The Last Poll

There was lots of speculation in the penultimate days before the Canadian election about which polls were the most accurate, and subsequently whose polling companies would be toast. Much speculation went on with the Blogging Tories that SES which did not give their party the big lead was out to lunch. Progressive Bloggers on the otherhand took heart over the SES polls. As it was SES faired the best being the closest to the actual vote count. Damn close. But for you poll junkies who need one more hit here is an interesting American site with an analysis of the polls. Political Arithmetik

And while we are speaking of polls the UBC Election Stockmarket Poll did amazingly well showing that a real marketplace in votes could replace Bay Street. Or was it Bay Street could marketize voting. If so you would not lose money betting on SES.

How well did we do? ESM vs. Pollsters (Popular Vote Predictions)
	LIB	CPC	NDP	BLQ	OTR	Absolute Error
1. SES Research (Jan. 22)	30.1	36.4	17.4	10.6	5.6	0.4
2. UBC ESM (Jan. 22)	28.1	37.6	17.8	10.5	6.0	4.2
3. Strategic Counsel (Jan. 22)	27	37	19	11	6	6.3
4. Ekos (Jan. 20)	26.9	37.1	19.5	11.5	4.6	8.1
5. Ipsos-Reid (Jan. 22)	27	38	19	12	4	9.6

How well did we do? ESM vs. Other Seat Projections
	LIB	CPC	NDP	BLQ	OTR	Absolute Error
1. Milton Chan (electionprediction.org)	104	118	29	56	1	12.0
2. UBC ESM	95.5	125.2	32.6	53.4	1.2	15.0
3. Gregory Morrow (democraticspace.com)	94	128	29	56	1	18.0
4. SES Research	84	134	34	55	1	36.0
5. Laurier Institute (LISPOP)	78	140	33	56	1	50.0
6. Jordan O'Brien (jord.ca)	72	135	38	62	1	62.0
7. Strategic Counsel	56	149	41	61	1	94.0
8. EKOS (Jan 20)	53	151	41	62	1	100.0
9. Ipsos-Reid (Jan 21)	46	157	42	62	1	114.0

Tags
Canada
Federal Election
Politics
Polls
Polling
Canadian Election