[005] Breakthroughs...? Lessons from AnalgesicGame5

In my previous post 3 months ago I proposed some ideas for exploration through the Shuffled World - https://han-tani.itch.io/shuffled-world-codename/devlog/498463/004-shuffled-worl... . Although they were still loose, and I wasn't particularly excited about prototyping any of them.

Recently, I've been doing a lot of level design for our next game #AnalgesicGame5! In the process I've been learning all sorts of things about action games and design.

Now, with Shuffled World, what I was shying away from was the task of implementing some kind of core that could carry the granular stuff you do in a room. Sure, there were perfunctory solutions - shallow combat or repetitive puzzles - but those didn't seem satisfactory. There's a lot of interesting 'design containers' to the exploration - things that could contain and lead to cool design, but without a strong 'core', I think I never felt motivated to do much more than the simple exploration prototype a few months back.

But as with all game ideas, sometimes time away really helps you figure out what matters the most and what doesn't. And after considering the work I've been doing on Analgesic Game 5, I think I have a decent idea for how my Shuffled World idea could work! I've created two prototypable game designs.

The first, I'm more excited about:

 1. Shuffled World x Analgesic Game 5 (AG5)

I feel pretty confident about the level design and combat in AG5. It can create both simple and relaxing levels, as well as extremely intense, demanding, yet fair levels. Situations can range from puzzly to vanilla action to strange one-offs. Rooms of enemies can be designed to be easier with certain weapons vs. others. 

Designing so many levels for AG5 has led me to wonder what the combat core would feel like in other areas or contexts. One of those has been roguelikes. We probably would make more money making one of those. And hey, I'm not really against the idea - I think AG5's core is great. Of course, I don't think I would want to do something typical. Like I thought over two years ago at the conception of this project, I mainly like the mystery and secrets behind roguelikes. That's always what I've been after with this project and its design/narrative goals.

In previous devlogs, I've lamented over how it would be hard to marry the mapping puzzles with exploration. There's something that doesn't sit right with a overly-random, perfunctory 'exploration' aspect to finding the treasure at the end of the map. Narratively that feels a bit flat - you have a 'a-ha!' moment solving the map puzzle, and then you spend 5 minutes walking boringly through small rooms until you can re-create that map.

So here's what I decided...

The Combat Core

• The exploration still takes places in rooms with up to 4 cardinal direction exits.
• Each room has enemies in it. I design the rooms. I can additionally define parameters for how the room can get 'harder' (maybe not all enemies spawn, maybe their HP or attack scale, their movement speed change, etc.)
• Thanks to the richness of AG5's combat design, it's very easy and possible to make themed 'roomsets'. E.g. "River rooms" which contain river enemies or something.

Making Maps

• To find a Treasure, you have to create the Map! A Map is a blueprint of one possible route to the treasure.
• That is, even after making a map, it's NOT the version of the Shuffled World you're exploring.
• Often Treasures require the map to have rooms from one or two roomsets in it. For example, the ruby might only appear if you travel through three grass rooms, then three fire rooms, in a connected path of 6 rooms.
  • But remember, a map having a Grass room on it will give you a RANDOM grass room when you actually try to explore the Shuffled World!
• Valid routes might be fuzzy, meaning that even if you don't execute the map perfectly, there's still a chance you could find the treasure. This fuzziness is not explicitly defined to the player, rather it's more of a failsafe
• To make a map, you place colored squares (corresponding to rooms from roomsets) on a grid. The game shows you if your map is valid as you edit it.
• You get to see Clue Maps, too - failed examples of Maps to a treasure, that show you roughly how 'right' they are, in terms of their Journey and Neighborhood
  • A map is VALID  if it has a a valid Journey and Neighborhood. The Journey is the path to the treasure - e.g. "The Ruby appears only along a path of X Y Z", and the Neighborhood is the area around the treasure "The Ruby must be within a wide lake. (e.g. Lake Rooms from the Lake Roomset)"
  • So if you made a map for the Ruby, but it has the wrong path or neighborhood, then it wouldn't work as a map.
• After creating a valid map, you can then set out on your journey into the shuffled world, setting it as your 'blueprint' which you try to build the shuffled world around. How does that building work...?

Finding Treasure

• Here's the catch! Each Treasure has a precondition to spawning (which is revealed). Maybe this is a particular weapon/item loadout, or something else, or the base difficulty.
• You can take up to 4 roomsets with you (number might change, idk.)
• You always set out from your home town at the center of the Shuffled world - (0,0). You get to pick the color (that is, the roomset) of the first room.
• Now, to start building out the world in accordance to your map - you must defeat the enemies in the room! This isn't necessarily every enemy (and this is a key concept to AG5's room design) - maybe it's only the weak ones, maybe it's just one weird one out of the crowd. Maybe there are other conditions (breaking a pillar while dodging attacks... etc) You must do this without spending a reset.
• What's a reset? It's a limited resource you get for each new adventure into a Shuffled World. It lets you retry a room. Why would you reset a room?
• If you run out of health in a room, you can't pick the color of the next door. This means maybe finding the treasure won't pan out. But thanks to fuzziness, maybe it will! Maybe you'll get the right room you want! In either case, if you run out of health in a room you'll probably want to reset. And the more roomsets you decide to bring, the more dangerous it is to spawn a new door without clearing it successfully.
  
• Health is probably restored in each room and pretty low. Well, idk. This is something you decide after making levels.
• Even after picking the color of a door, if you want, you can reset the room, to try and change them.
• By clearing enough rooms successfully, you can get to the treasure.

Now, while that would work well enough, there's still a lack of drama in reaching the treasure, especially as it relates to the map. That's where difficulty scaling can come in...

• Rooms all have 'difficulty ratings' I assign.
• The deeper into a color you are along a path, the harder the rooms from that roomset that will appear.
  • Put differently, if you're on a path that's 10 units away from Home, and spawn a Lake room - if it's your first it'll probably be one of the easier ones. But if it's the 11th, you're probably going to get a very hard one!

Here's where it becomes important for AG5's design to be solid. In many games the combat doesn't really get harder, it just becomes more ridiculous or the numbers get higher so you need better equipment. Now, I'm not 100% against that kind of scaling, but I think it works best when the combat system itself allows for no-damage, high-level play. Something like the combat in AG5 can get very hard, but it's theoretically possible to perform perfectly and never get hit.

It's also important for there to be lots of rooms, that still feel unique. AG5 is set up for that kind of quick, quantity-heavy design that still feels unique.

I think this can create some drama in the mapping section of Shuffled World. You might be piecing together a map, and notice that the treasure has to lay at the end of a 10-deep path of Lakes. Maybe you've encountered a tricky room from the Lakes before... how hard would it be if it had another enemy? If they moved slightly faster? I want that kind of knowledge to feed into what you might bring with you to even the odds, to create a kind of mild preparation step.

Now what if you switch from one color of room to another? Well, generally probably nothing. But I could see devious Treasure Maps where switching color requires you to give up the weapons you used in the previous room.. and so on. Maybe you're forced into status conditions or stat debuffs. Maybe a terrifying map tends to have things that apply time conditions to successful clearing.. or require a certain weapon.

The overall point is that I think the solid combat core within the exploration allows for the mapping puzzles to also feel strategic, maybe a bit terrifying. What if a treasure forced you to get through the run without using a single reset? That's kind of frightening, even as I type here. There seems to be lots of room for rich variation upon this "Map" formula.

Narrative Possibility

What's even more exciting to me is the narrative possibility here. AG5's level design is already kind of 'expressive' in this abstract, narrative way... the way I can place enemies, modify their behavior through code, vary the spatial layouts, feels really rich to me, even without NPCs or very many words. 

What I think this "Shuffled World" version of AG5 does is that it allows the dungeon (or the Map) to become a flexible, expressive entity of its own. Single rooms aren't locked to a single level like in AG5, but might reappear mixed into other Maps. There's the expressiveness of being able to hide secret rooms in the Shuffled World, to hide entire towns. I can let the player save and reload maps, and set out from somewhere within a particular Shuffled World. Or - what if stumbling upon a random room PREVENTS you from saving the map? Maybe you'll never meet a certain character again...

We can create a sense of depth and distance - what if some rooms that are very very far from Home end up staticy and weird? Weird things could happen and interrupt exploring a Map with a certain buddy NPC. What if certain characters can 'attract' certain rooms into the Shuffled World?

 What if you're looking for a Treasure and somehow have created a path that is valid for two Treasures? What if this is narratively relevant? Etc, and so on. I can create all kinds of twists - maybe I force the world to be 'unshuffled' and let you into a pre-designed dungeon. 

It should be easy to shift registers between that Yume Nikki/LSD-dream-emulator mysteriousness, to the more familiar worldbuilding of games like AG5 or Zeldas!

The Roguelike Influence

I think this captures what I like about roguelikes - that slight sense of chance or improvisation, the mystery, but while preserving the kind of designer-led storytelling I personally like, and while keeping an interesting sense of level design both at the room level and the more meta dungeon level.

Now, what about progression mechanics? I think after working on AG5 a lot I'm not a huge fan of them, usually they water down the core or something. I think with Shuffled World, if I DID have EXP/leveling, they would only be in the form of one-time EXP rewards upon finding Treasures. (That's how we've decided to do it for AG5). That way there's not really grinding - leveling is mostly a way to slowly introduce complexity to the player by giving more options. Having trouble clearing a treasure? Maybe finding other ones could get you an item to even the odds.

In Conclusion

That's my general idea for how Shuffled World could work. I actually really like the idea now, especially because I feel like I  have concrete evidence it would be a lot of fun thanks to having that 'tested' with the level design I'm doing in AG5. Sure, there will be things to figure out along the way while making (well, if I ever make) Shuffled World, but I feel a lot more confident now that I have this 'core toolset' of the combat and dungeon stuff.

But, oh yeah, what was that second idea?

2. Shuffled World x Yume Nikki

This idea is like the previous idea, but it's not as intense about having to be so dramatically-driven. You'll still solve puzzles and maps, but exploring the shuffled world is a lot more straightforward, maybe even perfunctory. It's mostly about building out the Shuffled World by opening doors however you want, and maybe these relate to little environmental puzzles (following a river), and you reach the destination and it's a little 3D zone to explore. There could be light gating/metroidvania elements, like needing climbing boots here or there... this would be a more loose, relaxing experience. Maybe you could throw in some light incremental mechanics to prevent you from making an infinitely  long map, etc. 

I'm not as interested in this version of the idea at the moment, but it does have its appeal (it would be a lot faster to make, since we don't have to make enemies or bosses). Maybe it could appear in the regular shuffled world as a variation! Who knows! Someone steal it...

Well, see you all next time, whenever that is.