The starting point to the puzzle was the original idea for a Different Neighbors Araf- I wanted to do something with the variant that could only be down with the variant, and I had two main ideas. The one I ended up going with was to have the solve driven by multiple large regions forced to be the same size and avoid touching- the T tetromino arrangement seemed to be the best way to start that off. Then I realized I needed ways to corral these large regions, and so I quickly settled into a visual theme of all T tetrominoes. This gave some interesting parity constraints and really let me drive the solve downwards. Sizewise, I actually didn't know how big the puzzle or the 36 regions were going to be until I was about 60% through constructing, as I originally had a 13x13 grid that needed more room to keep the two regions the same size. I also briefly toyed with 4 large regions on a 15x15 or other variations before getting started, but this definitely seemed best.
The biggest problem for constructing was getting the ending to be unique. My first version had 2 solutions as I miscounted something, my fix had 2 solutions, my fix to that had 2 solutions... the entire bottom third underwent several revisions and finagling something nice eventually required me to adjust the top left, too! Ultimately this was the 10th or so version I testsolved and it finally came out unique, but it also wasn't what I would consider a "nice" ending as it uses a global count to minimize a bunch of regions. I did manage to force the 36s to touch diagonally at the end which was nice, though.
I mentioned having two main ideas for the Sunday Stumper, and I'll probably make the other one someday - I originally thought of forcing a checkerboard coloring (say, regions of size 6 and 7 stemming from only 5s and 8s) but the idea shifted a bit. Hopefully the puzzlemaking bug will strike again soon because I do want the adjusted idea to exist, but you know, I have to make it first :)
No comments:
Post a Comment