Ah okay I see. I'll give it some thought and adjust my test case to be clearer.

Added random boards. Does this solve the issue?

Thanks, your help has been really great! Nice solution btw :). What do you mean by "Positions after first move generally don't count."?

Ah yes I see what you're saying. I've had to adjust my code somewhat. It looks like I wasn't counting unique solutions myself. I have adjusted however to not include end points that can reached in only one move which you have in your solution. So, for example [3, 7] cannot be reached in two moves. It can be reached in one, but a second move would move it away from that position. Should I make it clearer that that is one of the rules?

Which test case is this for? Also the start position should be included in the solution since the knight can return to its original position. I have edited the description to clarify this. Also I noticed a problem with my own solution and so have updated that too. Let me know how it goes this time. Thanks!