Sure I do know it is a horrible concept and never BIP39 compliant, however please assist fulfill my curiosity with this “thought experiment” and utilized maths train.
Two questions, one for 12 phrases, the opposite for twenty-four.
Assuming “completely random” entropy to begin, and sure the ultimate checksum phrase will get modified with the brand new ordering, not itself included within the alphabetical type.
Simply how a lot entropy is misplaced within the phrase’s re-sorting, in comparison with the unique random type?
I do not want the “present your work” maths, extra like virtually talking about present brute-forcing capabilities, “you’d lose 70% of your entropy, and 40% respectively”
If not too inconvenient, one thing ELI5 like, what number of supercomputer-years of effort’s distinction?