That isn't my reading of his algorithm. It looks to me like it can only choose 1 out of 24 squares - there isn't the possibility of moving diagonally. Also the mod means that the edge effects (should) disappear.
The difference between the two simulations is as you say, however: he is constrained to the limited path but Cristoforo is not.
In reply to Re^2: Running a simulation - expected outcome
by hippo
in thread Running a simulation - expected outcome
by Cristoforo
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |