Sunday, October 25, 2020

Puzzles #231, #232 & #233 - Libra (Minarism)

I'm running out of ways to introduce the Puzzle Zodiac but I'm only halfway through posting the puzzles from it help

Anyway part of my genre selection was to have 3 each of shading, loop/path, region division, and number/object placement puzzles. Even within those types, I wanted variety- and with 1.5 object placement and 1 latin square, naturally I went for another latin square in Minarism. Ripple Effect was something I considered here, with Balance Loop for Libra, but things matched up better and had a better aesthetic balance like this, I think.

For the example puzzle, I went with an antisymmetric theme which is always interesting in a type like this. The solve is pretty straightforward and the construction is like this more to look nice and for my own enjoyment than anything else.
Puzzle 13 on the test, worth 20 points, and themed on the Libra sign. I wanted to focus more on the difference clues here, and started with the 1-5 string along the top and built from there.
Puzzle 14 on the test, worth 70 points, and symmetrically themed. Minarism is one of those types where placing the same clues symmetrically doesn't result in the same deductions. As is probably obvious, as a counterpoint to puzzle 13 I went with an inequality focus here. This slot had the only puzzle other than the Foursight example I scrapped, because when constructing it I imagined the break-in and everything after it wasn't deducible. Somehow it still turned out unique, but it wasn't suited for posting on this blog, let alone a competition!

Saturday, October 24, 2020

Puzzles #228, #229 & #230 - Virgo (Star Battle Sudoku)

As much as it's not my favorite thing to solve, I knew for the Puzzle Zodiac I should include some sort of Sudoku puzzle. So when browsing variants I spotted Star Battle Sudoku and thought that it would be a perfect fit, merging number placement and object placement into one puzzle. For the puzz.link interfaces, the highest placeable number(s) represent stars, and the solution should contain (1-5 and 1 star) or (1-7 and 2 stars) in every row, column and region with no stars sharing an edge or corner.
For the example puzzle, my goal was to make sure that it was fairly easy. To that end, I placed an S into the givens because why not?
Puzzle 11 on the test, worth 20 points, and nicknamed "5 star confusion". This is a puzzle that just kind of ended up working out, as I placed in some numbers arbitrarily and deduced as much as I could, which was most of the grid. My only goal was to have 5 and star disambiguated by the Star Battle touching constraint, and this was free points for most people.
Puzzle 12 on the test, worth 50 points, and themed on the Virgo sign. It took me a very long time to hit on an arrangement of givens that looked like Virgo, and even once I did I had way too many givens to make an interesting puzzle and I kept running into dead ends. Eventually I removed some givens and tried moving the symbol around the grid before finally hitting on this arrangement. The rest of the fill took a bit of work to resolve, too. Ultimately any sudoku contest regulars blasted through this puzzle in no time and often earned over 20 points per minute. Does that mean it was overvalued? I don't think so- I balanced the points around more general puzzle people, which means that anyone who specializes in a type would earn more points per minute, compared to making a couple puzzles bad points per minute for most people.

Friday, October 23, 2020

Puzzles #225, #226 & #227 - Leo (Akari)

Fun fact about the Puzzle Zodiac: my original thought for Leo was Nurimisaki, not Akari, playing on cats hiding in corners instead of chasing lights. Both contest puzzles here are 10x10, the only time in the contest that both puzzles of the same genre are the same size. Even still, the second one is probably a good bit harder?
The fun thing about making example puzzles is that I didn't have to worry as much about an interesting solve path as much as demonstrating the key points of the type. I enjoyed trying to add in an aesthetic flair for these.
Puzzle 9 on the test, worth 20 points, and themed on the Leo sign. As usual for the zodiac themes, I placed down the blocks first and then tried to clue it in an interesting and non-trivial way.
Puzzle 10 on the test, worth 30 points, and no real theme beyond the L shaped blocks. I had a lot of trouble trying to come up with a good theme for this puzzle, and an even harder time trying to make it non-trivial. It's really easy for an Akari clue to break all the logic up to that point...

Thursday, October 22, 2020

Puzzles #221 - #224 - Cancer (Foursight)

So this is probably the weirdest and least thematic pick for the Puzzle Zodiac, but I swear I have a good...ish reason for it. So all the way back last year when I was first making Foursight puzzles, I hit on a nice given layout that, once I'd made the puzzle, I realized looked like the Cancer symbol. Afterwards I had an idea to theme a set of puzzles around all 13 zodiac signs, and shortly after had the idea to turn it into a contest. So when I sat down to pick the rest of the types for it, shading puzzles were the hardest to pick by a longshot just because of how strange this pick was.

My first attempt at an example puzzle... didn't go so well, so I made a second one to replace it for the instruction booklet.
Second attempt at an example, with a bit more interesting theming and logic. Still not the greatest example, in hindsight, but Foursight is not a type that works well on small grids.
Puzzle 7 on the test, worth 40 points, and themed on the Cancer sign. The puzzle that started it all, over a year ago... according to my personal spreadsheet, this was the 64th puzzle I made. The next number for a puzzle on this test? 116. Anyway, this isn't that tough for a Foursight but the type is inherently a slow and tricky solve.
Puzzle 8 on the test, worth 70 points, and themed on the main diagonal. I knew from the time I made the previous puzzle that I wanted to have some larger empty spaces and settled on this layout, but it took me until construction 136 to actually get around to making this puzzle. There's a couple parts that are tough to see here and this was the second least solved puzzle on the test- not really surprising to me given the type, however.

Wednesday, October 21, 2020

Puzzles #218, #219 & #220 - Gemini (Double Chocolate)

One of the easiest genre picks for the Puzzle Zodiac was Gemini. Double Chocolate literally has double in the name! I tried to keep a balance of different types of puzzle in the contest, and region division ended up filled first due to another obvious choice later on.

I themed the example puzzle on the fact that it's an example. Probably overclued, but I wanted to make sure that it showcased a reflection to make it clear that those are allowed.

Puzzle 5 on the test, worth 30 points, and themed on the Gemini sign. It took me longer to settle on a shaded cell pattern for this grid than it did to make the rest of the puzzle- I've always been partial to different kinds of symmetry for shaded cells and numbers in this type, as it can result in some wildly varied kinds of deductions.

Puzzle 6 on the test, worth a whopping 80 points, and themed vaguely on a chocolate wrapper. This is one of those grids where a couple things just kind of worked out, in large part due to the logical theming. But still, those corners gave me a ton of trouble and I kept running into endings that were impossible to satisfy at all when constructing this.

Make sure to check back tomorrow for the Foursights- they won't show up on puzz.link/db, which is where I know most of my traffic comes from :P

Tuesday, October 20, 2020

Puzzles #215, #216 & #217 - Taurus (Araf)

As I think might be clear by now, for Puzzle Zodiac I tried to pick puzzle genres that were at least somewhat thematic to the zodiac sign they represented. For Taurus, the most defining feature is the two large horns, and that reminded me of Araf for some reason.
For the example puzzle, I wasn't sure what kind of theming I wanted to go for until I had the thought of placing a 1234 in the top corner. I didn't dare dream that this sort of theming would end up creating a unique puzzle, but it did!
Puzzle 3 on the test, worth 30 points, and themed on the Taurus sign. I tinkered with placing givens on the border at first to try to get the layout correct but ultimately couldn't find a way to make it look nice except this 6x6. Once I had the layout in place, it was easy to embed a few deductions for a fairly smooth solve.
Puzzle 4 on the test, worth 50 points, and themed around partial arithmetic sequences on the outside edges, and a hidden logical theme. Araf can always be a bit of a fiddly type to get to resolve the way you want it to and this puzzle was no exception. It's always a good feeling when the acceptable lower and upper bounds for a clue are equal- see if you can find the one I'm referring to!

Monday, October 19, 2020

Puzzles #212, #213 & #214 - Aries (Slitherlink S&W)

My puzzle contest over at LMI, Puzzle Zodiac, concluded recently. The consensus seems to be that the set of 25 puzzles (and 13 examples) is of extremely high quality, and also quite tough as a competition set. This consensus is borne out in the results, as the contest had only one finisher! One of the toughest puzzles in the set was the second Slitherlink (Sheep and Wolves) puzzle, and I'm very pleased with how all of these puzzles turned out. For the puzz.link links, all sheep (S) are replaced by a 4 and all wolves (W) are replaced with a ?.

puzz.link interface      penpa interface

old puzz.link interface

This was the example puzzle for the type- I wanted to make a relatively simple puzzle with a touch of antisymmetry.

puzz.link interface      penpa interface

old puzz.link interface

Puzzle 1 on the test, worth 30 points, and themed on the Aries sign. I started by placing down all of the sheep and wolves, and then tried to keep the pattern largely intact with numeric clues to specify the rest of the puzzle. It turns out there's a lot of interesting deductions that follow from the layout here.

puzz.link interface      penpa interface

old puzz.link interface

Puzzle 2 on the test, worth 60 points, and themed around a domino pattern with sheep and wolves on opposite sides. In contrast to the other slitherlink, where I used the variant rule to drive the layout, I wanted to use the variant sparingly but still have it drive large parts of the solve. My favorite part of the solve is how the left side resolves- there's some intricate steps to discover that I put a lot of effort into trying to create and preserve.

Sunday, October 18, 2020

Puzzle #211 - Surprising Collision (Yajisan-Kazusan to Ringring)

Logic Showcase 19 recently concluded, which tasked constructors with making a Yajisan-Kazusan puzzle without any trivially unsatisfiable clues. Somehow, both me and someone else elected to make a relay puzzle to the same genre- Ringring! Relay puzzles involve solving one puzzle and using the solution as input for another puzzle. In this case, the Yajisan-Kazusan shaded cells must be solved as a Ringring puzzle. This puzzle placed in the middle of a strong field of entries, and I think it has a nice flow to the solve.

puzz.link interface      empty Ringring interface      penpa interface

Saturday, October 17, 2020

Puzzles #208, #209 & #210 - Medusa (Snake)

When I make Snake puzzles, it seems they turn out pretty non-standard. Something about the usual style just doesn't seem to click with me. The first one today is heavily themed around one specific deduction, the second one has no outside clues and a hopefully interesting solve path, and the third one is more visually themed and quite difficult. There do exist a set of reasonable deductions and lookaheads to solve it, though I wouldn't fault anyone if they don't find them.

One more puzzle until I'll start posting the Puzzle Zodiac puzzles here.

Wednesday, October 14, 2020

Puzzles #204 - #207 - Four Sided Fantasy (Shikaku)

I made a few Shikaku puzzles to experiment with unclued circles, as cluing every circle can easily short-circuit deductions and logical paths. These are also fairly easy to make quickly without compromising being interesting and varied, so that's neat too!