This is a fairly quick post (from a train! Ooh, technomology) on the final talk from Day 2 of the BSA SciComms conference 2010 as I know a few people had to miss it and includes things I wanted to write about before.
Matt Parker is a prominent contributor to the outreach movement; a “stand-up mathematician” currently based at QMUL (my uni!). He tries (and, I reckon, succeeds) to convey why he finds maths so fascinating and enjoyable like the behaviour, patterns and quirks of numbers – in contrast to the kind of stuff most of us are forced to do at school and consequently don’t tend to find exciting.
He started with a tale reminiscent of the work that’s been done on debunking ideas like the Bible Code – pointing out that, given a large enough data set, you can find just about any pattern you want, anywhere (then crazies like to ascribe some huge significance to it, and people believe it).
This case concerns our favourite ‘news’ rag, the Daily Mail, reporting that ‘prehistoric monuments form amazing isosceles triangles!’ in the South West of England. The article tries to portray the layout of these sites as deliberate and ‘sophisticated’ and I’ve just seen this amusing line: “It is known that many, if not all, early churches, abbeys and cathedrals were constructed on ancient sites and this diagram illustrates that point.” – well, yeah. Ancient things were build on sites that are now considered old. Durr.
Anyway, Matt set out to show how ridiculous these claims are by means of a similar exercise using… Woolworths stores. His results actually ended up being more accurate than the Mail’s within-100-metres, e.g. the Conway store being just 40ft from the exact line! He found over 85 million triangles could be made, compared to the prehistoric sites’ potential >100 million, further diminishing the impact of the Mail’s “revelation”.
Indeed probably the best thing to come out is the new term for such statistical manipulation; “Pick ‘n’ Mix data analysis”!!
There’s nothing in it
Matt went on to present some material he’d composed for the 10:23 Campaign.
He’d looked at the homeopathic remedies in Boots. They recommend arnica for bruises etc. and sell 30C* homeopathic arnica, which costsapproximately 6p/pill – sounds like a bargain, right?
How many pills would we need to take in order to get just one molecule of arneca??
7 million billion billion billion – in terms of cost that’d be greater than the GDP of the UK, or even of the world… since civilisation began.
You cannot buy 1 arneca molecule from Boots.
*30C means 1 drop into a solvent. Then dilute 1 drop of that into the next volume. Do it 30 times. That’s approximately a 1:1060 dilution. Sugar pills are then given a drop of the final solution and bottled up for sale.
To look at it another way, how would we take all these pills? Should we hypothetically acquire them.
– Fill Wembley stadium with people and pints of pills.
– Get each person to take 2 pints per hour.
– Get yourself 1 million Wembley stadia (holding 50,000 people each).
When would we need to get everyone to start taking the pills, to make sure just one person got one molecule?
4.5 billion years ago (i.e. when the planet was formed).
Oh, and you’ll need 737,000,000 planet Earths.
Of course, homeopaths argue that it’s not about how much of the substance you have, but that shaking (or ‘succussing‘, which I see spelled loads of different ways, amusingly) the original solution in a special way – though there is no standard, agreed-upon method – ‘imprints’ the ‘memory’ of the substance in the solvent.
As Tim Minchin points out, funny how it might then choose to remember a bit of onion juice over all the poo it’s had in it. This ‘theory’ is thoroughly deconstructed elsewhere and should I find myself with far too much free time (unlikely in the next 2 years) I might try to summarise the (lack of) evidence presented in the homeopathy Evidence Check report (produced for the government this year to assess the justification for funding homeopathy on the NHS) of which I have a copy. But again, others have done so.
Le Canard Noir makes an amusing illustration of the kind of point made above here.
6.022×1023 is quite special to us sciencey types. It’s Avogadro’s number – we use it to work out how many atoms of a substance are in any given amount of it. How big is the number? Consider the water on our planet.
There are 1,347,000,000 km3 H2O on planet Earth, which is equal to 2,689,000,000,000,000,000,000 (approx. 3×1021) pints.
1 pint contains 2×1025 H2O molecules; so, there are more water molecules in a pint than there are pints of water on the planet!
That’s what we call a mind-f*ck. Matt advises people to think about that next time they consider relieving themselves in the ocean.