8

The question Did Kim Jong-il Score 11 holes in one in golf? received an answer, that has since been deleted in response to flags.

The justification for the deletion seems to have been disputed in the (now deleted) comments, so this is an opportunity to continue the discussion in a more productive location.

The answer attempted to prove that Kim Jong-il could not have made the score, due to its statistical improbability.

The answer had two problems that were fixable:

  • its mathematical analysis contained major flaws and needed to be redone.
  • it made reference to a casually racist meme that had no bearing on the matter.

However, it had some fundamental flaws that made it a "broken window":

  • It did not attempt to introduce facts about the actual case; it did not address the claim.

To whit: The Korean government made a claim that Kim Jong-il performed a super-human (but not supernatural) feat, demonstrating he was no ordinary person. The answer showed that it was extremely unlikely that an ordinary person could achieve such a feat. There is no conflict between these two statements.

  • Arguably, it provided original research - a new model of the chances of getting holes-in-one. Personally, I think this issue is on a spectrum, so we can debate if this answer went too far.

Aside: The issue of whether mathematical proofs are allowed has come up before. We consistently see that:

  • the community is not terribly developed in its reviews of mathematical proofs; ones with major flaws get upvoted.
  • more importantly, mathematical proofs rely heavily on premises (i.e. assumptions). These assumptions need to be made explicit and clearly referenced. This is rarely done well.

We should be giving mathematical proofs far more scrutiny, and not voting them up because they have the appearance of rigour.


This meta-question is to inform people what happened, and if they are not happy, to give them a chance to come up with better community standards for the mods to follow in these cases.

7

Mathematical proofs are fine, if they start from facts and lead to facts. In fact, we rely on them all the time.

I disagree that the answer was bad simply because of the maths. It was bad because it took unsound assumptions and used incorrect maths to disprove a imprecise claim with a strawman, not answering the question in the least!

Let me simplify:

  • Q -- Was Kim Jong-Il so good at golf that he scored 11 aces in a 18 golf course?
  • A -- No because doing 11 holes-in-one is impossible.

11 holes in one is not impossible and I can prove it with maths.

Let's say that in order to declare this event "possible" it must have 1% chance of occurring randomly. There are 18 courses, 11 of which must be aced. How good must Kim Jong-Il be in order to have 1% chance of doing it by chance?

The result is 37%. So if he can make a hole-in-one 37% of the time, he can make 11 holes-in-one in a 18 course track 1% of the time.

The question now becomes: can we prove or disprove that he was that good? Maths is taking us from somewhere to somewhere else, but not answering the question for us.

A correct answer must:

  • verify the facts of the claim.
    In this particular case it's fairly easy to use google and trace back the original claim to 5 holes and not 11 -- thus disproving the claim directly. The answer obviously did not rely on any research in this direction. The OP did not investigate the facts.

  • distinguish between reality and toy models.
    The maths and statistics assumptions of the answer were all over the place, simply to justify a preconceived point of view. One can't use the Drake equation to prove that aliens exist. Drake's equation doesn't prove anything. Aliens exist or they do not - independently from the toy model of the equation.
    In physics, gendankenexperiments are useful theoretical tools, but thought experiments do not prove anything. Every theory must be verified by experiment. Just because it "works" in one's head, doesn't make it an accurate description of practical reality.

  • answer the particular question directly, not twist what is asked in order to answer a different question.
    The answer basicaly only proved that if a hole in one is really difficult, then eleven holes in one are much more difficult. On the other hand the claim was that Kim Jong-Il was an extremely good player that routinely scored holes-in-one. Therefore, the premise above is not valid.

1

If we are willing to accept a disproof of 11-holes in one based on math/stats/probability, would we accept a similar disproof for aliens or yowies?

If not, how are these distinguished?

  • Why do you think these disproof are similar? Some statistical method are more notorious than the other. – Metta World Peace Apr 13 '13 at 13:04
  • 1
    My answer (subject to being persuaded otherwise): We aren't willing to accept the antecedent, making the question moot. i.e. We don't accept a pure mathematical disproof on 11 holes; we don't accept a similar disproof for aliens/yowies; we don't distinguish between them. – Oddthinking Apr 14 '13 at 3:42
  • @Sancho: I'm a layman of statistics. But it seems to me, as to the question of "11-holes in one", there's a large sample of people who are better trained in Golf than Kim Jong-il, from which an upperbond of probability can be ubtained. But there's simply no sample available to deduce the probability of aliens or yowies. In other words, I show more confidence in large sample theory than Bayesian inference theory. – Metta World Peace Apr 14 '13 at 4:05
0

I think that answering certain kinds of question (Not necessarily this one) with Maths is a brilliant, scientific, producible way to answer a question. By its very nature it is hard fact. However you rightly point out the pitfalls when using it with assumptions, and that there should be a high standard of

  • Using sensible assumptions
  • Using reliable, pre-defined assumptions.

The first is a no-brainer, the second is essential to stop the temptation to use numbers that "just fit" and give the desired results.

I have one suggestion, the answer banners that say, for example "This answer references current events" and "This post does not cite references..." - how hard would it be to add one along the lines of

This answer relies on Mathematical assumptions, please use extra caution before upvoting an answer without checking the reliability of the assumptions used.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .