Friday, April 29, 2011

Your Daily Seat Projections

The big update will wait until the final round of polls come in this weekend but, for now, the latest from Nanos, Ekos, and Decima have been added.

CPC: 36.1% (-0.2)
NDP: 27.7% (+1.1)
Lib: 22.7% (-1.1)
Bloc: 6.6% (-0.3)
Green: 5.6% (+0.1)

As a reminder, the following projections are based on 10,000 simulations, taking into account the polling margins of error, how shifts in regional support have historically transferred to individual ridings, and the chance the pollsters could just miss the mark, like in some previous elections.

This is the only simulation model out there and, as such, it's far more effective when projecting tight races or three-way races since it recognizes the riding could go either way depending how the numbers break.

CPC: 137 to 162 seats (mean: 149.2)
NDP: 63 to 94 seats (mean: 78.1)
Lib: 44 to 68 seats (mean: 56.7)
Bloc: 12 to 35 seats (mean: 23.2)
Ind: 0.8

Odds of Tory victory: 100%
Odds of Tory majority: 20% (down from 26% on Wednesday)
Odds of NDP official opposition: 97%

Since Wednesday, the NDP are up another 9 seats on average, and are now projected to win 23 to 51 in Quebec, 15 to 22 in Ontario, and 7 to 14 in BC. Those confidence intervals should narrow over the weekend once we're hit with a batch of new large-sample polls, but there's a lot of volatility out there right now, so I suspect we'll have to wait until election day to see how a lot of seats break.

Remember, numbers could shift over the weekend. This isn't a prediction for Monday - it's an approximation of how the current polling numbers translate into seats.

(short methodology, long methodology)



  • So here's a 300 million dollar question. If the Conservatives do get 150 seats and the Liberals are in third place looking at oblivion, can the Conservatives convince 5 Liberals to cross the floor? Hmm and who would they be?

    I say Scott Brison is one of them.

    By Anonymous Anonymous, at 4:46 p.m.  

  • Why just call on Tories to cross the floor? A number of the newly elected NDP MPs will be people with very little tie to the party, and in some cases, very low chances of surviving a hotly contested nomination race, once people realize that those seats could represent a parliamentary meal ticket.

    If I was Ruth-Ellen Brousseau (trailing the Bloc candidate in Berthier-Maskinongue by a mere 7 points in the latest riding poll), and the Tories offered me a cabinet position, I'd jump at the opportunity.

    By Anonymous hosertohoosier, at 5:12 p.m.  

  • Depending how the independents break, and who gets named speaker, it's certainly possible the Tories could try to coax some opposition MPs over. They've done it before, after all.

    I think that point underscores how we shouldn't be as fixated on the "155" math as we are. MPs cross the floor, MPs resign, parties lose by elections.

    Yes, Harper needs 154 or 155 for that first throne speech and budget, but after that, things could shift along the way.

    In some ways, it's almost easier for the opposition parties if Harper gets 155 seats. They can then oppose him, but if his government becomes hugely unpopular, odds are he'll lose a by election or two, opening the door for them.

    By Blogger calgarygrit, at 5:27 p.m.  

  • Is it too late yet for me to predict that the NDP get 82 seats just in Quebec?

    I mean, we have all seen some pretty ridiculous predictions out there; I figure one more won't hurt.

    And it would be fun to see the Bloc (well, all the Parties I guess) *owe* the NDP seven seats.


    By Anonymous Anonymous, at 5:28 p.m.  

  • If the NDP gets 82 seats in Quebec we can call Jack Layton a dirty election cheater.

    Quebec only has 75 seats.

    By Anonymous Mark McLaughlin, at 5:35 p.m.  

  • mark, owing 7 should have been the giveaway.

    By Anonymous mikey, at 5:50 p.m.  

  • In your final evaluation, could you please also put a number on the odds that NDP + LP > 154

    By Anonymous Anonymous, at 5:56 p.m.  

  • Well Dan, I would say that 155 is still an important target for any chance of an NDP-Liberal coalition. The removal of the Bloc from any equations would grant them a level of legitimacy that they might need immediately. It would also ensure, because of the Bloc, that the Conservative total would be low enough to discourage floor crossing. I would still feel a lot more comfortable as the next government, whichever group it is, if I had 160+ seats. That said, as long as the first Throne Speech and budget can pass without the Bloc then dipping into a large minority afterwards is not the end of the world.

    Parliament could get very ugly with CPC at ~150, NDP-Lib at ~150 and Bloc at 5-15. I imagine with those numbers there would be all kinds of semi-dirty attempts to woo floor crossers. It really wouldn't help to reverse Canadians' opinions of Parliament and politicians.

    By Anonymous Robin, at 6:04 p.m.  

  • Brison? He would face some pretty pissed-off constituents here in NS.

    Dan, for whatever reason, I've only just noticed that your prediction model employs a nice Monte Carlo method to arrive at your point estimates and variances.

    By Blogger JG, at 6:08 p.m.  

  • 2 Bloc members support NDP
    CBC – Fri, 29 Apr, 2011 3:33 PM EDT.. .

    Two members of the Bloc Québécois are openly calling on sovereigntists to vote for the NDP on Election Day.

    Maxime Bellerose, the former president of the Hochelaga riding association, and Benoît Demuy, a one-time political adviser to former Bloc MP Réal Ménard, have drafted a letter in which they write that the Bloc has reached its limits in Ottawa and that it's time to move on to another option.

    In the letter, Bellerose and Demuy write that they are still convinced sovereignty is the right path for Quebec, but that the struggle needs to happen in Quebec, not in Ottawa.

    "While waiting for a winning referendum, voting NDP today is the solution to break the cycle of minority Conservative governments," the letter reads.

    It goes on to say that for the first time, social democracy is knocking on the doors of Parliament, and that it would be a shame for Quebecers not to take the opportunity to send MPs to Ottawa who champion the values of mutual help and justice held dearly in Quebec.

    By Blogger JimTan, at 6:40 p.m.  

  • BTW, Larry Campbell was saying that NDP and LPC should merge if LPC fails the test this election.

    How likely is that?

    By Blogger JimTan, at 6:44 p.m.  

  • Why should the NDP entertain a coalition with the Liberals? The Liberals will have nothing to offer.

    Certainly not a base in Quebec or the West, and only mariginally so in Ontario and the East.

    The situation is not at all like that of the PCs and Alliance pre-merger, where both parties could bring each other much-needed strengths, and where the ideological differences were much narrower.

    By Anonymous Nonpareil, at 7:26 p.m.  

  • If the NDP does in fact end up forming the Official Opposition, then, I think it is likely best for the long-term interest of the Liberal Party to have a majority Conservative government.

    Having to choose between supporting Layton and Harper is a Lose-Lose for the Liberals.

    Supporting Layton only cements further into the electorate's mind that the NDP is a viable governing alternative. Supporting the Conservatives only gives the NDP a bigger stick with which to beat up on the Liberals next election.

    A majority Conservative government allows the Liberal Party the time and space to redefine the party and properly assess all options.

    By Anonymous Anonymous, at 7:50 p.m.  

  • I think the Liberal chances of maintaining official opposition status just got better.

    By Anonymous hosertohoosier, at 10:05 p.m.  

  • Problem is that it's in the Sun. Who trusts the Sun?

    By Blogger sharonapple88, at 10:24 p.m.  

  • It seems like the standard deviation for each party is only 6-8 seats. That seems awfully low to me. If I remember correctly, the Liberals overshot most projections by more than that in 2004 and 2006, and undershot by at least that much in 2008.

    Of course, it could be that your model is much more reliable than most. Or maybe the past 3 elections were all freak. But I'm curious why the estimated uncertainty isn't higher.

    By Blogger Election Watcher, at 11:21 p.m.  

Post a Comment

<< Home