Friday, December 15, 2006

How do they get these results?

If you believe everything you read, star signs correlate to traffic statistics... Nifty.

However, What I don't understand is how they get the statistics...


So... we're assuming that we believe this at this point, that's interesting enough reading... they give the obligatory reasons why it is all so - that's fine. These figures are supposed to come from a combination of stats about accidents and tickets and then they go on to give the stats on tickets and accidents separately...

What I don't get, is how they get the main results they are spouting, to tout their book...

Lets take a look:

Their List
Worst Accidents Tickets Total of Rank
Libra 1 5 6
Aquarius 5 3 8
Aries 4 2 6
Pisces 7 1 8
Scorpio 2 7 9
Taurus 8 6 14
Sagittarius 6 11 17
Capricorn 3 4 7
Virgo 9 10 19
Cancer 11 9 20
Gemini 10 12 22
Leo 12 8 20

What I have done, is given a value for each sign as to how "dangerous" it is... 1 equalling the highest danger and 12 equalling the lowest danger. I then added them together, and in theory, the lower the number, the more dangerous the driver is.

Given this basic statistical modelling of the data they have presented to the world, I see that the rank should look a little more like:

Worst Total of Rank
Libra 6
Aries 6
Capricorn 7
Aquarius 8
Pisces 8
Scorpio 9
Taurus 14
Sagittarius 17
Virgo 19
Cancer 20
Leo 20
Gemini 22

That moves the list around quite a bit. Unless there is some serious weighting to some of the figures which there could be... (there might be more accidents caused by Aquarians than Capricorns, so therefore they move up the list as "dangerous drivers")... but they don't mention this on the page, in fact they don't mention any figures on that page other than 100,000 drivers and the rankings. (I'm a Capricorn by the way, so my star sign looses by me bringing these figures to the world's attention ;))

Personally, I think the proportion of people to star sign would be an interesting stat to look at with or without the claim of month of birth to accident/ticket ratio... I doubt that people in a non-geographically diverse area are statistically born all year round (if you took a world wide sample, you're more likely to get closer to an "all round" birth rate)... there is a strong likelihood that there are three or four star signs that are more prevalent in the study - and therefore, surely by pure virtue of being more populace, statistically more likely to have accidents/get tickets?

I dunno, I just juggle numbers to get the answer my boss wants half the time ;)

I also heard that 77% of all statistics are made up on the spot....


Maybe I should troll the website for some figures on gender to accident/ticket ratios... they prolly have some somewhere, if they want to sell a book about it...

Oh, and one last thing, what the hell is the "My Name is Earl" reference got to do with the damn story, or book for that matter... he got hit by a car, and then tries to modify his behaviour - oooo his star sign must be a Libra, or Aquarius, or..... If the driver that had hit Earl had read the book, such "hilarity" like "My Name is Earl" might never had existed...

...on second thoughts, this book should be recommended reading for all fictional/Hollywood drivers!!

*rreooooowwwr* !

No comments: