Making Descending
Want to try keeping a public devlog! I'm finishing up a game called Descending; it's about drills, paper, and poisonous gas. Feels odd do start the blog near the end, but whatever.
→ The game's out! Please play it here before reading.
Pushed the game to a playtestable state today, and had a playtest! As usual, confirmed two things that I manage to forget every new project—that I underestimate complexity, and that the tutorials should be way easier than it feels like.
Also today the project entered the stage where I actually feel like it works! It's always easier from here—I can focus on cool designery things like pacing and tension and stuff, instead of stumbling around in the dark. If I started the blog earlier, all notes would just be spiels about how nothing works. Could have entered this stage way earlier if I just pushed things I wasn't happy with to a prototype stage. Oh well!
I assumed a lot of completely obscure things to be obvious, and the playtest proved me very wrong. There's at least two things that I will throw away completely—the game is difficult enough right now, I'm only gonna be simplifying from here.
I want to put more non-systemic substance into the game. Right now it's all about raw mechanical tension and not so much about vibes. Want to strike a balance between the two. I usually fail at it and lean way towards systemic stuff.
I think temperature mechanics are fine, they just need better feedback and two posters of instructions instead of three. It feels weird if the first mechanic to learn requires two posters though, while pressure just needs one.
Pressure mechanics are definetely not good. I like the switch interactions, and controlling RPM feels good, but the current system (a specific configuration adds a specific number to the RPM, with an apply delay) is weird. The delay is also just obscure and I don't want to explain it. Maybe you look at the speed of change and stabilize based on it... sounds solid.
Battery mechanics... I like that the perfomance reports have utility. I also like that the batteries are mechanically different from restarting—otherwise why even bother with them? I should probably allow to change batteries at any point so it's actually like a downtime prevention thing. Also make getting a new battery very slow. Not sure how to properly feedback the failed battery though; just not applying it at all will be confusing, and causing slowdowns is obtuse. Actually, just not giving a full charge should be clear enough.
While I'm testing the mechanics, some more things to consider.
There's a lot of mechanics to learn: GAS, HOLES, OUTSIDE HOLES, BATTERIES, ADVANCED BATTERIES, TEMPERATURE, PRESSURE, and MALFUNCTIONS. I don't want the game to be overly long (well, I don't mind, but I don't think it's engaging enough to warrant it). I also don't want to throw a new mechanic into the player every day... I want some downtime to settle in. I can actually just unlock each mechanic for one well at a time so that there's growing complexity without new tutorials. The pacing is definetely something I can only understand properly through playtesting.
I want a reason to go outside a little more. Programmed a falling rock today, looks fun, so I want the player to see it properly and not just sit inside all day.
The clock doesn't work at all! There's no sense of urgency until the last few days. I'm not sure if I can just solve it with balancing, since there's a very large gap between perfect play and relaxed play. Because there's no sense of urgency, all the problems just feel mildly annoying instead of life-threatening. Maybe I can connect some control panel activity to gas outside... Gas also doesn't feel scary at all, it actually does nothing apart from "please seal holes beforehand thanks".
The new control panel mechanics seem to work fine as this point. Now experimenting with the pacing—will hold a playtest tomorrow to see how it feels.
What is the vibe that I want the game to have? Initial concept was methodical, but the design naturally went into the direction of being somewhat frantic. I think what I want is CONTRAST. Action-packed days followed by slow, more relaxed ones; tension & release, play & narrative.
Went for a walk and thought I had a neat solution for a few of my problems—just have the gas downstairs as well, making it inaccessible during a storm, and forcing you to plan more carefully. Went to implement it and realized it's incompatible with a visual decision I really like (the see-through grating on the floor would pass gas through). Was very frustrating and discouraging. Luckily, came up with a similar idea that seems to work fine—gas can sometimes tear the paper off!
This solves most of the problems mentioned before: lack of pressure, outside being useless, gas not doing much. I do have a worm in my brain begging me to find one more gameplay twist; I am considering continuing the game for a bit after the story twist, or making an alternative ending you can discover... but I don't think that's reasonable. Maybe it's okay for the game to just be what it is.
One big thing left (apart from actually pushing the direction over the finish line; very high-resistance activity) is the narrative. Morgan had an idea with a journal you can discover in a hidden locker, and I think that's sick. I can also tape some papers to the falling rocks, or have cryptic messages on the instructions that can reveal secret control panel codes. What do I want to say with it, though? As it stands, the game's story is annoyingly, painfully cliche. I'd like to make the sub-narrative have something special in it... just have to find what it is, I guess.
Right now the story is the classic "EVIL CORPORATION USES YOU THEN THROWS YOU AWAY". It's definitely not deeply interesting, just serviceable. What can I say with the sub-story? "Wow, it really sucks to be abused by an evil company"? "Actually, I'm the evil one"? "Actually, it's for the greater good"? "Actually, someone's good trying to help me"? None of those really speak to me, I don't care.
I think I care about personal stories. Can I make the narrative be about the previous guy? You find his notes, you get letters addressed to him. A good person, easy to care about. Leaving tips for the player. When the ending hits, you discover his last note; referencing your gameplay actions, having some message. Stronger punch for the large narrative and a proper climax for the personal one.
There's no need for an alternative ending, because the story isn't even about you at all. I like it.
Currently in the "it's so over" stage of the devlopment cycle.
For the last two days just quickly iterated—running through the game, writing down problems, fixing them, and running through the game again. The thing doesn't really seem to work, everything is way too complicated, underexplained, annoying. Still ran a playtest to not get too caught up in my own head; sure enough, it's really bad.
From here I think the only thing I can do is radical simplification, both in complexity and difficulty. Take my neat interconnected system and disassemble it, make it unattractive but serviceable, dig for the one core thing in the design that I enjoy and throw away as many other thing as I can... I've got some thinking to do. I've built up some trust in the process over the years, but this game is probably the most systemically complicated thing I made yet. Just need to push through I guess.
Took some days away from the project. Then simplified things as planned, but not too much; actually, just let the played take their time learning before escalating difficulty separately from introducing new mechanics, and it feels fine (playtests will show if I'm dead wrong soon). Still some unanswered questions, but it's a matter of shuffling things around now.
We found a small stray kitten after the last note, so I had my mind very occupied. But now it's time to really wrap up the project :)
Removed the battery mechanics completely; the other ones are complex enough. I think the paper management is way more essential to the design, so will try to push the day-end papers into that direction as well, by putting corrections to other papers there, or something like that.
Very obvious in hindsight decision that came to me way too late was to just keep it to one drill while the player learns the mechanics. Was crazy to just allow the player to move on before they learned how anything worked, as well as overwhelming them with two tasks at once. So now the escalation is MECHANICS → PAPER MANAGEMENT → TWO DRILLS → GAS BREACHES → TIMER and then the end. Doesn't sound like a short playtime, so I hope the few mechanics there are will be enough to keep the player entertained if I sprinkle some story beats and variation here and there.
Took some time to simplify testing for myself—made some console commands, editor tools, save loading. Definetely very worth it, would have saved myself hours of work if I did it all earlier.
This game is taking a very long time! First pre-production took me more than a month, and now I'm at a month of trying to push the thing over the finish line; for some reason it just keeps going and going. It's partially a million little technical problems or polish I'm chipping away at, but also the design iterations just take way longer when the game is so long (compared to games I usually make).
Kind of finalized progression today, don't think there's much left to change. Stuck to the escalation described in the last note. I tried "corrections" in papers I mentioned, they work fantastically from a systems standpoint... and I'm probably not keeping them, they feel like bloat. Eh.
Now that I got to almost properly play the game from start to finish, I don't feel like it's not working anymore, I just really hate it. I mean, I objectively can't properly judge it until I get it playtested, but from my point of view it's a deeply lame game—boring, slow, overwhelming, lacking direction. Gotta trust the process though.
Now that the system is playable, I realize that it's not a very good system. Sealing holes with papers feels okay, but the concept of "choose what information to give up" doesn't really work with such a rigid set of information: it's either trivial or extremely frustrating if you actually have to put away something important. Like, the only tradeoff you can get with this control panel is just COMPLETELY LOSE ACCESS TO A THING TO SEAL A HOLE which is not really a tradeoff you can consider. So papers are just a gimmick after all.
The final day of the game felt extremely bland (you just wake up to a drill failure), so I changed it up and it's actually pretty cool now, builds up the tension and ends with a little explosion. So that's cool.
One very fun observation is how much the game ended up matching its very first pitch—almost word for word!—after straying away pretty far from it. The difference is only in the paper system and in being way less complex (I can't believe I thought operating FOUR drills at once was a good idea).
The game's out! Didn't make many new design decisions since the last note, just playtests and fixes-fixes-fixes. Now it's time for some early retrospection.
I think the game is fine and plays well, even though not nearly as good as it could've been. The ending was the most important part to me and looks like it works, even though I still feel like it reeealy lacks a punch—but couldn't figure out how to make it punchier. I'm proud of some little design decisions, and UI is pretty good. I think I completely missed the mark on the VIBES but I still don't quite understand what's wrong. Is the room too big, is there not enough junk around, is the lighting wrong, or the sounds, or the writing? I now have a baseline to reflect on when playing games that are actually atmospheric.
The development took waaaaaay too long! I expected to finish in three weeks, and it took four months. I need to commit to decisions earlier, start playtesting earlier, and expect the initial scope to be too much (because it was). I also planned the game to be 15 minutes long, and it takes like 50 to beat (insane) and people seem to be engaged throughout (even more insane).
The game turned out to be extremely derivative of Unsorted Horror in its aesthetic and in many design elements. That makes me unhappy, I don't feel... a strong sense of authorship I guess. Nothing feels deserved. So from now I want to stray further from drawing too much from works of others.
All in all reception seems to be good and that's cool. I also enjoyed keeping public notes, even though they turned out to be very rambly and somewhat depressing.