Infinite Recursion of Increasing Complexity

Dr Who science abuse

Sun, 04 Jun 2006 03:33:05 GMT

I watched Dr Who on Saturday and was horrified beyond belief at the amount of murdering of science that went on. This is really awful science murder, worse than anything I've ever seen from any other science fiction.

The basic premise was a planet orbiting a black hole. Throughout the first half of the program or so it's impressed on people how this is completely impossible - the planet ought to be pulled in.

The writer (I was half expecting it to be Russell T. Davies, but it turns out to be Matt Jones) is falsely assuming that the gravitational field strength of a black hole is infinite, and that any object within a defined distance from it will be sucked in regardless of its motion - this is another way of saying that any orbits about the black hole are unstable. Clearly if this were true then all objects in the universe would be being pulled into the nearest black hole; the existence of galaxies orbiting supermassive black holes acts as a direct contradiction to this.

Importantly, the gravitational field strength g of a black hole is not infinite, but merely finite but so great that the escape velocity v = √2gr is greater than the speed of light c. The maths works out that the stability of an orbit around a black hole depends on the Schwarzschild radius r_s (the distance from the singularity to the event horizon) of the black hole; there is a theoretically stable orbit at 2r_s at which the slightest change in momentum will result in the orbit degenerating and the object falling in, and an infinite number of stable orbits outside 3r_s.

The result of this lack of scientific thought means that Jones has to resort to technobabble to explain how humanity got to the planet in question - he mentions a power source (the critical plot point of the episode, by the way) producing a "gravity funnel". The skeptic in me choked on his own bile at this point, since he knows full well that they just turned up and landed on this perfectly possible planet! (Alliteration is good.)

Of course, ignoring trivial things like scientific reality and causality is order of the day here - the Doctor is kept on the planet by the loss of the TARDIS, which collapsed into the mineshaft (they're drilling for the power source producing the technobabble) during an earthquake, conveniently ignoring the fact that the Doctor and Rose both have keys to the TARDIS which instantly warps them to wherever it is. At least it wasn't on Earth this time around...

Openmindedness

Fri, 02 Jun 2006 15:41:47 GMT

Mark Chu-Carroll of Good Math, Bad Math has been grappling with Gary Osborn, a new-age crank who claims to have found mythical significance in the specific angle 23.5 degrees from the horizontal or the vertical, and demonstrates it by showing its presence in various paintings ranging from the Ancient Egyptian to the Baroque period. He links this to the axial tilt of the Earth, and somehow goes from there to imperfection to death.

The ensuing debate (Gary, like all other bullshit-peddling mystics, remains firmly on the defensive when posed with simple questions regarding, among other things, his experimental protocol in determining these angles) spawned a couple of entries on the subject by Bronze Dog and Big Dumb Chimp, which so incensed Gary that he felt obliged to continue his tirade against scientific rigour in their comments section (and, like cyberz74, I'm inviting a bucketload of drama onto myself by blogging about it at all).

The crux of all this was when I decided to get off my arse and do some real experimental work. I grabbed my ruler and protractor, blew the dust off (I haven't actually used a ruler or protractor in a good few years now) and grabbed the nearest thing to hand - the pizza box my lunch today came in - and within five minutes using Gary's own technique found two 23.5-degree angles.

I guess Domino's Pizza are in on the secret.

Why I hate e-mail

Tue, 23 May 2006 11:05:06 GMT

I've recently considered taking a stand in deleting my e-mail address due to the recent influxes of spam which is put together in such a way that I can't block it even with Thunderbird's otherwise excellent Bayesian spam filter (it seems to be a notch above the old "v1agr4 n0w!!!!!!!!1 banana" style spam which Bayes' formula caught easily enough). I've reconsidered my earlier decision, and am using a ruthless spam filter called SpamAssassin which should filter 99% of spam (it uses the Realtime Blocking List, famous for blocking all undesirables although also blocking a fair amount of chaff; I recieve little enough e-mail that this won't be a problem).

Spam mail has become enough of a problem to warrant my taking action, and I endorse the ideas behind the ideas of Dan Bernstein and Jonathan de Boyne Pollard in their IM2000 specification. (Mr de Boyne Pollard may be a gigantic holier-than-thou asshole but he has the right idea here.)

The problem behind IM2000 is penetrating the market (what I often refer to as "the Internet at large" to contrast with "the Internet proper": the massive body of non-technerds who use the Internet for porn, talking to their son at Stanford and reading about Nintendo's latest inane meanderings) - people are too used to regular SMTP mail to want to use it. Oddly, JdeBP endorses a policy of forcing people into IM2000 by clogging up regular SMTP mail with enough spam to make even Joe from Texas complain. This would seem to be counterproductive and downright illegal, but hey, JdeBP doesn't have an SMTP e-mail address, so he doesn't care.

The best way to penetrate the market in my opinion is to rope in Google to come up with a decent client for it, then use Google's might to advertise this new form of e-mail emphasising its capacity for eliminating spam entirely. One would have to be careful Google doesn't try and pull a Bill and get all bitchy and proprietary; e-mail has to remain free for anybody's use.

IQ tests

Sun, 21 May 2006 19:33:33 GMT

sorakazi took an IQ test unaware that he'd been duped by a false test intended to demonstrate the falsity of IQ tests in general. I'll repeat in modified form what I sent in an e-mail to the person behind this LJ experiment.

The argument used on the website is flawed because it assumes the kind of trite "IQ test" you find on the Internet is actually a balanced assessment of intelligence. The truth is that an IQ test is designed such that the designer can easily skew the data to make people feel more or less special. The reason is statistical, and works as follows.

If one assesses the intelligence level of a large number of people in arbitrary units then the results will form what is called a bell curve - most people will fall into a small region in the centre with progressively fewer people at the two extremes (there are, contrary to popular belief, very few extremely stupid or extremely clever people compared to the entirety of the sample of humanity taken). There are a variety of ways of assessing these data, but relevant here is the method which uses the average value and what is called the standard deviation, or the mean difference between a value and the mean.

An IQ test is simply an agreed method of stating the arbitrary units; it takes the most common level of intelligence to be an IQ of 100. The variation in IQ tests is in the next point - the maker of the IQ tests picks the value he wants for the standard deviation. This means that every IQ test is different, and that IQ is a meaningless number unless you're also told what value was chosen for the standard deviation.

This is best illustrated with an example; let's say I take an IQ test and get a value of 175. I need to know the standard deviation - if it was chosen to be 50 then I'm more than one standard deviation from the mean, making me rather above average, but if it's set at 150 then I'm not really that much above average. (Note that either way one should take an IQ value with large quantities of salt; it's an abstract assessment which cannot possibly cover all aspects of intelligence, and everyone has strengths in something whereas IQ aims to assess more fundamental skills like abstract thinking and ability to visualise.)

That said, the person behind the LJ experiment manages to be right for the wrong reasons on two counts.

His point about the IQ tests found online being next to useless is right, but not for the reasons given on the page; I suspect he's oversimplifying for an overly stupid readership, whereas my personal credo is to give the facts plainly and be it on the head of whoever does not understand, since they teach this in every school in Britain to everyone who cares to listen and ask the right questions.

His point about organisations like MENSA being misguiding is right, but since it's based on the same flawed argument it falls down on the same merits.