Winnipeg's Probe Research Inc. issued a poll grading the support of the candidates for mayor, but with a 29 percent undecided response, the poll was less reliable than your daily horoscope. Ten percent undecided would be large. Twenty percent would be huge. But almost 30 percent is a joke, a bad joke that invalidates the alleged results.
In a desperate move to keep some credibility, Probe Research invented a momentum index, claiming to measure which candidates are moving up in public opinion and which down.
When was the last time you heard of something like this? Hint: NEVER.
But they had to deliver something to the private group, Manitoba Forward, and the United Fire Fighters of Winnipeg who fronted most of the cost of this exercise and who should complain to the BBB of polling -- if there is such a thing.
Do you want to know how useless this mayoral poll is? Just compare it to a similar poll conducted before the 2010 civic election:
Judy, Sam neck and neck in poll
Mayor a little behind challenger in one sample, a little ahead in other
By: Bartley Kives Posted: 10/2/2010
THE METHODOLOGY Probe Research conducted telephone interviews with 600 randomly selected Winnipeg adults between Sept. 16 and 30. One can say with 95 per cent certainty that the results, taken from a subsample of 439 people who intend to vote, have a margin of error of 4.7 per cent. The sample was statistically weighted to conform to Winnipeg’s demographics.
THE LARGER SAMPLE If you look at the entire 600-person sample — including people who do not intend to vote — Katz is ahead of Wasylycia-Leis by a three-point margin, 50 per cent to 47. The results remain a statistical dead heat, within the margin of error of four per cent. Probe Research president Scott MacKay considers this sample a less reliable snapshot of actual voting intentions.
THE race for mayor of Winnipeg is now a statistical dead heat. Incumbent Sam Katz has lost the polling lead he held all year and is now even with challenger Judy Wasylycia-Leis.
According to a Probe Research poll conducted in late September exclusively for the Free Press, Wasylycia-Leis is the preferred choice of 50 per cent of Winnipeg voters. Katz sits three points behind with 47 per cent support and Brad Gross and Rav Gill garnered the remaining three per cent. The margin of error for these results is 4.7 per cent.
The overall results point to Winnipeg's closest mayoral race since 1998, when Glen Murray squeaked by Peter Kauffman by 10,569 votes.
Well.....as it turned out that poll was indicative of absolutely nothing. Sam Katz trounced JustJudy by double digits. He collected just shy of 55 percent of the vote, she not quite 43 percent. She lost by almost 25,500 votes, hardly the squeaker predicted by Probe.
The only benefit of the 2014 Probe mayoral poll was as fodder for the news media's narrative of the election campaign. You see, modern journalists don't report the news anymore---they create the news, as CJOB's Richard Cloutier did when he disrupted a Gord Steeves news conference because Steeves wouldn't play the foil to CJOB's live broadcast.
The current narrative is 'left-wing candidate Judy Wasylycia-Leis (she's abandoned the folksy, aw shucks Just Judy persona) vs. a gang of right-wingers.'
It's a retread. That was last year's media narrative also.
When that wasn't enough, Free Press columnist Dan Lett resorted to outright fantasy and fabrication to smear Steeves:
"There was a time earlier in the campaign when it looked as if Steeves would drop out. He had a horrid start involving botched events and the resignation of some key campaign organizers.
Since then, however, he has delivered a barrage of campaign announcements, some intriguing, some poorly conceived. There was also the matter of an angry Facebook rant by his wife against homeless aboriginals."
That never happened.
She never mentioned homeless aboriginals. This is a complete and total invention by Dan Lett. Years past, Lett tried to show how clever he was by quoting his wife, who allegedly made the pithy observation that you're entitled to your own opinion, but not your own facts.
That, obviously, was before Lett found how much fun he had making up "facts".
This egregious attempt to malign Gord Steeves is just part of the pattern of Lett's writings. We addressed his confessed bias in his columns two years ago:
At a real newspaper, he would be disciplined for printing lies. But the attacks on Steeves are part of the Winnipeg Free Press narrative on the election, and will be rewarded, instead.
Speaking of mainstream media incompetence, here's some of the ever-popular:
"New crowd-funded poll gives Wasylycia-Leis 16-point lead in mayoral race
By: Mary Agnes Welch Winnipeg Free Press Posted: 08/28/2014 3:30 PM |
The poll was conducted over the weekend using a custom online panel and has a margin or error of +/- 4 per cent."
"Wasylycia-Leis tops 2nd mayoral poll
By Kristin Annable, Winnipeg Sun
First posted: Thursday, August 28, 2014 02:43 PM CDT |
The data for the country’s first crowdfunded election survey was collected from 603 Winnipeggers between Aug. 22 and Aug. 25, and is accurate to +/-4% at a 95% certainty level."
"Bowman Not Going To Leave Mayoral Race
Kim Lawson CJOBAugust 29, 2014 02:45 pm
The survey was conducted with 603 members of Probe Research’s proprietary online panel and has a margin of error of +/- 4 per cent, 19 times out of 20."
This survey was conducted with 603 members of Probe Research’s proprietary online panel.
Because this is a non-probability sample, no statistical margin of error can be ascribed.