Decromin developer's commentary

Initial idea

Some years, before the jam starts, I've had vague thoughts about what sort of thing I might want to make, if it fits with a more specific idea I have after I hear the theme. This year, earlier in the month, I'd had dental issues that caused a lot of pain, and I hadn't felt like doing much and didn't get much sleep, and while I was feeling better by the time the jam started, I still wasn't sure I'd have enough energy; combined with the fact that last year I got way worse sleep on the final night than even previous jams, my idea was "make something that takes less time/effort".

Also, last year, I'd seen one of those things where people post an image with all the months of the year and some art they'd made that month, and I'd made a few art things recently and I wanted to see if I could actually do at least one art thing each month. I hadn't done anything for this month yet, so I was hoping maybe I could do something that includes some sort of art.

And, finally, something I'd decided much earlier in the year that I wanted to do, but only remembered partway through the jam: I was thinking about how one of the things I liked in video games was having different world themes (like how Mario games tend to have desert, ice, etc.), and most of my jam games didn't, so I wanted to have some sort of different themes for different levels, even if it was something simple like a different palette or some graphic being hidden on some levels. (I think the only time I did that in a GMTK Jam was 1 Side 1 Jump, where different levels have different palettes, though I've also done it in games for other jams, and sort of in the post-jam version of Snail Scales.)

That said, after hearing the theme "count down", my ideas were:

The theme reveal video mentioned something about going backwards in time, so I had an idea to do one of those puzzle games where you have to interact with a recording of your past self, but your recording moves backwards in time. My idea was to have top-down grid-based movement, both to hopefully make things easier, and so that maybe counting the number of moves that you make could be involved (maybe something involving counting down the number of moves you have left, or something like that).
Something like Tetris, but instead of blocks falling, you have a timer counting down and you have to place the blocks before the timer runs out (and since they're not falling, you can just place them freely). I also considered having polyominos with more than four blocks, since those are interesting and I don't see them as much, and free placement might make that easier. (I think I also thought of the idea of decreasing numbers of blocks early on, since that would also fit with the countdown theme and would probably make things easier on the player, since you can fill in holes with smaller polyominos.)
Vaguer ideas: Something involving rockets, since there's the countdown to take-off. Something involving a cute-but-kind-of-underrated animal, since that seemed to work for Snail Scales (first thought was dragonfly, but then I realized it could be a bat, because Count Dracula).

I ended up going with the Tetris idea, because I felt that would be easier than something level-based (something level-based would require an extra step of actually doing level design), I couldn't think of anything interesting to put in the levels (I was sort of thinking some sort of Sokobon-ish stuff and maybe some laser reflection stuff? but those weren't that original and I didn't have any ideas for specific puzzles), and because the recording-of-your-past-self thing was something I'd already sort of done a couple years ago (the snails in Snail Scales are a recording of your past self, played at half speed, though it records your actions rather than your position so the snail might do something different than you did).

(Maybe I shouldn't care too much about similarity to what I did earlier years? Probably most people who play my current game won't be the same people who played my game previous years.)

More specific ideas

Once I thought of that, I was thinking, what should the condition for clearing the blocks/tiles be? Like, it could be that tiles clear when you complete a line, like Tetris, or when you get a group of tiles of the same color, but I didn't want to just copy existing games, and I wanted to incorporate counting down somehow. Maybe the tiles have numbers on them, and something involving consecutive numbers counting down.

Maybe adjacent consecutive numbers would get cleared, like having a 6 and a 5 next to each other, but then that wouldn't be about counting down, since the situation was symmetric.
Maybe you have to count down from the highest number? Then you're likely to end up clearing out 6's and 5's early before you can get to 2's and 1's (now that I'm thinking about it again, maybe that could be part of the challenge?).
An early thought I had was that if you put a six next to a five that would clear the six and five, but turn the five into something else (I was thinking of showing it as a rocket, because of the earlier idea to incorporate rockets somehow), and once you put a four next to that something else that would clear, and so on, but that seemed maybe too complicated?

Eventually I got the idea of only clearing tiles at the end of the round, which is what I ended up with in the final game.

Once I came up with that idea, I programmed it, but without the timer, just to get an idea of how well it would work, and it seemed like it was too hard to clear out the tiles. I tried a few things:

Changing how the numbers on the tiles are assigned. The initial hexomino is generated by starting with one tile in the center, then repeatedly adding tiles to a random free space adjacent to an existing tile; the polyominos after that are generated by removing the tiles in the opposite order they were added.
- I initially tried numbering the tiles in the order they were added, and that ended up with too many low numbers that were hard to clear (since smaller polyominos would only have low numbers).
- I tried numbering them in the opposite order, but that meant that fives were basically free, since the polyomino-generation algorithm always put fives next to sixes.
- I tried numbering them in a random order.
- I ended up with my initial idea, but having each polyomino increase the tile numbers by one, which means that lower numbers will tend to be near the middle where they're harder to get to, but rarer, but smaller polyominos will have higher numbers, solving the issue with my first idea.
I tried decreasing the initial polyomino to five tiles, and counting diagonals as adjacent. Without either of those changes it was too hard, with both of those changes or either by itself it was too easy, and then when I undid the changes it was just right.
I also implemented an idea I'd had earlier, where clearing a 1 tile would clear out adjacent tiles, which would both make it easier to clear more tiles and give an extra reward for clearing out a 1 tile.

Once things felt right (and I still hadn't implemented the timer), I noticed it seemed fun actually thinking through where the polyominos should be placed, so I decided to make an untimed variant (type A). I've done more testing with type A than with type B.

(In my original idea, I was thinking the main challenge would be finding a place where the polyomino fits, while leaving enough space for future polyominos, like in Tetris, but it turns out when I play I actually spend most of my time figuring out how to get all the numbers in the right places to get a good combo. Maybe changing that mindset is what made that easier, and also I think that influenced the untimed variant. When I play type B, I find myself spending more time making everything fit, like the original plan.)

I gave type A a maximum of six rounds, since an issue I've experienced with endless games is doing well enough that I can play them for so long I get kind of bored, but I don't want to intentionally lose. The game getting harder over time can help with that, which I think type B does sufficiently well (it gets faster and the board gets more tiles on it), but in type A the only way it gets harder is if the board gets more tiles, and it's possible to clear the whole board if you're good enough.

Graphics

The graphics have a beveled effect, like was common for buttons and such in computer interfaces before the shiny translucent appearance became more common, and then later boring flat design. I was thinking earlier this year about how I like the beveled appearance, and yet the menus in all of my games have flat buttons (or sometimes shadows); that might be why I chose this, I can't remember for sure. Anyhow, the beveled effect was one of the first graphical things I did, before I finished the main mechanics.

I had a graphical effect for the tiles breaking with some sparkles, because I remembered someone wanting more graphical effects in Corol Match. (Those graphical effects might not show up well on the background, though. That's why I added a dark option post-jam.)

I mentioned earlier that I wanted to do something graphical that was enough that I could count it as "art", and that I wanted some sort of world theme–like thing, and I had a game idea that was a bad fit for either of those. I could at least kind of sort of get something world-theme–like by changing the background color between rounds, and I had an idea for a background effect that involved 3D rainbowy lines and right angles (since that kind of fits in with the tiles being different colors), but I was worried that would be too distracting (maybe people would think the background was actually relevant to the gameplay) and also the background was one of the last things I did so I didn't have much time, so I went with a simplified, 2D version of the effect instead (and made my original idea as a separate thing outside the jam).

Music

In my recent jam games, I've been thinking, since I can compose music I probably should make some sort of new music for each game (and also not having things be silent will probably help my scores).

This time I came up with a melody, based on the idea of having the time signature of each measure count down, so I had two measures with six beats, then two with five, then two with four, etc. Once I wrote the melody, I copied it a couple times and made a few changes, like I did in Loop-D-Pool (my previous jam entry) (although I like how that music turned out better than this music).

I wanted it to be timed such that in type B, if you didn't click to place a polyomino, the 6/8 part would last until the first polyomino automatically placed, the 5/8 until the second one placed, and so on, but the amount of time I gave would make that too slow. (Also I tried at some point before increasing the amount of time you have to have it skip to the next part when you placed the piece, but I didn't like how that sounded.)

Transcript (PDF)

Things I'm not completely happy with

Game title: I was sort of thinking of calling it "Decromino", but while it uses polyominos, it's not based on the game Dominoes (and I'm not sure exactly how to play it), so I kind of didn't want to use that name, so I just dropped the "o". Not completely happy with it, but couldn't think of anything better.

Font: Most of my games use the browser's settings for serif, sans-serif, or monospace (exceptions being the series starting with 1S1J, which uses Verdana if you have it installed, and Walk and Jump, which uses a font made by someone on a forum I was on at the time). I wasn't completely happy with default monospace, but I kind of didn't want to include someone else's web font (since I don't usually use third-party stuff, also I'm not sure where to find good fonts) and couldn't think of a good font that people are likely to already have installed (and I don't even know what's likely to be installed these days).

Instructions: I kind of did the instructions at the last minute, so I did the easiest thing and just put a block of text on the screen. (I was thinking in the worst case scenario I could just put the instructions in the itch.io description, which is something I can do after the deadline, so I gave that lower priority.) I worried that the game would be too confusing. In Corol Match I had a full interactive tutorial, but that's because that was a part of the original plan, and I wasn't sure how to do that well for this game.

Not sure if the way I handled touch controls is ideal, but I'm also not sure if anyone rating games is actually using touch controls.

Timeline

Wednesday: spent most of the day just thinking about it in between watching YouTube videos; at the end of the day, started a project and wrote code to generate and draw a polyomino. (I was sort of thinking this was a particularly bad start, but the amount I got done is similar to the last two years.)

Thursday: got placing pieces and detecting which pieces should break done

Friday: got most of the game mechanics done, which was my goal (and have something that I feel happier with); also got an idea for music

Saturday: finished and uploaded a first version, then spent the rest of the day on graphics and audio

Sunday: some smallish changes, including making the timer for Type B more lenient