Polls Are For Trolls and Fools ~ Just Ignore Them!

Or Better yet, just stop taking them!

The Nik Nanos interview on CTV's Power Play  on Sunday Feb 20th (see the Nanos interview at the 1:38 minute mark) should have been billed as a Conservative Infomercial, or at the very least, a Con Game. And since it was not, I think some might have taken it seriously, the way Canadians do. Not that they tend to take it so seriously that they look at the disclaimers at the bottom of the stats, or even more importantly, the light that was made to shine on it just last week in print, and on "AT ISSUE" by Allan Greg. He is always described by Mansbridge as a long time Conservative Pollster, so it was refreshing to hear his refresher on the subject.


Allan Greg speaks to "The House" on CBC radio via CBC.ca
At any rate, The Cons are not worried, nor will they be until the Pollsters stop hyping them. Nik Nanos like all pollsters who insinuate that polls can be trusted, simply can ^NOT be trusted. And an increasing number of those of us who like to do a bit of math when we have doubts, see that the numbers are more political than scientific and our oddly widespread faith in them makes them much more able to affect outcomes than predict them. Polls don't just let us know who is going to win, they help determine it.

I Call in to "The Week" ~ A Sunday NOON ET show on CPAC with a few Questions and a few comments including one about Greg's idea that these Polls are now unreliable and as always an exercise in political interference. I ask them why in light of this, they still insist on pushing poll results.

Watch it Here ~ I'm on at the 23 minute mark,
but the whole show is well worth a look.

Pollsters call a few thousand or so (that's one in every 70,000 or so voters) mostly folks on land lines, only 15 to 20% of whom will talk to them at all and blithely come up with numbers which are (19 times out of twenty) somewhere ( according to them) within +/-3.4% (That's a possible 6.8 points off of, (-) or added to (+) the total spread) and rely on our being too damned lazy to do the math.

We could simply subtract that 6.8% and know that the current (supposed) 14% spread might only be 7.2% but they never bother to do that little bit of the real math on our behalf.

And don't forget that one in Twenty polls where even that huge inaccuracy is multiplied by built in error. (nineteen times out of twenty, remember)

Now Keep in mind that these polls are pretty much weekly and given the 52 weeks that is a year here in Canada, we are fed this twentieth poll two or three times a year. They don't bother to make that clear. Like I said they leave us to do that math. Why?


Is it because they know that it would show them to be irrelevant and leave them much less useful to the political parties that pay for them? I am afraid  that that is a sad and distinct possibility.

Food for thought.
A Review BY ALLAN GREGG of"
"Harperland" ~ The Politics of Control by Lawrence Martin published by Viking Canada