Get the puzzle pack here!
This post won't have preview images or rules for each puzzle, but I will link them all here. The pack has a penpa link with answer check and rules written in alongside an image, so the pack can be solved physically or digitally without issue. The U puzzle has the variant of unequal lengths.
F I L N P T U V W X Y Z
puzz.links for P, W and X can also be used. There was no way to make one for Z, unfortunately. Comments after the break.
I really wanted to have each pentomino themed puzzle use givens in the shape of that pentomino, with the puzzle name starting with the same letter. So the first thing I did was run down the list, bearing in mind that construction would need to be themed, and started with Fillomino for F. After toying with layouts a bit I found the one I used since it seemed workable. Coming up with the break-in took as long as making the rest of the puzzle, honestly!
Icebarn was an obvious choice to me for I, though constructing it proved challenging. I kept running into situations where I needed more useful crossings than I had available, until I figured out why and built the rest of the puzzle around that fact. In my opinion this is probably the 4th hardest puzzle in the set, and the 3 hardest aren't even close.
Originally for L I considered using a Liar puzzle, but couldn't come up with a good type to use for that. So I fell back on LITS and started with those 4 L shaped rooms. Placing some other deductions revealed that with a bit of extra effort, there was room for a fun surprise.
N was the only letter where I had to completely scrap a construction! Originally, I tried to make a Nurikabe (Pentomino) with only N clues... but after about an hour of trying on various grid sizes I was completely unable to even find a valid solution before placing clues. Nuribou wasn't an obvious pick and the theming is a bit less clear, but I love how this one turned out. Even if I ran into construction dead ends a ton of times on this one, too.
Pentominous was a no-brainer. What wasn't was settling on a given layout that I thought could specify the rest of the grid, and my first attempt failed completely at even having a solution. My second try went much better, having 2 solutions, and I was able to make it work by redoing the second half of the puzzle. As a secondary goal, I made sure no letter was duplicated :)
T and U ended up intertwined, as my plan was what I eventually went with: a Tapa and Unequal Lengths variant of a loop puzzle. However, after making the Tapa and feeling a bit dissatisfied from the clue density (though I made it work well) and starting on an UL Angle Loop, I had the idea to use Tapa-Like Loop and Uso-One. The theming might have even worked better there, but I didn't want to make an Uso-One and had a breakthrough on the Angle Loop. And so I ended up with what I view as the weakest pair in the set... while still being great puzzles!
There aren't a lot of V puzzles so I had to go with View. View puzzles don't tend to have dense clues, either, so when I ended up making a challenging one with this many clues and that much open space...I was very pleased. The original version was much easier, with a 5, but I was able to tweak it to something more logically interesting.
W had to be a word puzzle, and I remembered liking Word Division from the Casual and Word round on LMI last year. Theming a puzzle with a letter in every cell would still be doable with Ws of Ws, and so I came up with as many puzzle types with a W in the name as I could to pare down for my word list. I almost used Hashiwokakero here, and made sure to have a small surprise in store for those who don't check carefully... Constructing this was pretty easy once I made a frequency table, too.
X was even rarer so I went with the tried and true X-Sums Sudoku. This construction gave me such a difficult time because I made simple addition errors early in the puzzle twice, both times requiring everything afterwards to be scrapped. The third try ended up with two solutions and rather than backtrack and try to thread that needle again, I elected to just add the necessary clue in a place that wouldn't really damage my logic and its opposite. But this version still had an arithmetic error which fortunately didn't affect the solve- I could change the outside value to the correct one and everything still came out to what I expected. Crisis averted!
Yajilin. It couldn't be anything but Yajilin for Y. I went for an ambitious corners only placement for the givens and managed to make it work fairly, with what I believe is an unavoidable interesting step with the 3. This one only took about 10 minutes to make and test, which gave me enough time to try for the Z that night...
Zero TomTom. On a 9x9 grid. Oh boy, was this an ambitious theme to attempt. As seems to be tradition with me and large Latin Square genres, one of my very first deductions placed into the puzzle was faulty and I didn't catch it until after I'd already embedded and worked through some very intricate steps based on that placement- you know, the whole puzzle. Rather than scrap it and start over, I spent about 20 minutes thinking about how I could fix it without breaking the aesthetic (just giving the deduction in question was a lame possibility, if it came to that) and came to the conclusion that it couldn't be done. Then while relaxing before sleep I had an insight, added 2 clues, worked through the start until the point where the missed alternative didn't work and internally celebrated. Then the testsolve the next day ran into another problem after most of the tough steps, though this one was thankfully fixable by adding yet another clue. I almost clued the last 2 regions at this point just for completion.
I'm extremely proud of this set and especially the final Nuribou and Zero TomTom. I hope that this set is/was fun to solve and that my comments here were interesting. Until next time, happy puzzling!
No comments:
Post a Comment