ad hoc Gaussian Arrival Distribution
The context of the Law of Large Numbers related in this typical Novella hack job is the
ad hoc with intuitive averages (Gaussian) serendipity of "
If you consider the math more thoroughly (an ambiguous appeal to authority)
, you will see that amazing coincidences should happen all the time." "
If we do the math (Novella can't do the math), then it becomes clear that very unlikely events should happen all the time, given enough opportunity." This is the
ad hoc context of the LOLN.
(notice that Novella essentially said the same thing twice, in order to make his fluff piece appear more intimidating and substantial)
ex ante (Other than Gaussian)
ad hoc is not the same as an
ex ante forecast, as happened with Andrew here in this case. In a forecast, this is not an 'amazing coincidence', it is a predicted amazing coincidence, something completely different, - AND something which is the basis of deductive science (prediction). There is NOT 'enough opportunity' in Novella's pseudo-statistics lingo. It is
ex ante. Nor do we understand the concavity nor convexity of this particular
ex ante prediction over time or circumstance. Therefore it cannot be easily intuited as being subject to the LOLN in this manner.
The Law of Large Numbers has validity, but Steven Novella is not qualified to assign that validity to any particular case scenario, because:
1. He bears agency (not just bias)
2. He perceives all probability arrivals as being Gaussian
3. He is not trained in the disciplines of probability statistics and hypothesis/scenario modeling - nor does he understand it. He just holds a couple canned principles which he grasps about 70%.
4. He bungles the match between principle and example scenario regularly
5. He fails to perceive the difference between ad hoc and ex ante.
6. He bears a club celebrity conflict of interest - and cannot afford to actually look when he risks being wrong (anti-skepticism).
Nice try, but no cigar.