not so fast...SES POLL

[Montgomery Burns]

I think the Ipsos-Reid is more accurate. I believe the Conservatives have about a 10 point lead. I saw a Martin rally on CPAC yesterday and all the people standing behind him looked glum for much of his speech. Harper's rallies have been enthusiastic and boisterous while attracting 1000, 1200, 1400, and 1500 people. The Conservative base is motivated; I don't think one can say that about the Liberal base. Layton has conceded that the Conservatives are going to win. These parties do their own internal polling; they know what's going on. The only question is whether is will be a minority or majority victory.

[Canuck E Stan]

Now here's a poll,that's got it all.

Margin of error=/-1% with 10,000 participants.

From Kinsella's blog:

Was on CFRA this morning, and I got asked – as I have been too many times in this election – about SES. My advice to Steve Madely: read Ipsos. Here's the Ipsos release headline from last night, in case you don't want to read it all: "Conservatives On Cusp of Majority…Liberals and Bloc Battle it out for Offical Opposition As NDP Makes Impressive Gains." Here’s their seat model: "Conservatives -148-152; Liberals - 62-66; NDP - 34-38; and Bloc - 56-60." Here’s their poll percentages: "Conservatives 38%; Liberals 27%; NDP 19%; Bloc 12%; Green 4%." HERE’S THEIR SAMPLE SIZE: 10,500, Margin of Error +/- 1%. Got that? A SAMPLE SIZE OF MORE THAN 10,000 CANADIANS WHO ARE COMMITTED VOTERS. A M.O.E. OF ONE PER CENT. PLEASE, PLEASE, LET’S STOP TALKING ABOUT SES. THEY’RE GREAT PEOPLE AND ALL THAT. BUT STOP. STOP

[shoop]

Good point Stan.

There have been an number of issues with SES throughout this campaign.

Big Gunner whined about being attacked, but wouldn't define a *narrow minority* government.

What did he mean by that?

SES is majorly offside from other polls showing a 10-point CPC lead.

Not attacking anybody but you have to wonder about somebody posting the most maligned pollster in the campaign under the tittle (not so fast....)

[normanchateau]

Based on the 2006 election results, it appears than the January 22nd SES poll was the most accurate predictor of the popular vote and Democratic Space was the most accurate predictor of seats won. And the January 22nd Ipsos-Reid poll and the January 22nd Strategic Counsel poll were both way off a mere day before the election. Good job SES.

[shoop]

SES was uncannily close.

Off by *exactly* 0.1% for each of the four major parties. Unfreakingbelievable.

There methodology seemed a little whacky, but they got it right. Kudos to them.

[geoffrey]

SES was uncannily close.

Off by *exactly* 0.1% for each of the four major parties. Unfreakingbelievable.

There methodology seemed a little whacky, but they got it right. Kudos to them.

Their rolling 3 day that is... their night before single day was off by a considerable margin.

[shoop]

Your point being?

They only released the 3-day rollings during the campaign. They readily admitted that their single days had far to great a margin of error...

Their rolling 3 day that is... their night before single day was off by a considerable margin.

[Hicksey]

I think one of the bigger surprises was the big swing in Ontario. The night before the election Ontario was in favor of the CPC by a 4 point margin. When the dust settled, they lost the popular vote in Ontario by nearly 5 points.

Wonder why nobody called that?

[shoop]

SES probably came the closest in Ontario as well.

They called the Liberals at 38% in Ontario and the Liberals came in at 39.9%.

They called the CPC at 36% in Ontario and the CPC came in at 36%.

[normanchateau]

Democratiic Space was remarkably close in their seat projections.

[shoop]

Didn't he miss 20 or 25 seats?

[normanchateau]

Didn't he miss 20 or 25 seats?

Here's what Democratic Space projected on January 22nd:

CPC 128 seats

Liberal 94 seats

BQ 56 seats

NDP 29 seats

Independent 1 seat

Most of us would be pleased with ourselves if we'd been that close the day before.

[shoop]

Take a look at his individual seat projections. He did miss 20 or so of his individual seat projections.

He couldn't even be arsed to take the time to point out which individual seats he missed. Not too classy....

At leat electionprediction did that. They missed 30 seats individually probably a few more than democratic space missed. At least they had the gumption to point out how many seats they missed very prominently on their Web site.

Here's what Democratic Space projected on January 22nd:

CPC 128 seats

Liberal 94 seats

BQ 56 seats

NDP 29 seats

Independent 1 seat

Most of us would be pleased with ourselves if we'd been that close the day before.