I’ve been neglecting the blog while working on game development, and I’m not really okay with that. So I’m posting on an other-than-usual topic, one that’s on my mind, rather than not post at all.
For the past two years, I’ve been working on-and-off on a game called Hecatomb. It’s a different kind of “zombie survival” game, inspired by Dwarf Fortress. You play a necromancer raising zombie servants and building a base in the wilderness. It’s a “sandbox” game – like meaning there is no singular goal, other than surviving and doing cool things.
Right now my game works, but isn’t very fun because there simply isn’t that much stuff to do. That, I think, is the secret to games that generate stories rather than having preset stories – they need to generate interest by having a wider-than-usual number of “subsystems” (loosely defined) that interact in potentially interesting ways. Take, for example, Don’t Starve. The way I’m thinking about it, these could all be considered subsystems:
- Basic resource-gathering and crafting.
- Night and day, light and darkness.
- Food and the crock pot.
- Seasons and weather.
- Trees, wood, charcoal, and tree guards.
- Shovels, replanting, and fertilizing.
- Maxwell’s things.
- Birds and bird cages.
- Various things found in swamps.
- Gears and chess.
- Science and prototyping.
- Choice of character.
- Hound attacks.
- Sanity management.
…and if I’m stretching, walrus camps, koalefants, and so on. In Minecraft:
- Dungeons and mob farms.
- Cave and mineshaft exploration.
- Night and light.
- Nether portals.
- Nether fortresses.
- Villages and trading.
- Swamp biome resources.
- Desert biome resources.
- Jungle biome resources.
- The End.
- Taming pets and mounts.
- Furnaces and smelting.
…and several smaller ones, like fishing and boating.
Now, compare that games that are influenced by roguelikes but have clear goals.
- Ship choices.
- Alien species.
- Stores, parts, and augmentations.
- Ship systems, upgrades, and repairs.
- Shields, lasers, ion weapons, and missiles.
- Event choices.
- Unlock quests.
- The rebel fleet.
- The rebel flagship.
- Character classes.
- Adventuring areas.
- Town upgrades.
- Quirks and diseases.
- Leveling up.
- Boss battles.
- The Darkest Dungeon.
- Picking stuff up and throwing it.
- Plot-related items.
- Shopkeepers and power-up items.
- Ropes and bombs.
- The ghost.
- Unlocking shortcuts.
- Differences between the four zones.
- Hidden zones.
At the opposite extreme, Flappy Bird has only one “sub”-system: flapping.
There seems to be a clear pattern, that games with nebulous goals need more subsystems to maintain interest. This makes the process of developing a sandbox game tricky, because before a certain point, there simply aren’t enough subsystems to make it fun (how did early Minecraft and Don’t Starve version handle this? Did they simply appeal to a more restricted fanbase?) Thus far, Hecatomb has:
- Lumber and getting attacked by dryads.
- Trading at the black market.
- Building defenses against sieges.
- Getting attacked by humans.
- Managing spell energy to raise zombies.
Even if all those subsystems were fleshed out well (they’re not) that’s not enough to make a sandbox game fun. Eventually, I plan to add:
- Ghouls and other nighttime monsters.
- Cave exploration.
- Rival necromancers.
- Special stuff if you dig really deep.
- Dens of monsters on the surface.
- Ant colonies.
- Corpse management and rot.
- Advanced resources.
- Attacks by very large monsters like dragons.
…which gets me closer to the number of subsystems that a game like Don’t Starve or Minecraft has.