecided
Same 1000 people also polled
350 think Horwath is best premier
300 think Ford is
150 think Wynne is
200 are undecided
What this tells me is the NDP will likely take a larger than expected share of the undecided voters.
Regardless, the PC lead is significant, and as such, my final projection, is as follows:
SPREADSHEETS
60 PC 37.86%
55 NDP 36.80%
8 LIB 20.89%
1 GRN 4.45%
Ford, and Wynne, both will be (re)-elected in their seats.
Mississauga will be where the election is decided. Had Ford won all of Mississauga he would have 62 seats, enough for a barebones majority, especially if he can convince an opposition MPP to become speaker. Had the NDP won the 3 seats they will be close in, they would have finished head of Ford. Mississauga really ends up becoming the city that decided the winner, and that winner, is Doug Ford.
Regardless, with only 60 seats, there remains a chance that a NDP-OLP coalition will take office instead. The two parties would need to agree to a solid "deal" before this happens, no informal deal will do, not with a 5 seat lead in the legislature. This means either a Peterson-Rae style "accord" or an outright coalition. Given the way the election has played out, this could well be easily possible.
Do not ignore the Greens either. They may offer Ford a way out, carbon taxes bring in a lot of revenue, revenue he will need to pay for his tax cuts. He could make the excuse that he requires Green support, and thus, needs to delete the carbon tax "later" and not now. Consider that if I'm off by even a single seat, anywhere, all of the math could be very different. A 61 seat PC party and a 1 seat Green party could form a majority, with an opposition speaker.
In the end there are two winners. Andrea Horwath, and the PC Party. It only remains to be seen whose win is bigger.
There's actually a lot less I have to say on this than I otherwise thought. People don't want Doug Ford, but they do want the PC Party. The problem is that this sort of thing has occurred before, and, generally, what happens is people vote for the party and not for the leader. This is not always the case, and the leader can drive the vote, but I'm not seeing that being enough here.
The other problem is that there is a lot of evidence that many people don't trust the PC Party, and, they do generally trust the NDP with another chance to govern.
In short what we are seeing happen is just enough Liberals going NDP to give Doug Ford a government, without enough Liberals staying with that party to block such a government. The NDP would need either 5 more points from the Liberals for their own Majority, or, 5 less points from the Liberals for a minority (as, many of the Liberal seats are ones that would be won against the Tories, not the NDP)
That being said, I am unconvinced that the PC lead is as strong as many people think. Ford is weak. Additionally, most of the polls showing the PC party far ahead have very uneven weightings for age. In short; I am unconvinced the people these polls are reaching are actually representative of the voting population.
Despite that, I can not ignore the overwhelming evidence that the PC party is doing well, and is likely to win tomorrow evening.
However, again, the evidence that the NDP will outperform expectations. When they ask people how they are going to vote, the same number of people say they are voting PC as say Ford is the best premier. Less people are undecided on the best premier than on the vote. To explain, look at these fictionalized numbers I've provided as an example.
1000 people polled
300 are voting PC
250 NDP
150 Liberal
300 are undecided
Same 1000 people also polled
350 think Horwath is best premier
300 think Ford is
150 think Wynne is
200 are undecided
What this tells me is the NDP will likely take a larger than expected share of the undecided voters.
Regardless, the PC lead is significant, and as such, my final projection, is as follows:
SPREADSHEETS
60 PC 37.86%
55 NDP 36.80%
8 LIB 20.89%
1 GRN 4.45%
Ford, and Wynne, both will be (re)-elected in their seats.
Mississauga will be where the election is decided. Had Ford won all of Mississauga he would have 62 seats, enough for a barebones majority, especially if he can convince an opposition MPP to become speaker. Had the NDP won the 3 seats they will be close in, they would have finished head of Ford. Mississauga really ends up becoming the city that decided the winner, and that winner, is Doug Ford.
Regardless, with only 60 seats, there remains a chance that a NDP-OLP coalition will take office instead. The two parties would need to agree to a solid "deal" before this happens, no informal deal will do, not with a 5 seat lead in the legislature. This means either a Peterson-Rae style "accord" or an outright coalition. Given the way the election has played out, this could well be easily possible.
Do not ignore the Greens either. They may offer Ford a way out, carbon taxes bring in a lot of revenue, revenue he will need to pay for his tax cuts. He could make the excuse that he requires Green support, and thus, needs to delete the carbon tax "later" and not now. Consider that if I'm off by even a single seat, anywhere, all of the math could be very different. A 61 seat PC party and a 1 seat Green party could form a majority, with an opposition speaker.
In the end there are two winners. Andrea Horwath, and the PC Party. It only remains to be seen whose win is bigger.
Penultimate Projection
I've added my "gut" back into the math.
62 N 38.52%
54 P 34.86%
7 L 20.64%
1 G 5.98%
This evening, the final projection (which will be much more wordy)
sheets: https://docs.google.com/spreadsheets/d/1B5E3KyuL0X9LudhYES-DbAl8UkeNV_LaJVuBOszJ4_I/edit?usp=sharing
(math note, after the speaker selection, which normally comes from the governing party, there would be 61 NDP seats, 61 from the Libs+PC, and 1 Green)
62 N 38.52%
54 P 34.86%
7 L 20.64%
1 G 5.98%
This evening, the final projection (which will be much more wordy)
sheets: https://docs.google.com/spreadsheets/d/1B5E3KyuL0X9LudhYES-DbAl8UkeNV_LaJVuBOszJ4_I/edit?usp=sharing
(math note, after the speaker selection, which normally comes from the governing party, there would be 61 NDP seats, 61 from the Libs+PC, and 1 Green)
Tuesday, June 5, 2018
Pre-Penultimate Projection
This will be my Pre-Penultimate Projection. Tomorrow morning, my Penultimate projection will come out, followed by the final projection which will post at 11:59pm on the 6th.
Since my last post, I've been directly contacted by Quito Maggi, who leads Mainstreet, and given access to the riding polls. I am deeply thankful for this and wish to make note of this for all readers.
I've used these to fine tune my spreadsheet. I note, however, that I've adjusted no riding to match the riding polls. The only riding that was changed, at all, was Guelph, and only in response to the other changes I made to the sheet. My information that "The Greens are leading by probably 1,000 or so votes" comes from a source in the Green Party that I trust, and not from the riding polls. All other changes are simply based on a far more accurate province-wide vote translation. In particular, the FF has been reduced to 0, and this has created far more realistic projections, including (somehow, and oddly) a better showing in urban areas. The LSD has been significantly toned down, and the AFF toned down as well as a result. All of this is why Guelph required an adjustment.
Note as well this projection has 0 "gut" in it. My penultimate, and final projections, will have far more of my intuition and judgement based into the numbers.
61 PC 36.91%
56 NDP 36.69%
6 LIB 20.64%
1 GRN 5.76%*
*Note that the "pop vote" for the Greens includes votes for all other parties, and as such, should be read as being reduced by 1 or 2 points.
The result of this would likely see a Ford minority, with the Libs abstaining on the throne speech, in return for a (public or private) commitment that the budget will not come until spring 2019. This would give the party time to find a new leader, for people to look at and re-judge all 3 (now 4 I suppose) parties, and allow the opposition to force an election on an actual Ford budget, one where any cuts he may need to make are public. For Ford it gives him time to get used to the position and put any party scandals behind him, and allows him to build up some experience in provincial politics.
Since my last post, I've been directly contacted by Quito Maggi, who leads Mainstreet, and given access to the riding polls. I am deeply thankful for this and wish to make note of this for all readers.
I've used these to fine tune my spreadsheet. I note, however, that I've adjusted no riding to match the riding polls. The only riding that was changed, at all, was Guelph, and only in response to the other changes I made to the sheet. My information that "The Greens are leading by probably 1,000 or so votes" comes from a source in the Green Party that I trust, and not from the riding polls. All other changes are simply based on a far more accurate province-wide vote translation. In particular, the FF has been reduced to 0, and this has created far more realistic projections, including (somehow, and oddly) a better showing in urban areas. The LSD has been significantly toned down, and the AFF toned down as well as a result. All of this is why Guelph required an adjustment.
Note as well this projection has 0 "gut" in it. My penultimate, and final projections, will have far more of my intuition and judgement based into the numbers.
61 PC 36.91%
56 NDP 36.69%
6 LIB 20.64%
1 GRN 5.76%*
*Note that the "pop vote" for the Greens includes votes for all other parties, and as such, should be read as being reduced by 1 or 2 points.
The result of this would likely see a Ford minority, with the Libs abstaining on the throne speech, in return for a (public or private) commitment that the budget will not come until spring 2019. This would give the party time to find a new leader, for people to look at and re-judge all 3 (now 4 I suppose) parties, and allow the opposition to force an election on an actual Ford budget, one where any cuts he may need to make are public. For Ford it gives him time to get used to the position and put any party scandals behind him, and allows him to build up some experience in provincial politics.
Monday, June 4, 2018
Monday Morning Update - Ontario
A shorter post today. Until I see more polls that have polled the period of time after Wynne's announcement, there is not much I can do wrt updating the projection.
Instead, I wanted to address another issue. That of riding polls which have been locked behind a paywall.
For new readers; I am on ODSP. This means that I get $1151 a month, which has to cover everything. Half of that goes to rent for my literally 10 foot by 10 foot sized apartment, and after you take off the internet, I have $465 left for the month. If I were to buy access to these riding polls, it would have to come from this; the same budget for my groceries.
I am fortunate, however, in that some people donate to my Patreon. This is currently at $5, almost all of which is taken up by donations to people who provide me with hours upon hours upon hours of free entertainment that I enjoy on a daily basis. I want to give them more.
The riding polls, I know from those who have access to them, show a strong PC result. However, no provincial polls are telling me the PC party is as strong as the riding polls are telling them.
Bluntly, this puts me in a very difficult situation.
Should the riding polls turn out to be correct, I may need to stop projections altogether. I'll literally have been priced from the market. I simply can not justify spending a whopping 12% of my post-rent budget on access to polls. To help understand how much money this is for me, I've done some math.
Supposedly, the median income for Canada is $76,000 per family, while the average rent is $989. Assuming other similar bills (like a car payment) which I will include in rent, this would leave a person with roughly $5000 a month, meaning 12% is equal to $600.
I hope this helps explain exactly why I can not possibly hope to afford this.
In short; if the provincial polls continue to show no rush to the PC Party, the riding polls do, and the PC party does indeed win by the margins the riding polls suggested, then I will have to stop making projections. There's no point in spending hours of research every day to simply embarrass myself.
Instead, I wanted to address another issue. That of riding polls which have been locked behind a paywall.
For new readers; I am on ODSP. This means that I get $1151 a month, which has to cover everything. Half of that goes to rent for my literally 10 foot by 10 foot sized apartment, and after you take off the internet, I have $465 left for the month. If I were to buy access to these riding polls, it would have to come from this; the same budget for my groceries.
I am fortunate, however, in that some people donate to my Patreon. This is currently at $5, almost all of which is taken up by donations to people who provide me with hours upon hours upon hours of free entertainment that I enjoy on a daily basis. I want to give them more.
The riding polls, I know from those who have access to them, show a strong PC result. However, no provincial polls are telling me the PC party is as strong as the riding polls are telling them.
Bluntly, this puts me in a very difficult situation.
Should the riding polls turn out to be correct, I may need to stop projections altogether. I'll literally have been priced from the market. I simply can not justify spending a whopping 12% of my post-rent budget on access to polls. To help understand how much money this is for me, I've done some math.
Supposedly, the median income for Canada is $76,000 per family, while the average rent is $989. Assuming other similar bills (like a car payment) which I will include in rent, this would leave a person with roughly $5000 a month, meaning 12% is equal to $600.
I hope this helps explain exactly why I can not possibly hope to afford this.
In short; if the provincial polls continue to show no rush to the PC Party, the riding polls do, and the PC party does indeed win by the margins the riding polls suggested, then I will have to stop making projections. There's no point in spending hours of research every day to simply embarrass myself.
Sunday, June 3, 2018
Narrative Change: Ontario
Wynne's announcement has focused the narratives around the election. By explicitly asking those swing voters who can't make up their mind on the NDP or the Tories, there is a good chance that she will bolster her own party in places where they could stop the other two parties.
How, you may wonder, would you ever calculate which ridings this would work in? The anwer goes to something I've had on my projection sheets for quite some time.
Download your copy of the updated sheets:
https://docs.google.com/spreadsheets/d/1B5E3KyuL0X9LudhYES-DbAl8UkeNV_LaJVuBOszJ4_I/edit?usp=sharing
The LSD, or Liberal Swing Drop. This replicates the exact vote pattern you would expect to see should Wynne's gambit pay off.
As such I've raised the LSD to 7, which is extremely high given what the formula actually does. This is so high, in fact, it's impacted the AFF in a minor way, and, has forced me to use the higher of the two FF numbers. To compensate I've increased the Greens in Guelph yet again (though the current projection still shows them behind)
52 - NDP - 37.17%
49 - PC - 36.65%
23 - LIB - 21.83%
Saturday, June 2, 2018
Two tales of election: Ontario
There seem to be two different tales of this election occuring at once.
One in which Doug Ford can not be trusted and people are swinging away from him to vote NDP.
One in which Doug Ford can be trusted, and is maintaining his standing among those wanting Change.
To that end, today, I present not one, but two projections. One in which Ford maintains his position, and one in which he fails to do so.
64 P 39.27%
48 N 35.61%
12 L 21.53%
https://docs.google.com/spreadsheets/d/1WvTuHrZOiqcVBQ4VNWPMO03oqfPJCY1H9DGliUtsgL4/edit?usp=sharing
remember to make your own copy to make your own projections
66 N 40.88%
42 P 34.05%
16 L 21.46%
https://docs.google.com/spreadsheets/d/1V9YHBfCF8VdGZAGb9H6zOUWzM13XUZMzvhzYGn0rqgI/edit?usp=sharing
different from the sheet above only in topline numbers.
One in which Doug Ford can not be trusted and people are swinging away from him to vote NDP.
One in which Doug Ford can be trusted, and is maintaining his standing among those wanting Change.
To that end, today, I present not one, but two projections. One in which Ford maintains his position, and one in which he fails to do so.
FORD HOLDS
64 P 39.27%
48 N 35.61%
12 L 21.53%
https://docs.google.com/spreadsheets/d/1WvTuHrZOiqcVBQ4VNWPMO03oqfPJCY1H9DGliUtsgL4/edit?usp=sharing
remember to make your own copy to make your own projections
FORD FAILS
66 N 40.88%
42 P 34.05%
16 L 21.46%
https://docs.google.com/spreadsheets/d/1V9YHBfCF8VdGZAGb9H6zOUWzM13XUZMzvhzYGn0rqgI/edit?usp=sharing
different from the sheet above only in topline numbers.
Subscribe to:
Posts (Atom)