Post-Mortem: Elections

• The Saskatchewan Party won the Saskatchewan Election (also see my post Saskatchewan Election v2).  Last September, I said they had a 90% chance of winning, which I recently increased to 95%.  I might have been slightly overconfident.
• The PCs won the Manitoba Election (also see my post Manitoba Election v2).  Last September, I said there was an 80% chance that they would win the election, which I increased to 91% in March.  These numbers seem about right.
• Mulcair lost the NDP leadership review (see my post Mulcair NDP Leader).  In December, I said there was a 65% chance that Mulcair would lose.  That might have been a bit low, given the low level of support received by Mulcair.

Manitoba Election v2

Which party will the most seats in the next Manitoba provincial election?
PC: 91%
NDP: 5%
Liberal: 4%

Since my last post, the PCs and NDP have slipped slightly in the polls, and the Liberals have gained slightly (see article here).  The PCs still have a substantial lead however, so I am increasing my confidence for a PC win.

Year-End Post-Mortem: Global Leaders and Elections

• Pierre Nkurunziza has kept the presidency of Burundi (see my post Nkurunziza in Burundi).  In May, I said there was a 35% chance of him leaving office, which may have been too conservative.
• Tammam Salam has remained leader of Lebannon (see my post Tammam Salam Leaving Office).  In August, I said there was a 30% chance he would leave office, which might have been too high, since these leader exit questions rarely trigger.
• The People's Party won the most seats in the Spanish Election.  In June, I said there was a 30% chance of this; I probably overestimated the chances of Podemos winning the most seats at that time.
• The Venezuela election was held on December 6th, and resulted in a victory for the MUD coalition (see my posts Venezuelan Election Date and Venezuelan Election Result).  In June, I said there was a 75% chance that an election would be held in 2015.  In July, I said there was a 70% chance that MUD would win the election.
• As far as I can tell, the For the Love of Egypt bloc won the most seats in the Egyptian Parliamentary Election.  I said there was a 35% chance that this would happen, and this was the most likely outcome according to my scoring.
• There was no Libyan Unity Government in 2015.  In October, I said there was a 20% chance such a government would be formed.

Saskatchewan Election v2

Which party will win the most seats in the next Saskatchewan general election?
Saskatchewan Party: 95%
NDP: 5%

The Saskatchewan Election will be held on April 4th, about a month from now (see my previous post here, and 308's coverage here).  The Saskatchewan Party has slipped slightly in the most recent poll, and the NDP have gained; the Saskatchewan Party now only has a 52% to 34% lead.  However, the NDP are running out of time to make further gains, so I am increasing my confidence in a Saskatchewan Party victory.

Austrian Presidential Election

Who will win the Austrian Presidential Election?
Alexander Van der Bellen: 35%
Rudolf Hundstorfer: 20%
Irmgard Griss: 15%
Norbert Hofer: 15%
Someone Else: 15%

The Austrian Presidential Election will be held in April, with a second round in May (see Wikipedia Article here).  Alexander Van der Bellen seems to be currently leading in the polls, with Irmgard Griss in second and Rudolf Hundstorfer in third.  I'm giving Hundstorfer higher odds because he seems to be associated with a major party.

Peruvian Election

Who will win the next Peruvian presidential election?
Keiko Fujimori: 60%
Julio Guzmán: 10%
Pedro Pablo Kuczynski (PPK): 10%
César Acuña: 10%
Someone Else: 10%

The first round of the Peruvian election is scheduled for April (with a second round in June).  The most recent poll gives Fujimori about 39% of the vote, Guzmán about 20%, PPK about 13%, and Acuña about 9% (see Wikipedia article here).  Fujimori is the clear favorite; she only narrowly lost the last election, and has a huge lead in the polls.  Guzmán has rapidly climbed in the polls recently, so I think that he is less likely to win than his standing in the polls suggests.  PPK and Acuña seem to be the main challengers; both could plausibly win if they gain a lot in the polls.  Also, one of the minor candidates could gain enough ground to win; I do not know enough about Peruvian politics to rule this out.

Dominican Republic Election

Who will win the Dominican Republic Presidential Election?
Danilo Medina: 77%
Luis Abinader: 20%
Someone Else: 3%
(If one of the named candidates dies or is incapacitated and is explicitly replaced by another candidate, that candidate success shall count as the named candidates success)

The Dominican Republic presidential election will take place in May, with a second round if necessary in June.  The current president, Danilo Medina, appears to have a significant lead, with around 50% of the vote, in comparison to 35% for his nearest challenger Luis Abinader (see article here).  I think the incumbent will likely win.

Haitian Election

Who will win the Haitian Presidential Election?
Jovenel Moïse: 75%
Jude Célestin: 20%
Someone Else: 5%
(If the current scheduled runoff is rescheduled, then the result of the rescheduled election will count, as long as it is held in 2016.  Initial results count as long as they clearly indicate a winner, even if they are later overturned)

The first round of the Haitian Presidential Election was held on October 25th, and resulted in Jovenel Moïse and Jude Célestin getting enough votes to get to the second round (see article here).  Jovenel Moïse is the favoured candidate of the current president, and Célestin and other opposition figures have accused him of widespread fraud.  Célestin in particular has refused to campaign for the runoff election.  Given that Moïse got the most votes in the first round, and Célestin refuses to campaign, it looks likely Moïse will win.  It is also possible (though unlikely) that someone else could win if the first round election results are somehow invalidated.

Spanish PM v2

To which party will the next Spanish Prime Minister belong?
People's Party (PP): 40%
Socialist Workers' Party (PSOE): 59%
Some Other Party: 1%

The Spanish election has resulted in a plurality, but not a majority, for the conservative People's Party (see article here, and my earlier prediction here).  A conservative coalition does not seem possible, but a left-wing coalition also seems to be difficult to organize.  I think that a left-wing coalition of some type will eventually be cobbled together.

Overall Riding Prediction Performance

Overall, my predictions for Ottawa ridings were generally not that accurate.  The Brier score for my initial round of predictions was 0.61, versus an expected Brier score of 0.28 (lower is better for Brier scores).  This suggests that I was overconfident in some of my predictions.  In fact, if I had just given 50% odds to the top two parties, I would have gotten a Brier score of 0.56 (I initially put the Liberals as the third most likely party to win in Gatineau, so the average Brier score would be above 0.5).  Having said that, since my riding errors were so heavily correlated (i.e. I consistently underestimated the Liberals), my first round of predictions were closer to a single prediction underestimating the Liberals than a bunch of independent incorrect predictions.  This limits the strength of the conclusion I can draw.

I did much better in my second round of predictions, achieving a Brier score of 0.33 versus an expected score of 0.31.  My second round predictions were evidently much better calibrated and more accurate.  I evidently underestimated the importance of changes in the polls during the campaign.

Election-Related Post-Mortem: Quebec and Rural Ridings

• The Liberals won in Hull—Aylmer (see also Hull-Alymer v2).  I initially gave them 13% odds, which I later increased to 45%
• The Liberals won in Gatineau (see also Gatineau v2).  I initially gave them 2% odds (I actually thought they were less likely to win then the BQ), which I later increased to 15%
• The Liberals won in Pontiac (see also Pontiac v2).  I initially gave them 12% odds, which I later increased to 65%.
• The Liberals won in Glengarry—Prescott—Russell (see also Glengarry—Prescott—Russell v2).  My odds went from 35% to 70%.
• The Conservatives won in Stormont—Dundas—South Glengarry.  I gave them 90% odds.
• The Conservatives won in Leeds—Grenville.  I gave them 90% odds there as well.
• The Conservatives won in Lanark—Frontenac—Kingston.  I gave them 92% odds.
• The Conservatives won in Renfrew—Nipissing—Pembroke.  I gave them 94% odds.
• The NDP won in Outremont.  I gave them a 65% chance of winning there.

The biggest miss was definitely in Gatineau.  The NDP won in 2011 in Gatineau with almost 62% of the vote, and they had a very strong candidate in Françoise Boivin.  The size of the vote swing there was much larger than I expected, and rather surprising.

Election-Related Post-Mortem: Ottawa Ridings

• The Liberals won in Ottawa Centre (see also Ottawa Centre v2).  I originally gave them a 4% chance of winning, which I later increased to 30%.
• The Liberals won in Ottawa-Vanier (see also Ottawa-Vanier v2).  My odds went from 90% to 97%.
• The Liberals won in Ottawa-West Nepean (see also Ottawa-West Nepean v2).  My odds went from 55% to 90%.
• The Liberals won in Ottawa South (see also Ottawa South v2).  My odds went from 95% to 99%.
• The Liberals won in Orleans (see also Orleans v2).  My odds went from 70% to 87%.
• The Liberals won in Nepean (see also Nepean v2).  My odds went from 25% to 65%.
• The Liberals won in Kanata-Carleton (see also Kanata-Carleton v2).  My odds went from 20% to 72%.
• The Conservatives won in Carleton (see also Carleton v2).  My odds went from 90% to 85%.

The most significant miss was definitely in Ottawa Centre, although I think it was justifiable in context.

Election Related Post-Mortem: National Results

• The Liberals won a majority in the federal election.  In my first post in January, Federal Election Result, I gave a Liberal majority a 20% chance, and a Liberal government a 60% chance.  In my second post in April, Federal Election Result v2, I gave a Liberal majority a 10% chance. and a Liberal government a 40% chance.  In my third post in May, Federal Election Result v3, I stayed with the 10% chance for a Liberal majority, but decreased the chance of a Liberal government to 30%.  I don't think these numbers were too far off; the Liberal surge late in the campaign was genuinely surprising.
• The party that won the most seats is indeed forming the government (see my post Next Government Plurality).  I said this had a 93% chance of happening.  This question would have been more interesting under a minority government scenario
• The Federal Election Voter Turnnout was 68.5%.  I gave 5% odds to a turnout higher than 65%.  I was definitely somewhat overconfident about this question, although the model I used suggested I should have been even more confident (I was justifiably somewhat sceptical of the model, and maybe I should have been even more sceptical).  The high turnout at the advance polls was perhaps more meaningful than I thought, as was the perceived closeness of the election.
• The Conservative Polling Error was about 1% (the Conservatives were projected to get 30.9% of the vote, and actually got 31.89%).  I gave a 65% chance of a greater than 0% polling error, and a 30% chance of a polling error between 0% and 2%.  The Conservatives thus outperformed the polls to about the extent I expected (although they didn't get as many seats because of the Liberal out-performance).
• The CBC called the election before 10pm (see my post Timing of CBC Election Call).  I gave them 40% odds of calling the election before 11pm.  The results came in faster than I expected.

Election Related Post-Mortem: Campaign

• There were five Federal Election TV Debates.  I said there was a 60% chance of at least this number of debates, which might have been slightly overconfident given how much uncertainty seemed to be associated with the process.
• Neither Stephen Harper nor Rob Nicholson announced that they would retire before the federal election (see my posts Stephen Harper Retirement and Rob Nicholson Retirement).  I said Harper had a 35% chance of such an announcement, which was definitely too high in the absence of suggestive evidence.  I gave Nicholson a 15% of retiring, which was more reasonable.
• Negotiations have concluded on the Trans-Pacific Partnership (also see my Canada-specific predictions Trans-Pacific Partnership Canada and Trans-Pacific Partnership Canada v2).  I originally said a deal had a 15% chance of being reached between September and December, and a 35% chance of being reached in 2015.  I also said a deal involving Canada had a 20% chance of being reached in 2015, which I later increased to 45%.  I may have been a bit too pessimistic about the likelihood of a deal being reached, especially one involving Canada.
• Mulcair did not renounce his French citizenship (see my post Mulcair French Citizenship).  I gave him a 10% chance of doing so.

Questions Resolved in October: International Elections

A lot of questions have resolved in October, mainly related to the federal election.  I'm going to devote this week to going through them, starting with international election questions.

• The Centre Party won the Finnish Election.  I said they had a 70% of winning, which seems about right.
• The Portugal Ahead coalition won the Portuguese Election (also see my second post Portuguese Election v2).  I originally said the Social Democratic Party (the largest party in the Portugal Ahead coalition) had a 29% chance of winning the most seats, which I later increased to 40% after the coalition formed.  The formation of a coalition is one of those awkward events that can increase the odds of a party that is well behind winning an election.
• The new Portuguese prime minister, Pedro Coelho, belongs to the Portugal Ahead party, which I said had a 15% chance of happening.  I expected the left-wing parties to form an a coalition government, which did not happen.
• The Swiss People's party won the Swiss Election.  I said they had a 60% chance of doing so.
• Jimmy Morales won the Guatemalan Election (also see my other posts Guatemalan Election v2 and Guatemalan Election v3). In my original post, I didn't list him as a candidate, although I did give "Someone Else" a 15% chance of winning.  My second post listed him and gave him a 30% chance of winning, while my third post (after the run-off) gave him a 65% chance of winning.  I might have underestimated Morales, but his victory does seem to be genuinely surprising.
• Alpha Condé won the Guinean Presidential Election.  I gave him a 75% chance of winning, which might have been too low given how popular he seemed to be.

Burmese Election v2

Q141: What party will win the most seats in the House of Representatives in the next Burmese election?
UDSP: 25%
NLD: 73%
Some Other Party: 2%
(The rules for my previous post still apply)

The Burmese election will be held tomorrow (see article here).  As discussed in my last post, the opposition NLD is expected to do very well, although there are some concerns that the vote will be rigged or not recognized by the ruling UDSP party.  Given how smoothly the election has gone so far, I am increasing my odds for the NLD.  I am also decreasing my odds for the other option, as no other credible party seems to have emerged.

Pontiac v2

Liberal: 65%
NDP: 25%
Conservative: 10%

Glengarry—Prescott—Russell v2

Liberal: 70%
Conservative: 30%

Gatineau v2

NDP: 85%
Liberal: 15%

Hull—Aylmer v2

NDP: 55%
Liberal: 45%