Thursday, April 17, 2008 ... //

Nico Marquardt & asteroid: not really

This 13-year-old German kid, Nico Marquardt, made his own calculations of the probability that the asteroid 99942 Apophis - the same asteroid that we have already discussed under the name 2004 MN4 in 2004 - will hit the Earth. What a smart boy. The courage, enthusiasm, and talent of this young scientist must be appreciated.

The original collision was scheduled for Friday, April 13th (!), 2029. However, the probability of a judgement day on this particular black Friday has plummeted and now we talk about a much more uncertain possible encounter later in 2036 - one influenced by the asteroid's hitting a 600-meter "keyhole" in 2029. Instead of NASA's estimate 1:45,000, Nico obtained 1:450.

See the current Impact risk table: Apophis is back to Torino scale 0
Yesterday, a rumor began to propagate that NASA admitted that they were wrong and the boy was right. Recall what things you have to know, what data & software you need to have and master, and what errors you have to avoid in order to do such calculations properly and try to answer the following question: What is your estimate of the probability that such an incident in which the top world's space organization is humiliated could occur? It is insanely unlikely. My estimate is 10^{-12} per the whole childhood of a schoolkid below 14 years of age which means that such a story could actually occur once per 10,000 years.

Nevertheless, the media immediately started to print this sensational story as a fact. The journalists must have thought that the insanely tiny probability was not only above 50% but it was actually close enough to 100% so that they could present the story as a fact. They even didn't seem to have the idea to ask someone sane (NASA, for example) and check whether the story is sane or insane. "A boy debunks & outsmarts NASA and we are all doomed," we could read from:
AFP
Economic Times, India
Defamer
SlashDot
ABC News, Australia
The Sun, U.K.
and dozens of others, including most of German outlets. A Czech server compared Nico to Adam Bernau, a 20th century genius from a Czech science-fiction soap opera, "The Visitors" (who came here from the year 2484 to find his know-how about the transfer of continents to save their world in 2484 endangered by ... an asteroid). And I have even received uncritical e-mail messages from climate skeptics who really didn't turn out to be excessively skeptical in this particular case. ;-)

They had to believe the media and they had to believe that there is a high probability of a major catastrophe - but on the other hand, such beliefs allowed them to think that the people in NASA are less competent than 13-year-old schoolboys. It is apparently a very attractive combination of beliefs. It's the ultimate paradise. It's the end of the discrimination of children by the adults. Well, people often believe what they want to believe.

Of course, the story is nonsense and
NASA denies any mistake or communication with Nico. See also Nude Socialist.
Nowadays, if a story sounds sensational and if an underdog beats a top dog at the end, the story will immediately propagate. Too many people with too much influence are too unreasonable and under too comprehensive control of wishful thinking. And most journalists and editors are complete morons these days. They only know how to fine-tune the grammar of their sensational stories but they have no idea about the content, about the workings of the world, about the structure of human knowledge, and other things.

What a crazy world.
And similar bad judgements seem to be everywhere. For example, many - if not most - people think that it is comparably likely for one of the not-leading subdisciplines of theoretical physics such as loop quantum gravity to outsmart the top theoretical physicists and make a breakthrough that they can't make. Well, in principle, it can happen. But the probability that a community that hasn't ever been attracting the smartest people interested in theoretical physics could "outsmart" string theory is just very tiny, certainly much smaller than 10^{-3}.

The composition and chances of loop quantum gravity in the present era are something very different than string theory in the early days (late 1960s) because it was definitely attracting some of the best people - as measured by their early achievements in related but established subdisciplines such as hadronic physics.

In the history of physics, it has (almost) never happened that a complete underdog would outsmart those who could be expected, because of rational reasons, to be the leaders. Such things are of course always possible but they are unlikely and if we make extreme comparisons, they are extremely unlikely. Note that this probabilistic treatment is meant to generate realistic expectations, not guaranteed correct answers.

Einstein's chances

For example, it is a popular myth to think of Einstein as the ultimate underdog. But Einstein was no underdog. In 1905, he was certainly not as well-known as other physicists but he has been among the brightest students in his class and he was the brightest clerk in a patent office that is tightly connected with physics anyway. Whether he was employed by a patent office or a university clearly couldn't change his inherent ability to revolutionize our understanding of space and time.

Such things primarily depend on his talent, sharp thinking, right choice of principles, enthusiasm, patience, and hard work. Einstein has always had these advantages and all sane people had to know these things. But these features are something very different than fame, money, formal jobs, or social connections: these four things were clearly irrelevant for Einstein or any other man who revolutionized science.

Even if you made an estimate in which Einstein's chances were lower than the chances of Poincaré or Lorentz, they would be at most roughly 100 times smaller but you could find roughly 300 people who were on par with Einstein. When added together, their chances to discover special relativity were higher than the chances of e.g. Lorentz himself.

Similarly, Steve McIntyre was never a real underdog when he tried to fix the statistical treatment of surface temperature data or the "hockey stick". He has really been a child prodigy who is arguably smarter than any "conventional" climate scientist we know and he was doing pretty much the same work as GISS or MBH did and probably even more carefully - the only difference is that he was doing it at home and not at GISS but this difference is clearly not the aspect that could decide. And James Hansen is not exactly an extraordinary scientist who would be hard to be outsmarted.

But in more extreme situations, this counting involves very different numbers and ends up with a very different result. The probability that a 13-year-old kid learns everything so well that he can make a collision calculation more reliably than NASA is so small that the large number of kids in the world simply can't (practically) compensate this suppression.

A revolutionary insight about physics may arise in a patent office but probably not in an elementary school (or the younger portion of a classical gymnasium such as Nico's school).

Every reader who has been a child prodigy must remember that even though his or her abilities during the childhood were kind of extraordinary, he or she was still holding a lot of childish and naive opinions about advanced technical questions - e.g. those related to advanced quantum field theory and about the very question how much one has to learn in order to have qualified ideas about fundamental physics - that (nearly or completely) everyone in the world only learns after he or she is 18. A lot of learning, training, trials, and errors are usually necessary.

I am formulating these things in an excessively technical way because such conclusions should follow from common sense. In the history of mankind, it has probably never happened that a 13-year-old child outsmarted a leading institution of applied science to obtain a very different answer to a rather important question. Because it hasn't happened for thousands of years, a sane person should be able to conclude that it is unlikely that such a story takes place in April 2008. It doesn't mean that it can't happen but it does mean that a similar story is an extraordinary statement that requires extraordinary evidence.

However, people are not sane in this sense and they are ready to believe certain insanely unlikely stories without any evidence whatsoever.