@louiscouture there are infinitely many such equations that would not converge to anything, so if you just pick even a bajilion of them at random you will never get one that does the thing you want. If it was that easy all problems would be solved by now, by just trying out some random formulas.
The process is clearly not picked at random, one condition is meant to increase the number while the other decrease it, so you gotta explain why n/2 always wins at the end and gets you down to 1, instead of 3n+1 flinging you off to infinity.