Friday, March 23, 2012

Final NDP Power Rankings

The NDP crowns its next leader tomorrow, and the best we can do at this point is narrow the field down to 5 possible winners. That's a far cry from September when the Brian Topp juggernaut was described as Martinesque.

I made my "secret Liberal memo" endorsements earlier this month, and my opinion hasn't changed since then - I feel Mulcair is the strongest candidate, and as a partisan Liberal I'll be rooting for Peggy Nash or Brian Topp. Or maybe Nathan Cullen, because he's the most interesting candidate. Or maybe Paul Dewar, because that's how I'd likely vote if my blood ran orange. Or maybe Martin Singh, because I'd sure love to see what commercial the NDP prepared for him.

As for who will win, I don't know how many membership forms each camp sold so my best guess is nothing more than a wild guess. But here goes: I'll predict Mulcair comes in around 30%, with Cullen in second around 20%, and Nash, Dewar, and Topp hot on his heels. Mulcair over Nash on the final ballot.

To help size the race up, here's an update on the NDP Power Rankings - how the candidates fare in terms of fundraising, social media, endorsements, and buzz. (click for full size)

The "average share" column is simply an average of each candidate's share of the pie on these 9 indicators. It's by no means intended to be a predictor of first ballot support but, that said, I wouldn't be shocked to see numbers similar to this on Saturday:

Mulcair 26%
Nash 17%
Topp 17%
Cullen 16%
Dewar 15%
Ashton 6%
Singh 3%

The momentum numbers show how these pie slices have changed over the past week and the past month - in both instances, Mulcair and Cullen are gaining the most ground. Mulcair's gains have been primarily due to increased media attention, while Cullen has benefited from some very real gains in donations and social media support.

And this momentum is part of the reason I've predicted Cullen to finish second on the first ballot (that and the large number of BC NDP members). Although Cullen lags far behind on endorsements and his fundraising numbers aren't anything to write home about, he has now matched Mulcair when it comes to the total number of donors, and leads the Facebook "like" race.

February's Dewar poll and HosertoHoosier's analysis of the online preferential ballot, both suggest that Cullen's growth potential is limited, likely due to the divisiveness of his "co-operation" plan. But I expect him to raise some eyebrows when the first ballot results are read off at 10 am Saturday.

Labels: , ,


  • For handy reference, the first ballot results (as provided by

    Thomas Mulcair 19,728
    Brian Topp 13,915
    Nathan Cullen 10,671
    Peggy Nash 8,353
    Paul Dewar 4,883
    Martin Singh 3,821
    Niki Ashton 3,737

    Ashton is dropped from the second ballot, and Singh and Dewar voluntarily withdrew from the second ballot.

    By Anonymous Anonymous, at 12:41 p.m.  

  • The one to watch will be whether Nash picks up enough support to edge Cullen. There are some indications that she has a lot of growth potential. If not, it looks like a Topp-Mulcair showdown to me (I'm not convinced Cullen has a lot of growth potential).

    By Anonymous hosertohoosier, at 12:49 p.m.  

  • Over the next two ballots, Topp will need to pick up over 2/3rds of the support of Nash (who looked like she wanted to endorse Topp, but had read the tea leaves) and Cullen - and must do so without the endorsement of the former. That seems like a tall order.

    By Anonymous hosertohoosier, at 2:11 p.m.  

  • Interesting to note that, despite all the rhetoric out of the NDP about equity, each of the first two ballots knocked women off the ballot (two men withdrew voluntarily from the second ballot), and only men remain on the subsequent ballots.

    By Anonymous Anonymous, at 4:14 p.m.  

  • For future reference, we now know which numbers to pay attention to:
    number of MP endorsements and total donations.

    In both these categories, Topp came second, reflecting the spot he held after the first ballot and will hold until he concedes or loses.

    By Blogger Brian Henry, at 6:43 p.m.  

  • Out of curiosity, I looked at which of CG's measures best predicted the first ballot results. Here is the absolute value of the total error for each measure.

    Media mentions: 18.93
    Fundraising: 21.39
    Average share: 24
    Poll: 30.4
    Twitter mentions: 38.2
    Directed donors: 39.62
    Endorsements: 40.9
    Twitter: 49.22
    MP endorsements: 51.9
    Facebook: 66.35

    So much for social media!

    By Anonymous hosertohoosier, at 6:46 p.m.  

  • H2H look at the money raised column and line up who got the most to who got the least. It's exactly the same line up as who got the most votes to who got the least.

    It's a crude measure, but the only one that got it right, straight down the line.

    Number of MPs and Endorsements also succeed in predicting Mulcair in first place and Topp in second.

    And since it turns out the vote was purely a Mulcair vs Topp fight, those were pretty good predictors too.

    On the other hands, the social media measures aren't terrible ... if you turn them upside down and assume that domination on twitter means you have no grip at all on the real world.

    By Blogger Brian Henry, at 8:39 p.m.  

  • P.S. I'm tired tonight after a long day's work. Could you take pity and explain what "Endorsements" actually refers to? Endorsements from whom?

    By Blogger Brian Henry, at 8:40 p.m.  

  • Brian, rankings are a very noisy measure, and not a great indicator. I mean even Ekos got the ranking right in the last election - the real question was how close their predictions were to the actual results.

    If you look at each candidate's share of MP endorsements, the metric substantially overestimated Mulcair's support, and underestimated that of most other candidates.

    Which looks more like the results we saw today?

    That predicted by media mentions, or that predicted by MP endorsements?

    Media mentions
    Mulcair: 27.1% (vs. 30%)
    Nash: 12.2% (vs. 12.8%)
    Topp: 16.8% (vs. 21.4%)
    Cullen: 15.3% (vs. 16.4%)
    Dewar: 12.6% (vs. 7.5%)
    Ashton: 7.9% (vs. 5.9%)
    Singh: 8.1% (vs. 5.9%)

    MP endorsements
    Mulcair: 53.75%
    Nash: 11.25%
    Topp: 15%
    Cullen: 5%
    Dewar: 10%
    Ashton: 5%
    Singh: 0

    Media mentions was a better predictor of support than MP endorsements for Mulcair, Nash, Topp, Cullen and Singh, and wasn't way off for Mulcair.

    By Anonymous hosertohoosier, at 9:46 p.m.  

  • I hear what you're saying, and you're right - media mentions works very well. Thanks for pointing that our. I hadn't even noticed.

    But I don't know that I agree with your premise:

    Yes, for elections, the question is how close a predictor comes to actual results. After all, a few percentage points may result in a big change in the seat count.

    Also, in an election, we're not just interested in who comes first. We want to know if they're going to get a majority, and we want to know who the oppostition is going to be, and if the Bloc will finally be sent to hell, etc.

    But in a leadership race, ranking is everything; specifically, we're asking, who's going to win?

    Beyond that, it would be nice to have a finely tuned prediction of the actual vote count. But in a leadership race, we're not going to get the sort of accurate polling we get in an election.

    And given the absence of accurate polling, I'd suggest that a crude but effective method is actually better.

    Here's the bottom line: Mulcair walked away with the vote.

    Which measure best predicted that result? The number of MPs he lined up.

    Of course, I don't follow these things as closely as you.

    Could be there have been other recent leadership contests in which the number of MPs lined up was dead wrong in predicting who would win and that money raised also turned out to be a bad predictor of the winner, while media mentions turned out to be the crystal ball every time.

    I really have no idea, and I'm perfectly willing to bow to the historical evidence.

    By Blogger Brian Henry, at 8:21 a.m.  

  • Calgary Grit,
    What are "Endorsements"? Endorsements by whom?

    By Blogger Brian Henry, at 8:22 a.m.  

  • By Blogger Unknown, at 1:59 a.m.  

Post a Comment

<< Home