Federal Elections 2011: Seat Predictions

[nittanylionstorm07]

Now we can all do our own calculations.

http://www.tooclosetocall.ca/p/canada-simulator.html

Thanks, though that's a weird website. I edited Alberta down to 51/13/22/16 and it still shows all seats going to the CPC. Then for fun, I gave all of the votes to the NDP and it still shows the CPC with 19 seats versus the NDP with 8 and the GRN with 1 lol.

I edited my above seat totals to include Lizzie May, who I forgot about. I still think she'll win Saanich.

[punked]

Thanks, though that's a weird website. I edited Alberta down to 51/13/22/16 and it still shows all seats going to the CPC. Then for fun, I gave all of the votes to the NDP and it still shows the CPC with 19 seats versus the NDP with 8 and the GRN with 1 lol.

I edited my above seat totals to include Lizzie May, who I forgot about. I still think she'll win Saanich.

Well the maker of the projection admits his has the same problem all the other ones do. No one knows what happens when the NDP goes above 20%. There are ceilings built into all swing models. That is the biggest problem with this election and what makes it so exciting.

[August1991]

Similar to the link above, Hill & Knowlton, the PR firm, has a nifty seat predictor that allows you to shift popular votes around between parties (using the 2008 election results as a baseline) and see what effect they have on seat totals.

Using the same idea, I downloaded the 2008 results from a Simon Fraser University website and played around with the Quebec numbers in my own spreadsheet. In particular, I tried shifting chunks of voters on a riding-by-riding basis from the Bloc to the NDP.

What is striking is that despite different "reasonable" scenarios, the Bloc always gets about 45 seats and the Conservatives keep their 10 or so seats. At most, it is merely the Liberals and NDP that trade seats among themselves.

Here's an example:

Popular Vote

Bloc 31.9%

CPC 18.5%

Green 2.7%

LPC 18.7%

NDP 26.3%

Seat Totals

Bloc 48

CPC 10

Green 0

LPC 12

NDP 5

The NDP would win Outremont, Hull, Gatineau, Jeanne-Le Ber and Pontiac.

----

I ssupect that the main impact of the surge of the NDP in Quebec will not be seats in Quebec but rather the effects of this in Ontario and BC.

Edited April 23, 2011 by August1991

[cybercoma]

I don't know about that system. I dumped in the most favourable split for the NDP and it shows the Conservatives gaining 5 seats in Quebec, the Liberals picking up a highly unlikely 8 more seats in Quebec and the NDP only picking up 1 more.

I'm not sure about that.

[cybercoma]

When I just put the Quebec numbers from Ipsos-Reid in, the predictor gives me 67 seats for Gilles Duceppe, 1 for Harper, 1 for Layton and 6 for Ignatieff. Huh?

[Posc Student]

I always find weird flaws in that seat predictor, like the NDP going from like 40% in 2008 to literally 0% and being bet my the Greens who are at 1.3%.

[Mr.Canada]

harper is going to get a majority and theirs nothing anyone can do about it. Project away but it won't change reality.

[nittanylionstorm07]

harper is going to get a majority and theirs nothing anyone can do about it. Project away but it won't change reality.

lol delusional Tories. Why do you fail to understand that this is a completely new situation?

[Wild Bill]

lol delusional Tories. Why do you fail to understand that this is a completely new situation?

Another one for the list!

[Harry]

The games people play.

One of the seat projectionists is presently forecasting 74 seats for the NDP. I'm not surprised seeing all the absolute nonsense in the mainstream press.

Rubber Hits the Road for Parties … and Seat Projectors

This may surprise some people, but very few of the current amateur seat projection websites have even a single federal general election track record under their belts. And none of them has had to predict an election where so many assumptions have been upended, and so many tectonic shifts have been telegraphed in leading indicators whose full effect has yet to be seen in the horserace numbers.

Here was the track record of the final 2008 seat predictions from various seat predictors in the field during the last election (thanks to Paulitics for archiving them, since many of the authors have subsequently taken theirs down):

http://www.punditsguide.ca/2011/04/rubber-hits-the-road-for-parties-and-seat-projectors/

Edited April 25, 2011 by Harry

[Bryan]

The games people play.

One of the seat projectionists is presently forecasting 74 seats for the NDP. I'm not surprised seeing all the absolute nonsense in the mainstream press.

That article points out something important to remember not just when predicting seats, but also when reading the daily polling. Going back several elections now, pretty much every pollster and projector has shown Conservative support as being considerably lower than what they actually get on election day. Conversely, they pretty much all over-estimate the Liberal support. This is even the case in provincial elections, as the article points out.

What's most interesting to me is the BQ and NDP numbers for the most part are much closer to the actual result. That is why the NDP surge in Quebec is so interesting, because they really could be getting the support the polls show. It makes using seat projection models (which are already suspect at best) tough, because previous data of a similar NDP surge can't be factored into the model.

[bloodyminded]

Another one for the list!

Interestingly, you missed the declarative prediction directly before the one you implicitly criticize...one far more appropriate to your list.

[cybercoma]

That article points out something important to remember not just when predicting seats, but also when reading the daily polling. Going back several elections now, pretty much every pollster and projector has shown Conservative support as being considerably lower than what they actually get on election day. Conversely, they pretty much all over-estimate the Liberal support. This is even the case in provincial elections, as the article points out.

What's most interesting to me is the BQ and NDP numbers for the most part are much closer to the actual result. That is why the NDP surge in Quebec is so interesting, because they really could be getting the support the polls show. It makes using seat projection models (which are already suspect at best) tough, because previous data of a similar NDP surge can't be factored into the model.

That was answered by the article and to be frank it doesn't bode well for the Conservatives:

The inherent bias in seat prediction methodologies to favour previous election results means they tend to overly favour parties set to lose seats, such as the Liberals and Bloc Québécois in the last election. They also tend to miss the likelihood of parties on the rise to gain seats, such as was the case with the Conservatives in the last election. Only the NDP, whose vote intention numbers showed little gain by the end of the 2008 campaign, saw seat count predictions on both sides of its eventual total.

The Liberals were over-estimated because they lost seats during the last campaign. The Conservatives were considerably under-represented because they were set to overthrow the Liberals. The seat predictors do not capture swings in the electorate very well. They're "sticky" if you will. It could very well be the Conservatives over-represented this time around and the NDP under-represented.

Edited April 25, 2011 by cybercoma

[cybercoma]

Oh and it also doesn't bode well for the Bloc.

[sk1d]

http://www.projectdemocracy.ca/

This site shows the projected voting by riding. Enter your postal code and it will tell you how many votes they think each candidate will get get in your riding. There's also an option to show different projections based on different polls.

[capricorn]

A list of 95 ridings to watch in the election.

http://www.thehilltimes.ca/page/view/95ridings-04-25-2011

[nittanylionstorm07]

A list of 95 ridings to watch in the election.

http://www.thehilltimes.ca/page/view/95ridings-04-25-2011

A terribly written part with respect to Quebec. There is no way the Liberals are "in" that many ridings with respect to the BQ. It's BQ vs. NDP not BQ vs. Liberals in this election.

[Evening Star]

A list of 95 ridings to watch in the election.

http://www.thehilltimes.ca/page/view/95ridings-04-25-2011

Interesting. They're not predicting that vote splits will help Conservatives anywhere. The number of NDP-CPC battles out West is intriguing.

[Handsome Rob]

Interesting. They're not predicting that vote splits will help Conservatives anywhere. The number of NDP-CPC battles out West is intriguing.

We have our fair share of lefties, but lefties and righties alike can share in the dislike for the Liberal party. I wouldn't be surprised if they ceased to exist west of Winnipeg soon.

[August1991]

A terribly written part with respect to Quebec. There is no way the Liberals are "in" that many ridings with respect to the BQ. It's BQ vs. NDP not BQ vs. Liberals in this election.

I don't quite understand what you mean but I looked at the Hill Times predictions for Quebec and they make sense.

The Liberals and Conservatives have their vote concentrated in certain ridings. The Bloc, and in particular the NDP, have their votes spread out. If you shift 10% of the Bloc vote to the NDP in Quebec (very possible), the NDP still only wins a handful of seats and the Bloc stays with over 40.

Interesting. They're not predicting that vote splits will help Conservatives anywhere. The number of NDP-CPC battles out West is intriguing.

Someone elsewhere said that if you draw a north-south line just west of London, Ontario, the NDP dominates the non-Tory vote in the West and the Liberals dominate in the East.

Nevertheless, the issue in this election is not which party is going to win the most seats: that without doubt will be the Conservatives. The question is whether the Conservatives receive more than 154 seats with about 38% of the popular vote. To do, they will have to be lucky and get some fortunate vote splits.

Anyone know how to play bridge? If the suits split badly, it can make for a rough game.

[Harry]

A list of 95 ridings to watch in the election.

http://www.thehilltimes.ca/page/view/95ridings-04-25-2011

A lot of their forecasts seem to be ancient history now.

[Evening Star]

Someone elsewhere said that if you draw a north-south line just west of London, Ontario, the NDP dominates the non-Tory vote in the West and the Liberals dominate in the East.

That's very apt!

[Harry]

Finally the seat projectors are stating to move - they are a bit behind the times but better late than never I suppose.

Election Almanac just made a change based on today's polls - we're getting there slowly but surely:

Party / 08 GE / Apr 20 / Apr 21 Apr 23 / Apr 24 / Change

Cons / 143 / 181 / 141 / 143 / 130 / Down 13 seats

NDP / 37 / 49/ 50 / 48 / 64 / Up 27 seats

Libs / 77 / 47 / 77 / 78 / 76 / Down 1 seat

Bloc / 49 / 31/ 40 / 39 / 38 / Down 11 seats

http://www.electionalmanac.com/canada/projections.php

[Peter F]

My Guess:

CPC 144

LPC 75

BQ 45

NDP 42

Ind 2

[Harry]

Election Almanac before the Angus Reid poll that came out today showing the NDP at 30%

Cons - 128 seats

NDP - 103 seats

Libs - 65 seats

Bloc - 12 seats

http://www.electionalmanac.com/canada/