Coda (III)

Iberville and Champlain both fall into the PQ column; the former quite decisively so.

Coda (I)

Wow, Angus Reid nearly nailed it. In your face, Léger and CROP.

Unliveblogging the Quebec election

Well, that was less exciting than the last time.

9:00 — Tale of the tape: The ADQ won 44% of the vote in the 41 ridings they won last spring. Tonight, they are at 27% in those same ridings. The PQ is leading in 19 of the 41, the PLQ in 16. Mario Dumont is clinging to 6.

9:02 — PLQ is ahead in the Maggies.

9:04 — The adéquiste is running third in 24 of the 41 ADQ-won ridings from 2007.

9:14 — Amir Khadr is holding to a narrow lead for Québec solidaire in Mercier, which is one riding over from Mulcair's federal NDP riding. Françoise David is trailing narrowly for QS in neighbouring Gouin. They are also on the board in Outremont, Laurier-Dorion, Sainte-Marie—Saint-Jacques, Hochelaga-Maissonneuve, and Rosemont.

9:14 — Chevarie is clinging to a narrow PLQ lead over the PQ down in the islands.

9:17 — How many guys are there named Jean-Yves Roy in politics down in the Matapedia-Matane area, anyway?

9:20 — Hey, CBC: the Lower North Shore polls always come in early. Lorraine Richard is racking up an absolute majority of the vote for the PQ in Duplessis, just like in the old days. Another election, and she'll be in Denis Perron country.

9:21 — Hmm... the electoral map at this hour, provincially, looks an awful lot like it did at this hour, federally, in November 2000.

9:30 — See:

9:36 — Well, crud; does this mean a spherical three-D vote model for the 2012 election?

9:39 — With a biggish number of polls still left to go in Dubuc, the PQ margin of lead is two votes.
9:44 — No huge surprise, but Maka Kotto's blue parachute has landed him safely in Bourget.

9:48 — The PQ owns Saguenay—Lac-Saint-Jean again.

9:56 — In ridings that the ADQ won in 2007, the ADQ is down 17%, the PLQ up 10%, and the PQ up 7%. In ridings that the PQ won, the ADQ is down 14%, and the other parties are each up about 6%.

10:00 — PLQ pickups from the PQ: Les Iles, Gaspé, Rouyn-Noranda-Témiscamingue, Abitibi-Est. The eastern and western extremities of the province are redder than they have been in years.

10:03 — Bill Clennett is on the board in Hull for Québec solidaire. Without googling, who is Bill Clennett?

10:07 — Charest takes 58% in Sherbrooke to the PQ's 29%; slightly more comfortable than the 37%-33% margin last year.

10:11 — The ADQ's least worst showing is down in the Maggies, down from 5% to 1.5%, for a loss of "only" 3.5%. Next time... next time.

10:12 — Y'know, Charest is only a couple of by-elections away from losing that hard-earned second, non-consecutive, majority.

10:23 — The final tale of the tape: the Liberal vote share is up in 121 ridings, the PQ in 117, the ADQ in none.

Nighty-night!

OK, maybe not:

11:04 — Dubuc ends up in the PLQ column by about two points.

63towin (VII)

The evolution of the National Assembly seat projections since the last election, as given by running the polls graphed below through two differents votes-into-seats models. Each data point represents the average of the two projections for each party in each poll. For graphic purposes, the mid-point of the polling period is used as the date of the poll. The solid line is a four-poll moving average of seat projections. The vertical scale is in cohorts of nine; 63, an even multiple of nine, is a bare majority:

And again, the same data since the start of the campaign:

63towin (VI)

The evolution of the provincial vote-intent in Quebec since the March 2007 election. Each data point represents a poll by CROP, Léger, or another pollster. For graphic purposes, the mid-point of the polling period is used as the date of the poll. The vertical scale shows popular support, in percent. The solid lines are the four-poll moving averages:

And the same data, showing the change during the polling-heavy period since the election was called on November 6:

63towin (V)

The latest, and presumably last, batch of polls are in on the provincial vote-intent question in Quebec, which goes to the "urns" tomorrow.

First up, the outfit whose figures are usually taken as gospel, CROP. Unfortunately for poll junkies they were in the field the longest, and longest ago; they had the popular vote pegged at PLQ 45, PQ 29, ADQ 15, between November 28th and December 3rd, inclusive. Those numbers project the following seats:

Next, Léger Marketing, with the later and narrower field period of December 2nd-4th. Léger pegged the vote intent at PLQ 45, PQ 32, and ADQ 15, yielding the following seat count:

And finally, the two-day Angus Reid figures, also the latest, being collected on December 4th and 5th. Alas, they were also collected in an on-line sample, not a phone poll. They put the parties' support at PLQ 42, PQ 36, and ADQ 13, generating the following makeup for the National Assembly:
Taken at face value — a dangerous game with three different firms, three different methods margins of error and all that; but hey, let's have fun — these figures might suggest a race that is narrowing significantly between the Charest Liberals and Marois Péquistes in its final hours. Indeed, the Angus Reid numbers put the PLQ on the edge of losing the majority, even the whopping majority, that has been within Charest's grasp for most of the year, and most of the campaign.

They also suggest the ADQ on the edge of kablooieville, with most of the seats that Dumont won in 2007, tilting towards the Liberals in every election-period poll up until Angus Reid. The Angus Reid poll would give the PLQ and PQ an almost even split of the ADQ's entrails, a scenario which looks an awful lot like Philippe Gohier's 50-50 split, except that the PQ can pick up more than half of the ADQ's seats by picking up less than half of the vote that the ADQ is bleeding... but the PLQ has to have a five or six point lead over the PQ in voter-pickups from the ADQ, if it is to split the difference in seats with the PQ.

Charest still has the edge tonight, Election Eve... but by how much? And can Mario Dumont salvage enough seats to continue the minority status of the legislatur?

On verra sous peu.

63towin (IV)

On debate day, two projections for the price of one! First, the Léger figures for fieldwork between November 18th and 23rd inclusive, reported in today's Gazoo:




And the slightly-longer fieldwork done by CROP from the 17th to the 23rd, reported by La Presse:


These charts are actually the average of two models, one of which shows the ADQ winning narrowly in up to four seats, and the other, well, not so much.

63towin (III)

The latest Léger numbers, crunched. Clever colour scheme explained here.

63towin (II)

The latest CROP numbers (November 6-13) run through Calcutron yield this seat breakdown in the Quebec National Assembly:


For an explanation of the clever colour key, see the first posting in the series: 63towin (I)