OT

Oct 31, 2015 59 Replies
Saved reading
Jump to replies (59)

Until this summer I would have said that all air crashes especially those of airliners take place within a few days of a large earthquake. Yesterday I forecast another such earthquake for Sunday.



I would have confined such posts to a less OT forum in times past but why should they have all the fun?



formatting link


So presumably you have a nice timeline showing the distribution of aircrashes and earthquakes?

Depending on the definition of "large", enough earthquakes happen every day *somewhere* to make a correlation with something approaching certainty. But correlation and causation are not the same thing.

The one I liked best was the price of corn and the number of drawn matches in county cricket games in the 19th century...

Today, we would be producing learned studies on how lack of starch in the diet adversely affected the ability to play aggressive cricket...

even better is the rising crime rate since 1945 and the number of fridges in the home.

If you set out to produce a link between two things, it's usually possible.

Processed meats and cancer springs to mind. ;-)

You can get causation between things that are totally rubbish and the real world. Take astrology and the stock market.. if some astrologer predicts a market fall there are enough idiots trading to actually start selling and make the fall happen.

Hangon! that also equates to Climate change! Its down to FRIDGES

I think that last sentence is one that WL has difficulty with. ;-)

Tim

Well, CFCs were certainly said to damage the ozone layer, and plenty of the older ones have been banned from use in fridges etc. So perhaps some truth in it.

I think it was the hairspray usage in the 80s!

formatting link

Why do you presume that? It would involve matching two sets of statistics. And all it would do is convince me. Nobody ever got convinced by statistics that wasn't easily gulled or didn't already believe it.

If you want data there are several websites dedicated to them and there are scads of videos about them. You may even find a list at Winkiepaedia.

I would have thought a large earthquake occurring in the same spell as an air crash was by definition a correlation.

What sort of difficulties do you prefer with your statistics?

Well you seemed to believe it until this summer. Have you changed your mind then?

Tim

Trouble is, even some judges seem unable to manage statistics. As witness all these people convicted for murder because they have three cot deaths.

F*ck all have and those that have been have other evidence that they are murderers.

Wrong. A conviction was famously overturned when it was realised that the judge had just multiplied the normal incidence of cot deaths to give a fabulously unlikely probability that it could have occurred by chance.

What he failed to comprehend was that the chance of another cot death drops dramatically with each cot death.

Tim

We'll see...

That's just one, not 'all those people'

You presumably meant increases dramatically. No it does not.

F*ck all have 3 or more cot deaths.

On 31 Oct 2015, charles grunted:

Or my lecturer at uni, who demonstrated a close correlation between the annual birthrate in France and the years he visited the country on holiday.

On 31 Oct 2015, Tim+ grunted:

Thing is, those convictions were overturned because of that flawed expert evidence, meaning that the defendants were pronounced 'not guilty' under English law. If that evidence hadn't been used then one or more of them could well have been found guilty regardless.

For what it's worth - not a lot, I agree - I have a little inside knowledge of one of the cases which wasn't aired in court or made public (and which I am certainly not going to expand upon here). All the newspaper headlines about the hell that the 'poor innocent woman' had gone through due to her false conviction really stuck in my craw.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required