Paizo Top Nav Branding
  • Hello, Guest! |
  • Sign In |
  • My Account |
  • Shopping Cart |
  • Help/FAQ
About Paizo Messageboards News Paizo Blog Help/FAQ
Pathfinder Roleplaying Game
Pathfinder Society

Pathfinder Beginner Box

Pathfinder Adventure Card Game

Pathfinder Comics

Pathfinder Legends

Dungeons and Quests - Built or Random?


Pathfinder Online


I have only just discovered this project but it has peaked my interest like no other recent MMO in development. Just recently I was talking to my GF about my dream MMO. Pathfinder Online has hit all of those notes. After absorbing the blogs and online info I want to open a discussion on dungeons.

In my opinion the best way for Pathfinder online to have discover-able dungeons is to use randomized dungeons. Instead of putting huge amount of effort into making individual dungeons how about making just the rules for the game to make them itself?

The specifics are less important, but with proper testing it should be possible to have everything a dungeon needs: Traps, secret doors, simple puzzles, monsters, etc. Obviously the dungeons skins and monsters will depend on the region it spawns.

Once the system for generating dungeons is made adding templates or pieces would be much easier. This allows dungeons to change over time as content is developed.

As an aside, I have daydreamed of a system where the quests themselves are also random, but smart. Much like how a dungeon master may roll on a chart to help her with a complication in the campaign. For example:

1. quest to get x from a dungeon
2. complication, another person wants x for a good reason
3. choice

simple, NPC's could be assigned rolls (and depending on the NPC perhaps the choice has an affect on you standings with each NPC's represented faction, or your alignment) on the fly (like in Skyrim), and the simple complication adds depth where none existed before. Writers write scrips that are reused but are modular enough that the average player wont see the ruse (or at least will be able to have some suspension of disbelief.)

Quests like this can also be expanded infinity. Now that your choice NPC has x another quest opens. This new quest never existed before. Instead its generated by the script. New NPC's can be introduced. The script would chose them based on their role within the game (priest, bartender, part of a certain organization, etc). New complications are also generated.

If we want to get really crazy perhaps the NPC who wanted x also got y and z from other players (just by chance). At some point the generator rolls that the players who got got x, y, and z are invited to go on a joint mission! That way player actions affect each-others outcomes for their quests!

Basically what I am saying is invent a game that has infinite story, all created through player actions.

I would be interested to hear what others have to say on the subject.


It sounds like a crazy good concept, but is it manageable? Sounds similar almost to neverwinter nights, and a lot of what it was trying to achieve, by creating dungeon editting software for Dungeon masters. Either way sounds like almost any Dnd/ RPG gamers dream. :)


1 person marked this as a favorite.

Hi there Samping, welcome!

This is all great thinking, and a lot of it I'm fairly sure is already incorporated into the design principles.

Regarding dungeons, I agree completely: Modularly designed, procedurally generated dungeons are the way to go! After setting up a framework, they provide much more content for less work (albeit at the cost of design quality, less so with better framework/pieces).

However, it seems like you're thinking a little more in 'theme park' mentality as far as the quests go. How it looks like PFO intends to work is much like how you described, but you cut out the NPC and a lot of the scriptwriting. Here's the kind of thing that I expect to happen:

Lets say Chartered Company (player organization) Money-Grubbing Merchants has found a potential profit in the Wolf Fang Necklace market over in Westville. There aren't any wolves near Westville (which is why the price is so high there!). So they send Merchie to Eastville, which is right by the Wolfy Woods. Merchie sets up some quests for the locals: Bring wolf fangs, and make wolf fang necklaces. In the meantime, some unfavorable characters notice this, and set up to ambush Merchie when he leaves town. Knowing this likelihood, Merchie hires Somedude to move the goods, and Somedude hires some guards. One of Somedude's guards turns out to be a bandit, and arranges for the party to get jumped and the goods to be stolen. Then Merchie talks to his CC to hire a hit squad to get the goods back and take revenge. This potentially turns into a war between a bandit organization and the Money-Grubbing Merchants.

Thats a lot of player-generated content that didn't require a single line of developer script-writing. Now you could add the occasional big-picture quest, which would require (and generate!) interaction with a number of other players, and that way turn one developer-made-quest into dozens of quest-like missions!

Edit: And this is why sandboxes with simulated economies are really cool!

Goblin Squad Member

Hi Sampling, interesting ideas on dungeons. As Waffleyone says:

Waffleyone wrote:
Regarding dungeons, I agree completely: Modularly designed, procedurally generated dungeons are the way to go! After setting up a framework, they provide much more content for less work (albeit at the cost of design quality, less so with better framework/pieces).

Additionally there should be/are some mega dungeons eg Thornkeep that ARE well crafted (not sure they should have randomization of mobs/traps also?). But in terms of "discoverability" of dungeons out in the map, definitely procedural generated/randomized dungeons with variable challenge and variable resources can work better as a staple dungeon.

Again with Waffley on this, contracts seem like an infinitely more plastic way for players to do meaningful missions in PfO. Thornkeep mentions regular quests, as well, so there could be scope for interesting ways to combine these more than the bog-standard fedex-10-rats type?? Need to think about that. Tbh after reading about contracts, not sure I ever want to do another quest again. :)

Goblin Squad Member

Yes, nice idea, but one caveat:

This will be a game where control of resources is very important.

If players can add content, this should not be hugely profitable but rather some sort of fluff or else people would create their own content to exclusively farm resources from.

In this light I think that, while probably cool, this feature will be in the "nice to have" folder for quite some time.


I think some of you misunderstand. Players would have no control over the content beyond choices made during a quest. The dungeons are created in the background along with the quests. Those dungeons can be discovered by players through exploration as the blog stated. Thus, to answer MicMan, there is still a scarce resource as the person who first finds the dungeon keeps it.

Again, i am not talking about player made dungeons or quests

In my quest example the choices created in the background and presented to the players. I agree that a quest system like this is far more unlikely (although Skyrim did a good first shot at it) so perhaps its best if we focus on dungeons.

Also, to Waffleyone, I come from an EVE online background so trust me when I say I am not trying to push in a themepark direction. In fact dungeons that are you cannot predict are far more sandbox than setpieces that can be memorized or looked up online (like EVE online or WoW dungeons).

I think of dungeons and quests as being part of a larger economy. Some things can only be found in dungeons for example. Dungeon delves are kind of like different resource gatherers and just as important as miners. I imagine dungeons being like discover-able sites in EVE online where you can find bad guys, good loot, and rare blueprints. Of course in pathfinder the rewards would be different but the concept is similar.

Randomly generated dungeons would also add to player interaction because the threats cannot be guessed at (or looked up online). This lowers the chance of people farming easy/exploitable dungeons solo (looking at you Dire Maul from WoW) (although many dungeons should probably be able to do solo in my opinion)

I agree with Waffle that player interaction should be a main driver of content. I have many stories of EVE online that attest to how his can happen. On the other hand players need something to do other fighting each-other or mining/crafting. Dungeons are a great to build unit cohesion and are fun!

Grand Lodge Goblin Squad Member

I think both random dungeons and player made dungeons could be utilized, both though with different core assumptions.

Random dungeons should be the ones which can drive the economy on a bigger scale. ie Group finds random dungeon, clears it and then is able to farm the resources within... whereas

Player made dungeons should not have placeable loot (and depending on difficulty - random loot can be assigned), but the player made dungeons could be more for storyline and flavour. Also, player made dungeons could be converted into a 'random spawn dungeon'. ie Group finds a random NPc who asks for a favour (quest). Group heads out to complete quest and finish story.

Maybe all quests and dungeons provide some sort of 'faction' gain to help you deal with certain NPCs (a la Everquest) or give you more 'Renown'.


I imagine that quests for NPC's would have some other rewards so players who feel loyalty to NPC groups can work for them exclusively.

Goblin Squad Member

Sampling, had another, more deliberate browse of your idea about quests and think you're onto something.

So quests might be available for a time and ""anyone"" might be able to decide on it, with perhaps a notification of when people signed on for it. That way you get something of a race; which would be aided if it's multi-staged and if it requires several players then (decide how to split the loot intra-group and/or inter-groups!). Also like the idea of different quests influencing different factions/alignments. So you can choose to target alignments for/against or avoid such divisive quests and go for middle-of-the-road ones and so forth. Those are cool to add a loop to quests to the wider world. And the former adds collaboration/competition among players!

Also for interest, The Repopulation: Missions has Quest Generation System that generates "smart quests" for players to build on what they're doing/done.


Thats very interesting. It seems that the people working on Repopulation have similar ideas to me.

Paizo / Messageboards / Paizo Publishing / Paizo Licensed Products / Pathfinder Online / Dungeons and Quests - Built or Random? All Messageboards

Want to post a reply? Sign in.

©2002–2014 Paizo Inc.®. Need help? Email customer.service@paizo.com or call 425-250-0800 during our business hours: Monday–Friday, 10 AM–5 PM Pacific Time. View our privacy policy. Paizo Inc., Paizo, the Paizo golem logo, Pathfinder, the Pathfinder logo, Pathfinder Society, GameMastery, and Planet Stories are registered trademarks of Paizo Inc., and Pathfinder Roleplaying Game, Pathfinder Campaign Setting, Pathfinder Adventure Path, Pathfinder Adventure Card Game, Pathfinder Player Companion, Pathfinder Modules, Pathfinder Tales, Pathfinder Battles, Pathfinder Online, PaizoCon, RPG Superstar, The Golem's Got It, Titanic Games, the Titanic logo, and the Planet Stories planet logo are trademarks of Paizo Inc. Dungeons & Dragons, Dragon, Dungeon, and Polyhedron are registered trademarks of Wizards of the Coast, Inc., a subsidiary of Hasbro, Inc., and have been used by Paizo Inc. under license. Most product names are trademarks owned or used under license by the companies that publish those products; use of such names without mention of trademark status should not be construed as a challenge to such status.