sqbr: pretty purple pi (Default)
Sean ([personal profile] sqbr) wrote2006-08-29 05:53 pm
Entry tags:

Bad Science: have they not heard of the slayer??

*puts on science communicator hat*

Have any of you read this ABC article, about how physicists have "debunked" the paranormal? The logic given struck me as incredibly flawed but being the good little ex-Phd student I am I looked up the orginal paper (which despite being quite easy to read was not linked from the ABC article, luckily there aren't many Efthimious on arXiv) and while some of it was just badly paraphrased the vampire example is bad maths pure and simple.


Their argument is as follows: every time a vampire feeds it kills one human and creates one vampire. Thus every time the vampire population as a whole feed their numbers will double, quickly leading to an absurd situation with a hundred gazillion vampires and no people.

They haven't disproven vampires, they've "disproven" predators. Since yes, like any predator, vampires reduce the prey population and "reproduce". But according to every vampire legend I'm aware of they can also die. So if the human population got too low, they'd start competing for food and killing each other, not to mention the ever prsent danger of angry mobs of peasants etc. This is your typical predator-prey situation as familiar to any student of basic differential equations (which is what I am, at least about predator prey models).

Admittedly, if the death rate of either species is lower than their "birth" rate then the only stable population size is fractional, ie zero. So to work the vampires can't reproduce as often as the paper assumes (once a month), which does debunk the vampire myths where vampires hunt all the time and any of their victims which aren't staked become vampires too, unless there's a whole barrage of Van Helsings/Buffys/those-cute-guys-from-Supernaturals chasing them over the countryside. But afaict most, or at least many myths only have a fairly small proportion of victims turning into vampires, so as long as that proportion is less than the number of vampires who die due to infighting/peasants/cute guys etc then we're in gravy.


This may seem like a lot of effort to go to for a silly theory, but
(a) This kind of bad science plays into the hands of the pro-supernatural, since they can justifiably point to it as proof that science is "out to get them"
(b) This guy (or girl) is trying to encourage university science lecturers to use this kind of discussion to teach physics (this paper looks interesting too), which I think is a great idea in principle but not if the science is wrong.
(c) I'm a big fat nerd. C'mon, maths + fantasy, how could I resist.

Also, today I got paid to play with hexaflexagons. Admittedly, I would rather have had the day off properly (I was home sick from my shift, but this activity stuff needs to be done asap) but it was still fun if draining. Clearly it got my brain ticking over, though :)

[identity profile] penchaft.livejournal.com 2006-08-29 10:59 am (UTC)(link)
every time a vampire feeds it kills one human and creates one vampire.

NO, NO THEY DON'T. Where the hell did they get their "reserch" from?

[identity profile] greteldragon.livejournal.com 2006-08-29 11:02 am (UTC)(link)
Was that typo delibrate? *curious* It suits so well.

[identity profile] greteldragon.livejournal.com 2006-08-29 11:07 am (UTC)(link)
hehe. We need stickers saying "lern2reserch" to be stuck onto any bad bad papers. Or completely unthought out world building things. :P
alias_sqbr: the symbol pi on a pretty background (Default)

[personal profile] alias_sqbr 2006-08-29 11:22 am (UTC)(link)
Oh man, I wish I was on the panel for a journal now just so I could hand those out :) Hey, my old supervisor is on the Fields Medal committee now....Hmmmm
alias_sqbr: the symbol pi on a pretty background (Default)

[personal profile] alias_sqbr 2006-08-29 11:03 am (UTC)(link)
"Blade" :)

Although that's certainly got a fairly high vampire death rate from it's title character alone.

[identity profile] penchaft.livejournal.com 2006-08-29 11:07 am (UTC)(link)
So they completely didn't mention the born species of vampire? Or the fact that the majority of vampire fiction IS DIFFERENT FROM BLADE AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAARGH.

[identity profile] pooxs.livejournal.com 2006-08-29 12:24 pm (UTC)(link)
afaik blade is the _only_ vampire thing I've read/watched/whatever where it just takes one bite (drink?) to turn a human into a vampire. everything else they had to do something extra to turn

[identity profile] volandum.livejournal.com 2006-08-29 11:10 am (UTC)(link)
Are papers usually this patronising?
alias_sqbr: the symbol pi on a pretty background (Default)

[personal profile] alias_sqbr 2006-08-29 11:23 am (UTC)(link)
No. This isn't a real scientific paper, it's a patchy attempt at science communication. arXiv is a paper repository which will take anything, so there's some pretty random stuff on there.

To give a totally random example, consider my one published paper :) (my supervisors put it there so people could read it during the peer review process, it's in a Real Journal now) See, not patronising, just unreadably dense.

[identity profile] edible-hat.livejournal.com 2006-08-29 11:32 am (UTC)(link)
I wonder if they'll take my proof that nothing that happens is agains the will of God...
alias_sqbr: the symbol pi on a pretty background (Default)

[personal profile] alias_sqbr 2006-08-29 11:55 am (UTC)(link)
Possibly, if you format it right :)

[identity profile] volandum.livejournal.com 2006-08-29 02:21 pm (UTC)(link)
I see, thank you.

I'll learn to read it in a decade or so.
alias_sqbr: the symbol pi on a pretty background (Default)

[personal profile] alias_sqbr 2006-08-31 01:07 pm (UTC)(link)
You could instead just read the poster. It's pretty much just the same.

[identity profile] volandum.livejournal.com 2006-08-31 02:34 pm (UTC)(link)
Yep.

[identity profile] edible-hat.livejournal.com 2006-08-29 11:31 am (UTC)(link)
Somebody once said that to survive in the memescape, a pseudoscientiffic belief has to be bad enough to keep real scientists away but good enough to attract the attention of scientists on the worse side of mediocre to give it a little legitimacy.

[identity profile] edible-hat.livejournal.com 2006-08-29 11:31 am (UTC)(link)
(like, for example, the concept of a memescape)

have not read the article

[identity profile] anxiolytic.livejournal.com 2006-08-29 07:08 pm (UTC)(link)
I read your precis and from that I think you have a bunch of high school/first or second year uni students writing this paper.

And you non-life scientists have non-peer reviewed web sites that anyone can post to? AND you post your article whilst it is still in peer review? Screwy
alias_sqbr: the symbol pi on a pretty background (Default)

Re: have not read the article

[personal profile] alias_sqbr 2006-08-30 12:03 am (UTC)(link)
Disturbingly, no, you have a university lecturer writing the paper. But as I said, it's not a real physics paper at all, it's aimed at the general public, so they're trying for easy to understand rather than accurate. Not that physics lecturers would know much about modelling predators anyway.

And yes, yes we do. Given how much maths seems to based off other people's unpublished research they told you about in the pub(*), I think reading submitted-but-not-reviewed papers is a step up :)

(*)For example, one of my examiners complained that I hadn't mentioned one of the two seminal papers on a particular area in my lit review. When I asked him for a reference (since he was, coincidentally, the author) he thought about it for a while and remembered that while it's been well known for a decade and everyone refers to it, he still hadn't gotten around to publishing it.

[identity profile] kadeton.livejournal.com 2006-08-30 01:44 am (UTC)(link)
Wow, that whole article is an exercise in dumb.
alias_sqbr: the symbol pi on a pretty background (Default)

[personal profile] alias_sqbr 2006-08-30 02:11 am (UTC)(link)
I do wonder how slow a news day it must have been for the ABC to present as new science.