Random Dungeon Sagas

DS tilesAs you may have seen on the (huge) Kickstarter update 40, random dungeons are in, and they look like being a lot of fun. The random element can really capture the whole exploring the unknown feeling, with the potential for nasty surprises to be lurking round every corner. It’s also bit more anarchic and less finely balanced – so the dangers may be even greater than normal. There’s no way of knowing. Not a game for those of a nervous disposition.

In terms of rules design, the enormous number of different game modes, variable characters, random dungeons and so on offer a particular challenge. This is how do all these disparate elements fit together? 

In the good old days when roleplaying games were in their infancy, there was D&D. This was a pretty simple game. In contrast to this, there was Chivalry & Sorcery (C&S). This was not a simple game. However, it wasn’t the intricacy of individual rules that made C&S so complicated, it was the lack of integration. For example, in their book on magic they listed many different types of prestidigitation, and each one worked differently. This was great for adding character to the different spell casters, but required a huge feat of memory to get it all right if there were several types within the adventure. Whilst C&S always had its devotees, and while I rather wanted to play it too, the major impact it had on me was as a lesson in how not to do something. 

I’m sure some of you will tell me I’m being unfair, and that may be true. This is all said with the fuzziness of decades between me and it, and it’s a form of nostalgia. Like all nostalgia, close inspection may make it disappear. Either way, it’s a lesson I apply to all of my designs, and with Dungeon Sagas, because of its similar theme, I feel it even more strongly than usual. 

Heroes 1As all these new elements have been added to the Kickstarter (thanks to the generosity of all you guys!), I’ve been pondering how to integrate them all into one smooth system. Clearly some bits fit more closely than others. For example, creating your own characters and resolving experience gains for characters are a good fit and will be closely related. Some parts will simply be the same. Any experience system bleeds into what happens in the downtime between adventures, so that’s linked too. 

Building dungeons is a different subject. The random layout of dungeons is more than just placing rooms – it’s got to populate them too. The AI systems need to integrate with the ways that dungeons are built if I want to be able to include more character for the creatures that inhabit the dungeon (which I do). You can’t tailor AI to fit the behaviour of specific beasties if you don’t know what they are in advance, and random doesn’t suggest that you would. Or can you? I think I can see a way around this too. The AI and the dungeon generation (both planned and random) need to link up and impact on one another, so they will.

I’m thinking of using a split deck with one half being the layout of tiles and the other part being the creatures in it. Both halves would include a core set of cards that was always used (base tiles from the Core set on the one hand, giant spiders, rats, possibly traps, etc on the other). Each new tile we add in a quest pack would have a new card to go with it to add into the mix. Quest packs would have their specific race set of cards to add to the base set. The Advanced game that introduced these rules would have the card set for the Undead. 

This means that you will be able to tailor things to an extent, so you’re randomly drawing from within a specific group rather than all the possible critters (or tiles or traps or whatever). This makes it possible to tailor the AI deck in the same way and have them dovetail properly 🙂

Undead 1In practice, I’m thinking that you will draw a card to lay out one or more tiles (depending on how big they are), then draw a denizen card to see if anyone’s home. Finally, an AI card will tell you what they’re up to and how they react to their peace and quiet being disturbed by a bunch of rowdy adventurers. The doors will separate different sections of the dungeon as they do normally (almost as if it were planned). 

Finally, while I’ve been thinking of this, I realised that it would be good to be able to tailor a random dungeon to fit any group of heroes, with the cards responding to their level of experience as they go along. In normal Necro-built or pre-set scenarios you have to make the Heroes fit the level of difficulty of the dungeon. Random generation flips that on its head as there is no fixed dungeon to use as reference. Instead the Heroes themselves become the reference point and the cards can adapt to give the Heroes a good run for their money whatever their level. 

As this is all still in development these rules aren’t final. I just thought you might like a suggestion of the sort of things I’m aiming to include and the kind of way these can all be integrated without making the rules hugely complex. Once this has all been worked through in detail and played a few hundred times I think we’l be able to make it rather slick and incorporate a great deal of character in relatively simple processes. Which is what we want 🙂

This entry was posted in Board Gaming, Dungeon Sagas, Game Design Theory, Kickstarter. Bookmark the permalink.

39 Responses to Random Dungeon Sagas

  1. Guillermo J. Colón says:

    I think it’s a good idea, my mind turns immediately to Dork Tower’s monster generation technique.

  2. PikaRapH says:

    That sounds really good ! Can’t wait to try it ^^

  3. jason says:

    As far as tailoring the encounter cards to match the heroes level, maybe the card could be split into sections. If the hereos are lvl 1-3, X occurs. Lvl 4-6, Y occurs…etc

  4. eriochrome says:

    Seems like it might be hard to separate the monsters from the AI into two different sets. Might be able to have the AI deck more generic like aggressive (attack nearest enemy and chase after), neutral (attack nearest but no chasing), defensive (hold passage positions and wait for heroes to make a move) , sneaky (attack lowest armor or lowest health), flee, etc. Then have those keywords map to finite number of choicse on the monster card. A mindless type monster might have everything map to aggessive behaviour while goblins would probably have defensive, sneaky and flee actions. You could replace the AI deck with just a dice roll but the AI deck could be tuned for the type of scenario by varying the ratios of the cards.

    • Mi7 says:

      All sounds awesome just put a few ‘reshuffle this deck’ cards into the monster cards and ai cards to keep the heroes off balance i reckon

      Will the ai card cover one tactic for all the monster in a room of detail different tactics for each monster type?

      • eriochrome says:

        I would probably set up the dungeon deck to have some cards which link to full encounters (cards which map to special tiles with villians and maybe a few other monsters on them) and then most cards that have a map section and a guide for how many monster cards to draw based on the adventurer advancement state. In the system I describe, since one AI card maps to different states for different monsters I would probably just draw one card for all the monsters in that zone. That way the monsters all act according to their natures and you do not need to remember that goblin 1 is defense right now while orc 2 is sneaky, etc.

    • Quirkworthy says:

      @erio – that sounds similar to what I’ve been working on. Drawing only a single card to define all the actions of the creatures present. It gets way too fiddly to have one per beastie, or even one per group.

  5. PaulM says:

    I like your current approach to spawning the dungeon… only placing the visible sections and then place the next section when you open the door is great. This is how I’ve always played Space Hulk solitaire and it really makes it feel like you are exploring the unmapped depths.

    I’ve always been a fan of your spawning concept in Lost Patrol. The way the jungle and enemies would get deleted and then rebuilt if you push forward and then backtrack meant that you always had interesting choices for maneuverability. I don’t know if this dungeon is meant to be that dynamic but it can be frustrating to have no alternates pathing options in a linear dungeon. The Lost Patrol approach could be interesting if you want to create the feel of a running battle rather than just bashing your way past linear choke points.

    • eriochrome says:

      I could total imagine a special magical dungeon type were any tiles you could not see disappear but I think that would probably be a special case. I would think that the generator cards would probably set up[ between zero and 3 exits from the section with some rule about how to treat new and old tiles that are close like finding the back side of secret doors.

    • Quirkworthy says:

      @PaulM – thanks. I’m still very happy with Lost Patrol and it is an interesting layout. Not really suitable for most dungeons add the walls aren’t likely to shift. That doesn’t mean it couldn’t ever be used though…

  6. crusher says:

    Great ideas and youmshould also think on Unexpected events or initiative/ambus situations whe heroes can be surprised or caught by surprise the minions.

    Wanted also to ask about the idea of linking player objectives that are kept hidden from a common mission.

    • Quirkworthy says:

      Good question. Linking those hidden missions into a random dungeon might be rather tricky. On the other hand, perhaps a more effective approach is not worrying about a link. In the Advanced game there are so many variables that it gets to be more about telling the story, and maybe the side missions is just part of all that.

  7. Maciek says:

    Clever… 🙂 can’t wait !

  8. Danny says:

    I can’t express how happy I am that the random Dungeons are being included. It’s just such a fantastic way to experience the exploration element of these types of games. Any effort required to pull it off will be effort extremely well spent in my opinion.

    So, from what you’ve written, I’m seeing;

    1) A desire to create random, yet themed dungeons. These randomly card-drawn dungeon sections to be placed as the heroes explore deeper into the dungeon with each opening of a door.

    2) A desire to create random, yet themed AI dungeon inhabitants (themed both for their race and for the particular dungeon setting). Each dungeon section group of inhabitants to be determined by a card draw after a new dungeon section placement.

    3) A desire to include escalating AI difficulty to cater for hero advancement and/or hero type.

    4) A desire to allow players to tailor their adventure using the above systems.

    Excellent stuff!!! I **love** what I’m seeing here and I think you’re definitely on the right track in the way you propose to implement it.

    Allowing the decks to be tailored to suit the players tastes is fantastic, and something that the WHQ random system was sorely lacking in an official non-house ruled way. Having random dungeon inhabitants is cool, but its much, much cooler if those dungeon inhabitants could plausibly reside in the same dungeon in an at least semi-cooperative state. For instance, I could NOT see a small dungeon with Orcs, Zombies and Demons all living happily together. Yet I CAN certainly see a small dungeon containing Orcs, Goblins, Mawbeasts, a Troll or two and some Rats and Spiders, all residing in a somewhat stable state.

    So allowing the Monster deck to be stacked/tailored is great.

    Allowing the Dungeon Room deck to be stacked/tailored is great for the same reasons, plus an additional reason. That additional reason is, it allows the players to choose/control an **approximate** game time for completion.

    An hour long game = 3-4 x dungeon room cards
    A 2 hour long game = 6-8 dungeon room cards

    Something along those lines, just as an example.

    What is the reason for adventuring in this random dungeon? Will there be an objective room (a la WHQ), or a final goal to be achieved to **win** the game? How will the heroes win the game, how will this be determined?

    In such a randomly generated dungeon, how will treasure and equipment finds be handle? A draw from another deck after “clearing” a new dungeon section? A roll on a table? Will the AI cards incorporate XP and treasure values to specific monsters? Or will the rulebook do this?

    In regards to the AI, I still think that having a set of rules in the actual rulebook, which will dictate how certain troop types will act by default when using AI, is a good way to go. This can then be manipulated by using the AI cards, changing slightly the behavior of these troops by race with the inclusion of abilities or proposed slight behavioral changes. The card could also include monster “levels” so they would scale appropriately with hero levels. It could also include specific behavior towards certain hero types…although this may also be better left in the rulebook under the default troop type AI.

    In regards to the laying of random dungeon sections. One, not insurmountable problem, I do see though, is this;

    How do you determine how and in what way each new dungeon section drawn, will be placed against a previously placed dungeon tile? In WHQ this was fairly straight forward as the shapes of the rooms, corridors and intersections were very basic and limited, making it obvious in how they should be placed. However the DKH and Dungeon Saga tiles are great in number and variety of shape and are able to be placed together along any edge. There would need to be a strict guideline in place for the placement or joining of one new dungeon section to a previously placed dungeon section. One way to tackle this that quickly jumps to mind, is to have thin (red or yellow or whatever colour) lines printed onto specific edges of the dungeon tiles so that when the tiles are being use in the random dungeon generating game mode, it’s explicitly obvious where and how tiles may to be joined. Or wold you just simply leave it up to the players to choose how the new dungeon sections would be placed?

    Single Sections or Multiple Sections for each card draw – I’m still undecided whether or not it would be better to have cards that lay down single tiles or multiple tiles. I can see arguments for both, but I tend to feel that it would probably be better to lay down a set pattern of two or three tiles with each card draw…..HMmmmm….don’t know, not sure and undecided…

    Traps – Will trap cards be mixed in with the deck that determines the dungeon inhabitants of each new dungeon area? Each trap having its own AI

    Doors – I’m still a fan of having a separate door deck, to be drawn on when a door opens. Its very quick to resolve, adds an extra element of excitement and increases the exploration theme/atmosphere.

    As a last point of interest. I see this randomly generated dungeon most used when players are wanting to either play solo, or play Co-Op. However, it would be interesting to see if the same system could be used with a human Necromancer (or bad guy) player Vs human hero players, whereby the bad guy player is forced to use whatever lady fate delivers him to defeat the heroes with…nothing I am suggesting that should be included in the game, just a point of interest as another game mode.

    Loving this stuff Jake, please do continue to share with us as much and as openly as you have been, it very much appreciated 🙂

    • Quirkworthy says:

      Good point on game length. It’s desirable to be able to control it. Unfortunately that makes balancing the threat of the dungeon vs the Heroes more difficult. Not impossible, just more awkward. Maybe a way round that would be to add to the actual power of the Heroes for shorter dungeons, to give the dungeon a more powerful response. That might work.

      Joining tile sections is something that will need a lot of messing about with diagrams and layouts to get right. There’s a few day’s work in that alone. Rather than print something on the tiles themselves, I’d mark the joining areas on the cards. They will need pictures of the layouts anyway, and not all will want to join the same way. As you say, it needs to be obvious.

      One tile or three per card? I think a mixture will be best. Also, not all should have a door at the end, so you may draw more than one card at a time, laying the tiles out as you do so.

      I can think of a couple of ways to do random traps and will test both to see which works best in practice. Possibly mark them on the layout and then randomise their effect. You can then have multiple cards with the same tile layout, but different traps, to mix things up.

      Doors will be a set of counters with their different types on the back. That allows them to be used both for planned dungeons and randomly, and to sit on the board in both cases. If we get 3D doors then they will just sit on top of the counters.

      I’d considered a random dungeon with a Necro player, and I think it should work. You’d ignore the Ai bit, but the rest would be the same. The only missing thing would be a victory condition for the Necro, though crippling one of the Heroes is a good fall back.

      Thanks for your comments. They’re appreciated too 🙂

      • eriochrome says:

        Maybe for the Random dungeon with an AI you would generate the dungeon in larger sections at a time so that the Necromancer can try to position the monsters the way he wants. You can use door to seperate the sections that monsters should not move through.

  9. eriochrome says:

    Ofcourse the random dungeon part is where the D&D Coop games fall down since they use it for pretty much every adventure. The square interlocking tiles they use for it work great but the problem is more that you reveal a tile at the end of your turn and they all have a new monster on it that automatically attacks you, plus about half have an encounter card that also usually does damage to you. It is pretty much a health tax on each tile. You also find that no matter which way you go you will find the objective tile since it is placed in some range in the stack.

    If you decide on drawing monster cards will people get extra cards in the monster booster boxes so they have a matching number of cards to number of monsters.

  10. LCK says:

    Yeah, the D&D games actually have a good basic system but the surprise is gone. Every tile is 99% guaranteed a monster, plus 50% encounter card if it’s got a black arrow. Need to add Event cards to counterbalance that, or houserule to say onl draw encounter cards when a black arrow tile is drawn and ignore the rest. While Descent 2nd Ed. lays out all the tiles and monsters already. No surprise. Needs to be a good balance between the 2 and I hope DS is it.

  11. crusher says:

    Jake I created sometime ago a deck system to create random dungeons for a set in 10mm. Perhaps you could check it and get some ideas… 😉

    You can get it here:


  12. crusher says:

    One more idea that I don’t know if you got it from another post I did.

    One of the tjings that use not to be represented in dungeon crawlers is the time spent and the consequences of it to the heroes. The idea I wanted to drop here is about manging food and supplid to keep alive the heroes during their quests.

    The idea is that after a certain number of turns the Heroes will start “paying” with food to recreate heroes energy spent every x turns. So this way Heores are forced to run through the dungeon, loot for gold and gems but also they need to search for food items or water pools ( they could be cursed ;)). In your town for rest they could trade gold to buy supplies for the next quest of course.

    A hero who do not manage well the supplies could die to starvation in a quest (too much time spent). This way you can also balance the turns in the dungeon vs the supplies available. In other words not only gold or assets but supplies as a new resource to manage.

    Jake what do you think?

    • Torkel says:

      I know you didn’t ask me, but I’ll chime in.
      I like the “unpleasant to hang out in dungeon for too long” aspect, but I’m not a fan of extra resource management from food etc. I think that giving the heroes a sense of urgency can be done in a more interesting way.
      But having to look for water pools and such, I do like. That adds to the adventuring. It makes it more exciting to explore new rooms. One idea can be to have certain monsters affect the heroes with a certain condition, like a curse as you said, that they can only rid themselves of by finding this or that in the dungeon. (Or if they don’t find it by the end of the dungeon, they can visit somewhere during downtime to cure themselves.)

    • Quirkworthy says:

      I think that for most people, the idea of losing their Hero to starvation isn’t very heroic. They want to go down fighting the Big Bad (if they die at all). Adding tension and time pressure to the Heroes is a fine idea, it’s just that food and water is a bit mundane for most people.

      Many years ago I used to play a lot of RPGs, often running 2-3 sessions a week as a GM. In that context I did (and still would) make people think about supplies precisely because it is mundane and helps bring the characters to life. However, in a board game I think that we’re already removed from the Heroes and they need to be heroic. The different format requires a different approach.

      Of course, we can add a spin to this idea to spice it up. Perhaps a trap poisons a Hero who must then look for the antidote. This could be in a magic pool or a mushroom growing somewhere in the dungeon, creating the same effect.

  13. Jeppe Hatting says:

    Slightly unrelated, but I’m very curious about traps and what your plans are for them.
    How are they triggered? Can the players search for them like in HQ? Are they visible and needs a dice check in order to disarm?

  14. crusher says:

    Fine with me Jake you will create a very good game as usual. 🙂

  15. mastertugunegb says:

    I was just thinking… How hard would it be to combine the elements together on the same cards? The AI deck and the Random Dungeon Generation mechanics I mean. Your Dreadball Cards for example have a fair amount of variety for them from the random team mate numbers down the side, to the Cheer numbers down below and the Refbot move number up the top, not to mention card type and effects.

    You could probably have one deck that does all of these features (though the Necromancer Deck should still be separate probably) with maybe some symbols for monster behaviour and a pic of the tile to be deployed either down the bottom or as a background silhouette.

    The Necromancer Info Card could then have a quick reference of what the behaviour symbols mean.

    • Quirkworthy says:

      I’m not sure yet how hard this will be, which is why I’ve not suggested it. My guess is that it will be clearer with the cards separate, though I personally like cards that do several things. Needs more experimenting.

  16. eriochrome says:

    Given tiles from now 4 sets are the tiles going to have stuff on them that matters. I was reading through green menace and while you have the cave ins, the stuff on the tiles all seems to be cosmetic.

    • Quirkworthy says:

      The stuff on them is cosmetic, merely setting the scene. However, their shapes are not. I expect the expansions to have new tile shapes, which can make a big difference to play.

      • eriochrome says:

        But if you have no cosmetic stuff on them like chasms, magma pools, streams, ice sheets opens lots of new possible rules and character abilities. Might be a good way to really make that new tile set special and the expansion tiles different also.

        • Quirkworthy says:

          We’ll have to see. We’d need to be really clear if we do have tile art with game effects, though that’s not insurmountable. It does mean that people start looking at the art though, with a view to what it does in game, and this may in itself be confusing.

        • eriochrome says:

          It certainly requires more care in the tile art design but given that the tiles are a non negilible cost in the sets it would great to get as much game out of them as possible and not just have them be pretty movement borders.

  17. Danny says:

    If it was going to be done, perhaps it would be better to have specific counters to place on the tiles to show these effects. That way the tile can be used with or without the tile effect, and each game could have the counters placed in a different area of different tiles. For example, a chasm could be a separate token 1 space wide by 6 spaces long. This could be used to create a chasm on any row or column of a 6 x 6 tile, That same tile could be used without the chasm in other games too. It would make it more modular I suppose.

    • Quirkworthy says:

      Overlays have their own issues (don’t sit still, look odd having a hole being something that sits on top, etc). There’s nothing wrong with the idea of chasms and suchlike, it’s just how best to implement it.

  18. Paul says:

    Something that got asked on kickstarter, but hasn’t as far as I’m aware, been answered… will all of the expansions, especially the two tile packs, be coming with extra cards for the tile deck so we can use them in random dungeons?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s