denis, I did assume that my points about your unbiased* statistics had been discussed before, but I was trying to illustrate a more general problem.
I'll give you a TL;DR of my other post.
You seem to be unconcerned about the readers of your writings:
- Your notation puts an unnecessary burden on the reader.
- Your stats are based on an unusual, never explicitly defined population and the reader is left wondering how significant the results actually are.
- The results are then interpreted over a different population without any justification or notice to the reader (puzzles appearing in their respective grid's minlex form → all puzzles).
Nevertheless, (supposedly) not realising that insulting people may result in them getting insulted is (AFAIK) a new low for you and something I wouldn't have expected, not even from you.
ghfick wrote:You note the insulting "claim to have a PhD in stats".
I'm sorry, I quoted the line to provide context for the next one, I should have deleted the first half of it.
Edit:
denis, maybe in your mind the set of all puzzles has always been the population in your stats, in which case they never were unbiased, in this post I thought of them as of unbiased statistics over the population of puzzles appearing in their respective grid's minlex form then being interpreted over the population of all puzzles (which of course produces bias).
Marek