Fun with Numb3rs

Trudeau’s Win by the Numbers

Posted on by CalgaryGrit in --- 2013 LPC Leadership Race, Featured Posts, Fun with Numb3rs | 15 Comments

trudeau-family15nw5Over the past year, there have been thousands of articles written about Justin Trudeau, his father, and his leadership campaign. Since it hasn’t been a big secret he was going to come out on top, we’ve also seen thousands of articles about what his win means.

So rather than rehash what has already been written, allow me to provide the cold hard numbers behind his victory.


That’s more than voted in the most recent NDP (65,108) or Conservative (97,397) leadership races – indeed, it might very well be the most Canadians to ever vote directly for the leader in a federal leadership race. I say “federal”, because, despite what was claimed earlier today, the 2006 Alberta PC leadership race drew 144,289 votes.

Either way, I wouldn’t read too much into this. Both the BC and Alberta Liberals had high turnout leadership races in 2011, and it doesn’t appear to have translated to general public support. But at the very least, Justin Trudeau now has a lot of semi-engaged Liberals to draw from for donations and volunteers.


It’s difficult to compare this total to delegated conventions – especially delegated conventions from the good old days. But, for fun, Trudeau’s first ballot support ranks behind Martin (94%), is comparable to Pearson (78%), and is decidedly ahead of St. Laurent (69%), Chretien (57%), Turner (46%), King (36%), the other Trudeau (32%), and Dion (18%). Trudeau performed slightly better than Stephen Harper, who received 69% of the votes and 56% of the points (after they were weighted by riding) in 2004.


Trudeau’s crushing triumph certainly makes it look inevitable in hindsight. Maybe it was, but we’ve seen “can’t miss” candidates miss before.

If you look at the Intrade stock for a Trudeau victory, it ranged from 75% to 91%, showing that at least some people were willing to bet against him. Back in December, I asked readers of this blog to offer their predictions on the race, and while every entry except one had Trudeau winning, he was only given an average score of 41% on the first ballot. Remember, these are people who follow politics closely.

Even a few days ago, my poll of readers predicted an average first ballot figure of 65%, and only one-in-ten thought he’d crack 80%.

Of course, the support was always there, even if we didn’t all see it. But speaking as someone who was convinced to vote for Trudeau based on his performance during this race, I think the candidate and the campaign deserve a certain amount of credit for the magnitude of his victory.


Here are my final Power Rankings, with each metric converted to a percentage:

Total $ Donors Endorsement Media Facebook Twitter Power Rank
Justin Trudeau 63% 68% 90% 77% 84% 91% 78% (+3)
Joyce Murray 13% 16% 8% 7% 2% 3% 9% (–)
Martha Hall Findlay 11% 9% 1% 7% 10% 4% 6% (-1)
Martin Cauchon 9% 2% 1% 4% 3% 1% 4% (–)
Karen McCrimmon 2% 2% 0% 3% 0% 0% 1.7% (–)
Deborah Coyne 2% 3% 0% 2% 1% 1% 1.5% (-1)

Even though these power rankings weren’t intended to predict first ballot support, they came within 2 percentage points for every candidate:

Power Rank Actual
Justin Trudeau 78% 80%
Joyce Murray 9% 10%
Martha Hall Findlay 6% 6%
Martin Cauchon 4% 3%
Deborah Coyne 1% 1%
Karen McCrimmon 2% 1%

I’m sure some of that is luck, but this is definitely an exercise I plan to continue on future leadership races.

Final Power Rankings

Posted on by CalgaryGrit in --- 2013 LPC Leadership Race, Fun with Numb3rs | 10 Comments

Trudeau Sophie danse

There isn’t a lot of suspense surrounding Sunday’s Liberal leadership vote. Pick the metric of your choice – fundraising, endorsements, hair volume – and Trudeau leads his nearest challenger by at least a 4:1 ratio. I wouldn’t put a lot of stock in Twitter support, but Justin has 10 times more followers than the rest of the field combined.

The following table provides an overview of what little quantitative data we have on the race and offers a Power Rank, based on how these variables have translated to votes in past contests (methodology here).

Fundraising Endorsement Media Facebook Twitter Power Rank
Justin Trudeau $1,078,866 90% 77% 73,992 199,394 78% (+3)
Joyce Murray $225,310 8% 7% 2,147 5,848 9% (–)
Martha Hall Findlay $192,280 1% 7% 8,563 7,923 6% (-1)
Martin Cauchon $148,739 1% 4% 2,612 1,655 4% (–)
Karen McCrimmon $36,222 0% 3% 387 901 1.7% (–)
Deborah Coyne $31,651 0% 2% 501 2,199 1.5% (-1)

The bracketed number on the final column reflects changes from the last update – you can consider it a “momentum” score of sorts. Although that number shows Trudeau gaining ground, that’s completely a by-product of increased media attention – his share of the fundraising pie has actually dipped from 66% to 63%, with Joyce Murray and Martin Cauchon finishing strong on that front.

As I’ve stressed before, this isn’t a first ballot prediction, though it seems like as good a guess as any. I’d personally bet the “under” on 78% for Trudeau, but I do think he’s heading for a clear majority – and not just a “clear majority” by NDP standards.

In the end, whether Trudeau nabs 78%, or 63%, or the 112% some seem to be expecting, is irrelevant. Paul Martin received 94% of the vote and Michael Ignatieff got every vote, but both inherited deeply divided parties. While I have no doubt there will still be gripping from anonymous Liberals in the years to come, Team Trudeau has smartly run a positive and mostly unantagonistic campaign, that should leave Justin with relatively few enemies within the party.

That Trudeau exits this race untarnished and that the party exits this race united are far more important than whatever number is announced on Sunday.

What percentage of the weighted vote will Justin Trudeau get on the first ballot?

Updated Power Rankings Show Trudeau in Control

Posted on by CalgaryGrit in --- 2013 LPC Leadership Race, Federal Politics, Fun with Numb3rs | 18 Comments
Karen McCrimmon recognizes the state of the race better than most.

Karen McCrimmon recognizes the odds she’s facing

When I released my first set of LPC Power Rankings in early February, I was a bit surprised to see Justin Trudeau up at 66%. These rankings aren’t intended to be a first ballot predictor, but they came pretty close to the mark in the NDP contest and it was still a bit of shock to see Trudeau 54 points above his nearest competitor. But wouldn’t you know it, Marc Garneau’s mystery poll was essentially spot on my numbers. So maybe there’s something to this.

And if there is, we are heading to an absolute rout.

Fundraising Endorsement Media Facebook Twitter Power Rank
Justin Trudeau $1,001,060 94% 60% 71,773 195,672 75% (+9)
Joyce Murray $169,411 5% 13% 1,998 5,615 9% (+4)
Martha Hall Findlay $178,590 1% 10% 8,571 7,819 7% (+1)
Martin Cauchon $103,203 1% 7% 2,565 1,609 4% (+3)
Karen McCrimmon $26,259 0% 6% 375 848 2% (+1)
Deborah Coyne $27,385 0% 5% 479 2,155 2% (+1)

You can see the methodology behind these rankings here. Since the last update, I’ve sweetened the recipe with ever-so-small weights for number of donors and Facebook “talking abouts”, but it doesn’t change the rankings.

The bracketed number on the final column reflects changes from the last update – you can consider it a “momentum” score of sorts, with everyone picking up some of the pieces from the Garneau, Takach, and Bertschi campaigns. Trudeau’s +9 score is nearly as much as the rest of the field combined, and he shows no signs of slowing down the stretch.

Nearly doubling her Power Score since the last update is Joyce Murray, who has raised an additional $100,000, picked up 1200 new Twitter followers and 800 likes, while earning an endorsement by Ted Hsu.

This sets up an interesting battle for second between Murray and Hall Findlay, but it appears to be a battle for a very, very distant second.

Liberal Leadership Power Rankings

Posted on by CalgaryGrit in --- 2013 LPC Leadership Race, Fun with Numb3rs | 26 Comments
All signs point to a Justin Trudeau cakewalk

Justin Trudeau is head and shoulders ahead of his closest competitors

During the NDP leadership race, I got into the habit of tabulating “Power Rankings” of how the different candidates fared on fundraising, Facebook, Twitter, polls, and any other shred of quantitative data I could claw my hands onto. The exercise wasn’t intended to predict the first ballot vote, but it actually came surprisingly close to the mark.

After measuring how closely correlated those various metrics were to candidate support in seven recent races (including the NDP contest), I’m ready to launch the Liberal Leadership Power Rankings, based on:

1. Fundraising (30%): My research showed total dollars raised to be a better predictor than total number of donors, so that’s what I’ll be looking at – using the figures helpfully compiled by Pundits Guide.

2. Endorsements (30%): In past contests, a simple count of caucus endorsers has been as good a predictor of success as more elaborate systems, but there aren’t a lot of Liberal MPs to count so I’ll be relying on Eric Grenier’s endorsement scores. Alison Redford and Christy Clark both showed that endorsements can be overrated but, on the whole, they were a better predictor of support than fundraising numbers in the races I looked at – and they provided a good picture of the NDP contest.

3. Media Mentions (30%): This simple count of Google news stories is incredibly crude but, for whatever reason, it turns out to be one of the strongest predictors of first ballot support. Just goes to show the media may not be as clueless about leadership races as they’re often painted to be.

4. Social Media (10%): Turns out this has next to no relationship with support but, hey, it’s fun to count. So I’ll give 5% to Twitter followers and 5% to Facebook likes.

This formula may be tweaked if new data (i.e. polls among Liberal members) is released, but here are the preliminary rankings:

Fundraising Endorsement Media Facebook Twitter Power Rank
Justin Trudeau $673,157 78% 56% 66,132 183,370 66%
Marc Garneau $122,616 17% 11% 3,472 10,908 12%
Martha Hall Findlay $149,877 1% 7% 3,568 7,086 7%
Joyce Murray $56,554 3% 10% 1,200 4,471 5%
George Takach $106,233 0% 4% 1,071 1,698 4%
David Bertschi $0 1% 7% 498 1,287 2%
Martin Cauchon $0 1% 2% 2,519 1,320 1%
Karen McCrimmon $20,275 0% 2% 302 628 1%
Deborah Coyne $16,355 0% 2% 371 1,966 1%

Again, this isn’t a first ballot prediction. The fundraising numbers are out of date, and this magic formula overlooks the most important variable in all of this – the ground game. These rankings are intended only as a fun exercise to give a sense of candidate support and momentum.

So when you see that “66%” next to Justin Trudeau’s name, don’t take it as proof this race is over. It’s possible someone might break free of the pack and narrow that gap. But as it sits now, every sign points to a crushing Trudeau victory.

Convention Math

Posted on by CalgaryGrit in 2013 OLP Leadership Race, Fun with Numb3rs, Ontario Politics | 3 Comments

Wynne Pup

On Saturday, Maple Leaf Gardens was a cauldron of emotions. There were tears, broken promises, dashed dreams, and shrieks of pure unadulterated joy. That’s to be expected when you bring 2,000 people with very different motivations and beliefs together, and ask them to figure out who will govern a province of over 12 million people.

Yet despite all of that, in the end, what mattered more than the signs and scarves and speeches were the cold hard numbers. So it’s worth pausing to study the math behind Kathleen Wynne’s historic victory:

The First Ballot

The big story of the first ballot was Wynne’s jump to within 2 votes of Pupatello. To get a sense of where that support came from, it’s important to not look at the number of delegates elected but at the number who were actually registered at the convention and eligible to vote after backfills, alternate bumps, and no-shows are taken into account:

Pupatello   495
Wynne 454
Kennedy 253
Takhar 235
Sousa 200
Hoskins 100

There were also 67 registered independents and 320 registered ex-officios. Here’s how they broke in round one:

First Ballot

That means 40 delegates either didn’t vote or ate their ballots – including at least one Takhar delegate, unless Harrinder forgot to vote for himself.

The real story of this was Wynne picking up over 40% of the uncommitted voters. Pre-convention reports had her 10-15 behind Pupatello in the ex-officio count, so it seems likely she was able to snag most of the independent voters – and I guess that shouldn’t be too surprising, given many of them were former Glen Murray supporters.

Until we read the tell-all memoirs in 20 years, we won’t know for sure what was going through each candidate’s head at this point, but Wynne’s bounce certainly shifted the odds in her favour, perhaps prompting Hoskins to also shift his endorsement. It also meant the path to a Kennedy victory now relied on more aggresive math – with Hoskins moving to Wynne, Kennedy would have needed at least three-quarters of Takhar and Sousa’s delegates to reach the final ballot. This likely killed any talk of a third option, explaining Takhar’s bizarre move to Pupatello after the deadline to withdraw.

The Second Ballot

Second Ballot

With the race now clearly a two-woman show, Kennedy was only able to grow by 4 delegates, and Sousa fell by 19. While no candidate is ever able to deliver 100% of their delegates, both the numbers and what I saw on the floor suggest that’s almost exactly what happened. Pupatello’s vote jumped by 218 – toss in the 18 confused Takhar delegates who voted for their unofficially withdrawn leader, and you nearly hit Takhar’s first ballot number on the money. Similarly, Wynne’s gain of 153 was nearly spot-on to Hoskins’ first ballot total (though some high profile Hoskins supporters did go to Pupatello, including the Right Honourable John Turner).

While Kennedy and Sousa could have stayed around and pushed the inevitable back to midnight, both recognized the reality of the situation and withdrew. A Sousa-to-Pupatello and Kennedy-to-Wynne scenario would have set up an interesting final ballot, but it does not appear that either candidate nor their supporters had much appetite to back Pupatello, whose team had spent much of the campaign belitleling them.

Both men marched to Wynne, effectively sealing the deal. Unless the protestors outside burned the building to the ground, the math was now such that there was virtually no way for Pupatello to hold her lead.

The Third Ballot

As Jeff Jedras reports, some delegates decided it wasn’t worth waiting for the burst water pipes to be fixed and called it a day. Still, all but 57 stayed and voted:
Third Ballot

Together, Kennedy and Sousa moved 89% of their vote to Wynne on the final ballot – remarkable when you consider the historical norms, but likely in line with what Takhar and Hoskins also delivered.

Perhaps the timid nature of the campaign left most delegates without strong feelings towards either of the frontrunners, so they figured they might as well follow their man. Perhaps the short timeline gave candidates little time to woo delegates for second ballot support. Perhaps the four defeated candidates all commanded an unusually high sense of loyalty from their troops.

Whatever the reason, this convention came down to cold, hard, delegate math. And the math worked for Wynne a lot better than it did for Pupatello.

Alberta Votes Day 28: What to expect tomorrow

Posted on by CalgaryGrit in 2012 Alberta Election, Alberta Politics, Seat Projections | Leave a comment

I’ve already written about the difficulties of using seat projection models due to the unique circumstances of this Alberta election. So rather than crunch the numbers, I thought it might prove more useful to conduct a “wisdom of the masses” exercise, by looking at the predictions being made in  the Great Alberta Election pool. Here’s what the nearly 100 politicos who have entered so far are expecting tomorrow:

Seat Totals

The above graph shows the median prediction for each party, with the bars representing the first-to-third quartile range – that is, where the “middle half” of all entries fall. Looking strictly at the means, we get a “best guess” of 42 Wildrose seats, 37 PC seats, 4 NDP seats, 3 Liberal seats, and 1 for the Alberta Party. That’s a slightly narrower gap than my prediction of WR 44, PC 35, NDP 4, Lib 4, AP 0 – and it suggests there’s some skepticism out there about Smith’s ability to seal the deal. After all, three-in-ten respondents still predict a PC victory and only a minority (41%) expect Smith to win a majority.

Races to Watch

Nine-in-ten expect Redford to hold her seat, but respondents are less certain about Raj Sherman (43%) and Ted Morton’s (19%) chances. It’s a little counter intuitive for Morton to get drowned in the Wildrose wave given he’d be right at home in the party, but I have a hard time seeing any PCs left standing in southern rural Alberta.

As for the Wildrose Party’s more colourful candidates, only 36% expect Allan Hunsperger (of “gays burn in hell” fame) or Ron Leech (of “being white is an advantage” fame) to be elected Monday. I’d be shocked if Hunsperger won, but Leech took 26% of the vote running as an independent last election – add in the Wildrose boost, and he seems like the odds on favourite to find his way to Edmonton, perhaps as Danielle Smith’s Multiculturalism Minister*.

Edmonton Glenora figures to be one of the most hotly contested seats – the PCs beat the Liberals by 100 votes last time, but both the NDP and Alberta Party are running strong candidates and have targeted the riding. And with the Wildrose’s Edmonton poll numbers, even they could pull it out on the vote split. This unpredictability is born out in the pool, where 56% expect the PCs to hold the riding, 21% (myself included) see it as an NDP pick-up, 11% pick the Alberta Party, 10% pick the Liberals, and 2% pick the Wildrose.

The Alberta Party

This being their first election, it’s difficult to know what to expect from the Alberta Party. Only one-in-three predict they’ll win a seat, and the mean guess on their best riding vote is 15-20%. That’s the most I’d expect them to get anywhere, but many pool entries had them winning multiple seats, and up to 40 or 50 percent of the vote in some ridings. 

The Senate

The Wildrose Party is predicted to pick up 1.7 of the 3 Senate seats. Personally, I expect a clean sweep – given the low level attention paid to the Senate election, I expect most will vote party lines.

The Morning After

Close to half (44%) expect Liberal leader Raj Sherman to announce his resignation within 48 hours of the vote – perhaps not surprising since most expect Sherman to lose his seat and just 18% think the Liberals will win more votes than the NDP.

One-in-three expect Redford to resign after the election – presumably not the same people voting for her as “the most progressive option”. Personally, I’d be surprised if any of the four opposition leaders are around come next election – hell, I’d be surprised if all four opposition parties still exist come the next election. But my best guess is that they’ll all take a bit of time before coming to a decision…or being forced into one.

*That’s obviously a joke. We all know Danielle Smith won’t name a Multiculturalism Minister.

Alberta Votes Day 24: Sleeping Arrangements at Minority Motel

Posted on by CalgaryGrit in 2012 Alberta Election, Alberta Politics, Seat Projections | Leave a comment

Above: Preview of post-election Cabinet meetings?

Despite a lackluster debate performance by Alison Redford, a pair of post-debate polls show the Wildrose margin down to 7 points. With daily controversies dogging Danielle Smith, there’s no guarantee that lead will hold and, even if it does, it’s impossible to accurately project what the next legislature will look like.

The first problem with any seat projection is that all polls split Alberta into three regions for reporting purposes – Calgary, Edmonton, and rural Alberta. While that’s the most logical cut, just as downtown Calgary and suburban Calgary are two very different entities, lumping Red Deer and Peace River into the “rural Alberta” umbrella masks regional trends.

Moreover, the regional sample sizes in these polls are small, carrying large margins of error. It’s not at all surprising to see the PCs up by 4 in Calgary Monday morning, but trailing by 14 come dinner time. Yes, it’s possible a Wildrose chinook rolled across the city, but it’s more likely we’re dealing with the effects of small sample sizes and varying methodologies.

But even if we’re lucky enough to get completely accurate polls, extrapolating out seat totals is a fool’s errand. Seat projections in the last few federal elections have generally worked well, because we’ve had a good baseline to work with. If the Liberal vote is down 5 points in Ontario from 2008 to 2011, it’s not hard to do the math for each riding. However, the Wildrose Alliance was little more than a protest party last time Albertans went to the polls, so it’s counter productive to calculate their 2012 support using their 2008 vote. When a party rises from nothingness, it’s extremely difficult to predict vote patterns – there’s a reason seat projections showed the NDP winning anywhere from 20 to 70 seats in Quebec last spring, and a reason projects a Wildrose seat range of between 27 and 74 seats.

Toss in wild 4-way splits in Edmonton, incumbency effects, riding redistricting, and a map skewed to favour rural Alberta, and we’re dealing with Alberta’s most unpredictable election in a long time.

So the bottom line is we don’t know what the next legislature will look like, but if the vote stays close, Alberta’s first ever minority government is a distinct posibility. So what happens then?

The first thing to consider is the likelihood of defections. A floor crossing epidemic has spread across Alberta in recent years, so I wouldn’t be at all surprised to see a handful of Progressive Conservatives take a step to the right after the election. Few PC MLAs supported Redford when she ran for leader, and she hasn’t exactly endeared herself to them over the past month. Many in her caucus share the same Wildrose positions Redford spent the campaign ridiculing, and the prospect of an NDP-supported Redford budget might be too much for them to bear.

Hell, if Smith is only a handful of seats shy of a majority, expect an exodus. PC MLAs have never sat on the opposition benches, and many chose their party colours merely because they saw the PCs as the only avenue to power. It’s not like Danielle Smith has her hands full of “Cabinet material” candidates to choose from, so she wouldn’t hesitate for a second to offer Cabinet posts to entice one or two Tories across the floor.

But let’s assume for a moment that Redford holds the PC ship together – maybe the PCs even find themselves ahead of the Wildrosers. The focus then shifts to the Liberals and NDP, who will likely win 4 to 10 seats between them. The Saskatchewan Liberals found themselves in a similar situation in 1996, and their decision to take Cabinet posts in an NDP government eventually led to the demise of that party. So it was somewhat curious when Liberal leader Raj Sherman began the negotiation process during an online debate yesterday, by asking Smith and Redford if they’d name him Minister of Health.

While I don’t think the good doctor will get his wish, nearly everyone I’ve talked to assumes the PCs would cut some kind of deal with what’s left of the left – even if they win fewer seats than the Wildrosers. But I’m not convinced.

The Liberals and NDP have spent decades waiting for the PC empire to crumble and if Smith offers them a deal – say some democratic reform measures and a hold on the more objectionable parts of Wildrose platform – I wouldn’t be shocked to see an unholly alliance. After all, seeing the Wildrose Party in power would surely plunge the PCs into infighting and a leadership race.

For the Wildrose Party, the benefits of a deal with the left are obvious – the aforementioned PC chaos, and a chance to govern “non-scarily” thanks to the calming influence of the Liberals and NDP (think Stephen Harper from 2006 to 2011). Keep in mind, Wildrose campaign manager Tom Flanagan has gone on record as stating he saw nothing wrong with Harper’s threat to cut a deal with the separatists and socialists in 2004, after winning fewer seats than Paul Martin. Smith sounded like she was laying the groundwork for this type of arrangement during the leaders debate, when she went out of her way to compliment Raj Sherman and talked about how she’s been able to work with the NDP in the legislature.

As strange as it may sound, life may soon be imitating the scenes from a satirical 2010 press gallery video, featuring Brian Mason barking orders (“left, left, left“) at Danielle Smith and a Wildrose caucus decked out in orange NDP shirts. Hell, the video even features a cameo from then-independent MLA Raj Sherman.

Mulcair’s win was inevitable…but was it predictable?

Posted on by CalgaryGrit in Federal Politics, Fun with Numb3rs, NDP Leadership 2012 | Leave a comment

PunditsGuide concludes her top notch NDP leadership coverage with the vote breakdown of advance and in-person ballots. White Nathan Cullen “won” the convention vote, Mulcair had enough support among advanced voters that he could have dropped his pants and sung La Marseillaise during his Friday night speech, and likely still carried the day.

So the question is – if a Mulcair win was inevitable, was it predictable?

I compiled a series of 9 leadership metrics in the days leading up to the convention on attributes such as fundraising, endorsements, social media, and polling data. In my mind, this was a bit of a test drive for the Liberal leadership race, since I’d like to be able to post weekly or monthly “Power Rankings” during that contest, to gauge candidate support and momentum.

So here’s how the different indicators stacked up, sorted by their correlation to the first ballot results:

Poll: 0.899
Endorsement Score: 0.869
Media mentions: 0.868
MP Endorsements: 0.860
Donations: 0.836
Twitter mentions: 0.715
Donors: 0.683
Facebook “likes”: 0.308
Twitter followers: 0.170

As has been the case in past leadership contests, polls among party members proved to be the most accurate prediction of candidate support. And remember, in this case we were dealing with a pair of month-old polls conducted by two of the leadership camps. Of the two, Mulcair’s proved to be more accurate – mainly because (surprise, surprise) Paul Dewar’s overstated his support by a factor of 2. Let that serve as a cautionary tale anytime you see an “internal” campaign poll released.

Endorsements also score well, using either a simple count of MPs, or’s more complex system. Mind you, Premier Mar and Premier Falcon will tell you how much endorsements can be worth in some contests.

Fundraising numbers were a good indicator of candidate support, though not as much as the 2003 NDP leadership contest when they had a remarkable correlation of 0.998. As has been the case with the other half dozen leadership races I’ve crunched the numbers on, the total amount raised seems to be a better predictor than the total number of donors. Goes to show money matters, even when we’re talking about socialists.

As for social media? It may play a big role in modern elections, but it’s certainly not a very good indicator of who’s in front. Language may have played a role in it, but Mulcair was 5th in terms of Facebook “likes” and was barely ahead of Niki Ashton when it came to Twitter followers.

Surprisingly enough, something as simple as the number of times a candidate is mentioned by the media proved to be just as good a predictor of support as the factors above. So while we all like to think the media has no clue what goes on during internal party contests, they were basically on the mark this time. Goes to show you there’s some truth behind what’s heard at Hy’s.

If we toss all these numbers into the regression wood hopper, the best prediction models weights endorsements, donors, polling, and media buzz about evenly, and explains 92% of the variance. That’s a good number, but my stats teachers would slap me if they saw me creating a “magic regression equation” based on 8 candidates in one leadership race. After all, each contest has its own rules and personality, and it’s foolish to think the Liberal race will look like the NDP’s.

Still, the signs that mattered most all pointed to a significant first ballot lead for Mulcair. We shouldn’t be at all surprised that’s what happened.

Crunching Numbers

Posted on by CalgaryGrit in Fun with Numb3rs | Leave a comment

Now that the shock of Lise St-Denis’ floor crossing has worn off, it’s worth looking ahead to whether Jean Chretien’s old riding will stay Liberal red in the next election. Eric Grenier at 308 likes her chances:

If Lise St-Denis, the MP for Saint-Maurice–Champlain who defected to the Liberals from the NDP last week, decides to run for re-election under her new party’s banner, history suggests that, despite the wide margin between the two parties in the last election, she would have a good chance of winning the seat again.


Of those 34 floor-crossers who have stood for re-election, 22 were successful – including Richard John Cartwright. The success rate of floor crossers stands at 65 per cent.


More than half of floor crossers lose support when they run for re-election with their new party, with their average drop in vote share being eight points from one election to the next. However, the parties that welcome the new MPs almost always do better when the floor crosser runs for re-election. Fully 86 per cent of floor crossers have increased the vote share of their new parties in their riding, increasing their party’s support by an average of 11 points.

Ms. St-Denis received 39 per cent of the vote in Saint-Maurice–Champlain in May 2011, while the Liberals took 12 per cent. Applying these numbers to the average change in support when an MP has crossed the floor indicates she might expect to receive somewhere between 23 and 31 per cent of the vote in the next election, perhaps only enough to win a narrow victory.

There are a lot of ways to crunch the numbers, but before we try to draw conclusions based on 34 incredibly different data points from the past hundred plus years, we should probably look at the numbers that matter most – the votes in St. Maurice Champlain last election:

NDP 39%
BQ 29%
CPC 18%
Lib 12%

In terms of vote percentage, the riding ranked 33rd for the Liberals in Quebec (a sign of just how awful they did in La Belle Province last election). Maybe their electoral chances are better than that though – the 27 points the Liberals lost by makes it their 24th “best” riding in Quebec.

That’s not to say the Liberals can’t ride a 20 to 30 point red wave and win the seat next election, but it’s preposterous to argue St-Denis herself will have anything to do with that. Well known and well liked incumbents tend to be worth about 5 percentage points at the ballot box, and I wouldn’t use either of those terms to describe St-Denis. After all, she herself admits to being a no-name MP who was elected on Jack Layton’s coat tails.

The problem with looking at the historical numbers is that there’s little historical precedent for what St-Denis did. When Belinda Stronach moved from the Conservatives to the Liberals in 2005, she was jumping between parties which had been within 700 votes of each other in the previous election. In comparison, St-Denis just “gave up” 13,000 votes with her move. The more apt analogy might be kamikaze PC MP Jack Horner who won the bluest of blue ridings in Canada, Crowfoot, by 61 percentage points over the Liberals in 1974. He crossed the floor to take a position in Pierre Trudeau’s Cabinet and in the next election…lost by 59 percentage points. Yes, Jean Chretien’s old riding isn’t as hostile to the grits as rural Alberta during the Trudeau years, but Horner was a PC leadership candidate and had represented the riding for 20 years. All that was worth a few votes at the ballot box.

Let’s have some fun with numbers and create a single variable to measure just how much of an impact floor crossers have. I’ll take a page from baseball stat geeks and call it VORC – Value Over Replacement Candidate. Taking the Belinda Stronach example, the Liberal vote in the ridings around Newmarket Aurora fell 4 points between the 2004 and 2006 elections. So the fact that the Liberal vote actually went up by 5 points in Newmarket Aurora with her name on the ballot means she may have been worth 9 points. Of course, there are a gazillion other factors to consider, but it’s the best quick and dirty estimator of a candidate’s value we have.

So Belinda may have been worth 9 points at the ballot box, but using the same math, Garth Turner was worth a big “0” points to the Liberals in Halton in 2008, while Wajid Khan actually cost the Tories 3 points in Mississauga-Streetsville (KHAAAAAAAN!). So there’s no hard and fast rule.

Which, after that mathematical detour, is the point I’m trying to make with this post. When it comes to unique political events like by elections or floor crossings, it’s futile to look at past trends and averages to predict the future. Regardless of what may have happened to Jack Horner or Belinda Stronach or 1869 floor crosser Richard John Cartwright, Lise St-Denis’ situation is wholly unique – there aren’t many cases of first time MPs jumping to the party which finished fourth in their home riding.

So without precedent, all we can go on is political instinct and my political instincts tell me St-Denis’ value in the next election (should she even run, which seems unlikely) is negligible. That’s not to say her jump to the Liberals won’t help the party create a more favourable narrative province-wide, but on the ground in St. Maurice it won’t make a lick of difference come the 2015 election.

Redistribution Winners and Losers

Posted on by CalgaryGrit in Fun with Numb3rs | Leave a comment

Even though it was the right decision, the Harper government likely didn’t do itself any electoral favours by shutting Quebec out of the new ship building contract and the 15,000 jobs that go with it. However, they did announce two new jobs they will be creating in Quebec:

Ontario and B.C. will get fewer new seats than the Harper government originally promised, according to the Conservatives’ latest plan to redraw regional representation in the House of Commons. The move, meant to address growing populations in B.C., Ontario, Quebec and Alberta, adds seats in all of those provinces. Compared to an earlier plan, though, B.C will receive five rather than seven new seats, while Ontario will get 13, down from the 18 it was originally promised. The Tory stronghold of Alberta will receive six new seats, and two more will go to Quebec, which had argued its representation was disproportionately low compared to English Canada in the Conservative’s original redesign.

While everyone will try to predict what this means electorally, a lot will depend on what the new ridings look like. Even in a place like Saskatchewan (who will stay at 14 seats) the new boundaries could be a game changer. Redrawing the nonsensical rurban ridings that go from downtown Regina and Saskatoon to 100 miles outside city borders could cost the Tories seats…or it could mean the end of Ralph Goodale’s reign as the king of Wascana. We won’t know the new boundaries for another two years, so anything before then is just speculation.

But speculation is fun, and it’s better done using numbers than the guesswork we’ve seen in many newspapers this week. So I’ve taken a crack at analyzing the electoral implications of the new ridings in what is sure to be one of the most exciting edge-of-your-seat posts in the history of Calgary Grit.

Identifying New Seats

To figure out where the new seats will go, I took the average riding population post-redistribution and carved anything above this number off into new ridings. For example, once Alberta grows from 27 to 33 seats, the average riding population will be 87,000. Ridings smaller than this are assumed to stay the same, while 30,000 “excess” voters from Calgary Centre get sent to a new ridings…when tallying the numbers, this means Calgary Centre gets counted as about a third of a “new” riding.

In short, we know big chunks of big ridings will be carved up to make new ridings. This analysis identifies just how big those chunks are.

What It Means

Using this method, the popular vote for the 26 new ridings breaks down as follows:

CPC 46.9%
NDP 26.8%
Lib 20.0%
Green 4.6%
BQ 1.7%

That spells good news for the Tories and bad news for the NDP, but that’s mostly a reflection of Alberta getting three times as many new seats as Quebec. Within each province, the support level for each party in the “new” ridings is quite similar to their province-wide numbers.

It’s risky to project seats since a lot will change in 4 years, but based on 2011 election numbers, the new seats break down as follows:

CPC 17.8
NDP 5.7
Lib 2.3
BQ 0.1

Things will change, but this plan adds a de facto 10 seat cushion to the Tory majority. That’s a big deal.

This will make the job of the Liberals and NDP harder in 2015, but they have no one to blame but themselves. The Liberals especially, have written off Western Canada for over 30 years, even though the electoral math makes that strategy riskier and riskier as time goes by. Going 0 for 19 in Alberta in 1972 is one thing…going 0 for 33 in 2015 makes it that much harder to form government.

One Final Note

Even though the addition of these new seats will benefit the Tories, these gains may be offset by other intangibles. Redistribution hurts incumbents, since new borders will force some MPs to campaign in neighbourhoods where they are unknown to voters. Given the Tories have the most incumbents, this works against them.

Redistribution also risks opening up is a Copps-Valeri situation, where MPs are forced up against each other. There aren’t many benefits to only having 34 MPs, but this is a problem the Liberals won’t have to worry about this go around.

Plugin from the creators of Brindes Personalizados :: More at Plulz Wordpress Plugins