What Election Tell Us About Political Party Brands

[maplesyrup]

What the Election Tells Us About the Strength of Political Party Brands?

I read what appeared on the surface to be an interesting analysis of the recent election results, but then something started to gnaw away at me, and now I am beginning to have some reservations.

I seem to be having some difficulty pinpointing my concerns.

Suggestion: First of all try the math - check the figures in the chart. Do they all jive for you?

[maplesyrup]

First of all, do the 2nd Place Finishers add up to the correct total?

[kimmy]

That's one problem. As well, as far as I can figure out their Competitiveness % is calcluated by dividing the "<3k" column by the "2nd place finishes" column ... which should give the NDP a Competitiveness % of not 20%, but a whopping 37%, far better than the other national parties.

And I think the value of any calculation that concludes the NDP is Canada's most competitive political party is dubious to say the least.

-kimmy

[maplesyrup]

Well, initially it was Le BQ at 33%.

But you have come with the same figures I did, and it kind of skewers a lot of his analytical comments.

So are you saying that his analysis is not valid, or just that is not valid because it shows promising results for the New Democrats?

[kimmy]

Well, initially it was Le BQ at 33%.

But you have come with the same figures I did, and it kind of skewers a lot of his analytical comments.

So are you saying that his analysis is not valid, or just that is not valid because it shows promising results for the New Democrats? 

I just glanced over the analysis-- it seemed to me that his analysis had little to do with the Competitiveness statistic he invented.

I'm saying that I think he invented a statistic that is pretty much meaningless, and as an example I cite the strong rating it gives the NDP. I think that even Jack Layton himself would concede that the NDP was not the most competitive party in the last election, don't you?

I think the whole premise that dividing the # of close 2nd place finishes by the total # of 2nd place finishes is very hoaky. Suppose that some party, say the Rhinoceros Party finished dead last in every riding except for 1 riding, where they had a close 2nd place finish. So, they had 1 close second place finish out of 1 total second place finish... that would give... a 100% competitiveness rating! Wow! Even though they finished dead last in 307 out of 308 ridings!

I think this example highlights where this guy's thinking falls flat: it doesn't take into account the total number of ridings. It also doesn't take into account the number of victories (which are probably a better measure of a party's strength than close 2nd place finishes, I would think. )

What if we try a difference calculation:

-Victory = 3 points

-Close 2nd place finish = 2 points

-Non-close 2nd place finish = 1 point

-3rd place finishes or worse = no points.

Total # of points available = 3*308 = 924 (except for the BQ, which only ran in 75 ridings, so 3*75=225 points)

Using the numbers from his chart (which we already know don't add up, but I'm not going to go find the real numbers...)

Liberal:

3*135 = 405

2*38 = 76

1*109 = 109 (ie, 147 2nd place finishes minus 38 that were close = 109 that weren't close.)

total: 590/924 = 63.9%

Conservative

3*99 = 297

2*16 = 32

1*79 = 79

total: 408/924 = 44.2%

NDP

3*19 = 57

2*19 = 38

1*32 = 32

total: 127/924 = 13.7%

Bloc

3*54 = 162

2*21 = 42

1*0 = 0

total: 204/225 = 90.7% in Quebec.

I don't know if these calculations are very meaningful either, but I do think my idea makes a lot more sense than his!

My calculation includes victories, which are kind of important. My calculation also includes the total # of attempts, his just counts just the # of 2nd place finishes that weren't close.

-kimmy

[maplesyrup]

Kimmy.....good work. It looks like the author has about as much credibility in my eyes, as some of the polling companies such as Ipsos-Reid.

Those chart miscalculations means that substantial portions of his written analysis leave a lot to be desired.

[maplesyrup]

Now what can we assume about the author?

My impression is that he is a Liberal supporter as his written analysis shows a bias towards the Liberals.

Was he trying to be scientific, and just made honest mistakes like Ipsos-Reid supposedly did in their election polling for example, when they showed the New Democrats at 14% in BC?

Or, did the author more likely, in my opinion, already have a conclusion in mind, when he began his research, and was just looking for stats to support his biased propaganda.