October, on the other hand, I can’t seem to be able to solve at all. Ouch…

Okay, I managed to do it, it was way tough. November was hard, but not as much, and December was a pushover. I wonder if this pattern has any meaning to you?

Regarding the ‘aiming’, probably a good solution would have been to only read a ‘ray’ pointing in the precise cardinal directions of each box, and if the cursor is at an intersection of these, prefer the one closest, or if at the same distance, the one previously selected.

It was sometimes hard to tell which box is the “cursor” at any one moment– like, when I move the cursor in a certain way, which arrow will be pointing toward the cursor and which will be pointing away. When the boxes got very close it seemed like they would start “fighting” over which would point toward and which would point away.