26 November 2009

The Terra Nova by-election result

Just to drive home the point on just how inaccurate CRA polls really are here are some comparisons using the Terra Nova by-election.

1.
Actual Turnout
2.
Actual, Percent, Turnout
3.
Actual,
Percent, All
4.
CRA Nov Turnout
5.
CRA,
Corrected All
6.
CRA, Corrected, Turnout
LIB
1663
38
20
16/697
12/991
12/522
NDP
297
07
04
07/305
05/413
05/217
PC
2398
55
29
77/3356
58.5/4833
58.5/2549
DNV
3904
-
47

24/1982
Total Eligible
8262
100
100

The first column shows the actual turnout vote along with the number that didn’t show up.

The second column shows the turnout divvied up among the three parties, with the people who didn’t vote discounted. That’s basically the way CRA does its numbers as a percentage of “decideds”. Everything that didn’t make a vote choice is lumped in a big pile and tossed to one side.

The third column shows the actual percentages for each category, including those who didn’t vote.

The fourth column is where it gets funny. If we take the actual turnout of 4358 eligible voters and divide them according to the shares that CRA projected, this is the percent and the number you get. As you can see, these numbers aren’t even close to the actual result in any way shape or form.

That fifth column uses the corrected CRA results and plots out the projected returns in comparison to all eligible voters. Frankly, this is the one that CRA results should most closely model.

Well, it should but it doesn’t. You can see the undecided, will not vote is way off. The Tory vote is inflated, the Liberal vote is pushed down and the NDP number is pretty close to right, relatively speaking.

Then in that last column and just for even more fun, there’s a comparison of the CRA figures – corrected as a percentage of all – and then reapportioned as the turnout vote.

Well that one is a wee bit closer to what actually happened but for some bizarre reason CRA just can’t seem to get that Liberal vote even close to right.

Now to anyone who really wants to think hard about the results consider column three as your food for sustain you over the long cold winter. A higher percentage of people stayed home in Terra Nova than turned out to exercise their democratic right.

Almost half the voters in Terra Nova district opted to sit out rather than get involved.

Someone needs to probe the reasons why but your humble e-scribbler would venture there is a something that would motivate those people to get out and vote. There is something that none of the parties has currently figured out.

But if someone could energise those people, you could actually defeat any one of the existing parties – including Hisself – and do so handily. What would motivate those voters might even peel voters away from one or more of the other piles.

Now there’s an interesting idea to give some people the cold sweats at night.

-srbp-