Hi

Im doing a fantasy turnbased tactical RPG, with some elements from "Darkest Dungeon", "Battlebrothers" and tabletop games like "Hero Quest". You control up to 6 heroes in your party and travel a simplified overhead map where you can harvest resources, fight roaming bandits, visit hamlets (get new recruits, visit trader, crafting etc) and attempt to "clear dungeons". The last part is the main part of the game.

Dungeons are similar to in "Darkest Dungeon". They are randomly generated according to their type (cave, crypt, abandoned keep etc) and features corridors and rooms including stuff like:

fighting groups of enemies (main gameplay element)
dealing with traps
dealing with obstacles such as locked doors

A player can abandon a dungeon (go back to the map, for example to avoid getting heroes killed) but then that dungeon will remain "explored" and the player cannot enter again to try and get the "completion loot" (which is some better loot/reward if completing the dungeon).

-----

How to "complete" a dungeon? I have some ideas but want more input. The completion-requirement can be generated along with the dungeon content itself.
- Explore (visit 80 % of the rooms, the last room will be a treasury with some nice stuff and maybe extra monsters?)
- Locked treasury (one room contain the key, another room is locked and contrain the treasury)
- Find the boss (one room contain a Boss monster, kill it to complete the dungeon)

What else could there be as "objectives"? What seems fun or suitable? And what about the ideas I have already?

About the ideas you already have, most are pretty traditional and have proven to be effective. You can see they work well in Darkest Dungeon and similar games. I would suggest you study that game to find out what are the weaker points of it, what is lacking and what could have been done better or just differently.

For example, for a rpg, Darkest Dungeon is pretty light on the story/lore. The focus is more on tactics and resource management, but they could just as easily have had a more narrative-heavy game while keeping those. You could look into that to make your dungeons different and more interesting/engaging.

If you have a mini storyline for each dungeon, it opens up a lot of possibilities, for example you can have puzzles and enigmas as obstacles which require knowledge of your world and its logic from the player instead of just a specific type of resources (like the shovel in DD), you can have story-related objectives which are infinitely more engaging and fun and potentially diverse than the ones you have now and exist in DD. Those would underline the uniqueness of each dungeon and be very immersive instead of doing the complete opposite of that ("80% exploration" is both very gamey and very bland).

The challenge is then to manage to conciliate narrative and random generation. You have multiple options there, depending on the length of your game. If the number of dungeons the average player would visit in a single playthrough is small, you could just handcraft a large enough number of dungeons with unique storylines so that the player doesn't visit the same one twice. If the number is too big, then you will need to have more generic storylines and use modularity to your advantage.

It wouldn't require much apart from additional writing, you would still use the same art assets and generate the layout of each dungeon randomly, with maybe some extra rules and exceptions associated with each storyline to make it more interesting and if you are feeling courageous, some unique assets for each dungeon, like a unique boss for example.

On a basic level, there aren't that many different design options for objectives unless you are ready to introduce new mechanics, so "coating" them with a layer of narrative to make them appear more varied is a good alternative in my opinion.

Im not sure of the value of highly randomized/procedurally generated stories. And i think that part of Darkest Dungeon (that it wasnt much story) is fine for such a game.

Maybe a larger overarching goal for the group of adventures is needed though. Just "setting out to find gold" is not so inspiring:) Maybe other than finding the resources to press on, dungeons can provide something needed to reach the "end goal", even if that end goal is to locate/unlock the "final dungeon".

The narrative part doesn't have to be randomized or procedurally generated, which would be difficult to do. You can just slap it on top of procedurally generated content like the dungeon's layouts. To a degree, Darkest Dungeon did just that with the "boss dungeons". Although they were similarly randomly generated and used the same assets as regular dungeons, they also had a mini story associated with the unique asset of the dungeon, the end boss.

Then again, you are right and this type of games doesn't need a strong narrative as much as other RPG types, as proven by DD's success, but in the end it can only make the game better, so why not do it ?

It opens up an infinity of options for dungeon objectives, since they are then story related and not just design related, which solves your problem, and it also makes your game different.

If you don't want to do that, the number of possible objectives only depends on your mechanics. With the ones you listed so far you would have :

Exploration : explore a minimum number of rooms, all rooms, a specific room, a succession of specific rooms (key-treasure), a maximum number of rooms (you can only explore X rooms and need to find the final room before you do), follow a specific path (no backtracking allowed for example, tunnels collapse), avoid specific rooms/paths.

Traps/obstacles : encounter X number of traps/obstacles, overcome X number of traps/obstacles, avoid X number of traps/obstacles, find/overcome/avoid specific traps/obstacles, overcome obstacles in a specific way (a locked door might need a key but you could force it with a lockpick if you can't find the key, and doing so could be an objective or a failure condition).

Enemies : finding/killing/avoiding a certain number of enemies or specific enemies, surviving a specific enemy (this enemy cannot be defeated and instead you need to survive the battles with it), killing enemies in a specific way or in a specific room, chasing an enemy that can move from room to room, avoiding an enemy that can move.

Party : doing any of the above tasks in a limited time or in a limited number of turns/actions or using a limited amount of resources or specific resources/items, doing so using only certain heroes or X number of heroes or a specific formation/tactic, doing so while satisfying one or many conditions related to your party, like a no-death run, or completing the objective with all your heroes at max hp, or not taking any damage, or having a certain number of heroes die, or having them be extremely low hp, doing so while not using any item or mana, etc...

I think it's very doable to do a procedurally generated world yet still have key story elements rooted in the game. You just have to be careful about where you put them and make sure that people can find them. Same with objectives.