Dungeon layout, map flow and old school game design

[T. Foster]

While I disagree with most of the WotC article, this bit struck me as pretty sound conceptual advice:

Rather than start the PCs at one edge of your graph paper, put that first entry staircase in the middle of the map. Drop them into a room that gives three or four choices right off the bat, and your players will revel in the choice. Then, build your dungeon like a bullseye, with easier encounters near the middle of the map and the tough stuff tucked away at the edges and corners. Include periodic branches, especially ones that connect within the same “ring” of the bullseye. Now the players perceive meaningful choice, and you know the PCs won’t get to the corners without the prior experience they need.

which, I would point out, pretty closely parallels some advice Gygax and Arneson gave way back in 1974 (D&D vol. 3: The Underworld & Wilderness Adventures, p. 6):

In laying out your dungeons keep in mind that downward (and upward) mobility is desirable, for players will not find a game enjoyable which confines them too much. On the other hand unusual areas and rich treasures should be relatively difficult to locate, and access must be limited. The layout of a level will affect the route most often followed by players. Observation of the most frequently used passages and explored rooms will guide the referee in preparation of successive levels, which, of course, should be progressively more dangerous and difficult.

However, I'd add the caveat that while making level that is "bullseye" shaped on a conceptual/flowchart level (a la Melan's graphs in the OP) is a good idea, it's lame to make the map literally bullseye-shaped. it's much more interesting, I think, to have broad/long "high traffic" corridors that connect various sections of the dungeon/level with few, if any, "decision points" in-between, and then complex mazes of winding corridors and rooms that "fill in" the intervening space -- so the most 'remote' section of the dungeon (in terms of decision points and intermediate encounters/'challenges required to reach it from the 'start' point) might actually be physically very close to the start point -- which can cause interesting situations to arise with "treasure finding" spells and/or magic items (the players will know a great treasure (or great magic, or great source of evil, or whatever they are "detecting") is located close by, but not how to get to it (which is also an ideal "organic" plot hook -- find a way to get to the treasure/magic/bad guy/whatever that we already know is there)), makes spells like passwall more useful -- allowing the players to shortcut from point A to point Z (or vice versa) without necessarily having to pass through points B-Y (the author of the WotC article would presumably disagree with my assertion that this is a desirable possibility, but, well, I already said I disagreed with most of that article...), and also incentivizes players to make more careful and accurate maps so that they will, for instance, realize when they're approaching the area that their locate object spell had told them the treasure/whatever was in, or that they're in an area where a passwall or similar magic could allow them to make a quick retreat to the surface.

 

[Melan]

Also, I had a few questions for Melan, if he's still reading.

Sure.

1. What do you think about the old mapping techniques on 8 1/2" x 11" grid paper? Many dungeons used to play a metagame of finding secret rooms or ensuring a level was fully mapped by filliing in all the spaces on a single sheet of paper. The Prince in T1-4 is probably the most famous case.

That's not a bad method of hiding some secret places. It is put to good use in Palace of the Silver Princess, for example, where you can see it in a completely elemental form - not a bad thing for a module aimed at beginners. That said, I don't map on gridded paper anymore, for entirely personal reasosn. First, I don't usually have grid paper on hand when designing adventures, whereas clean printer paper is always in a steady supply. Second, when I use grid paper, I start to get anal about corridor width, angles and similar things, so in the end the map lacks that special organic feel. Of course, some dungeon designers like Gary Gygax and Bob Bledsaw (in Tegel Manor, which is basically a "completely filled space" kind of module) didn't have a problem like that - they designed great maps with the features you describe.

2. What do you think of classical mazes in games? The kind kids used to buy in paperbacks for long trips. The ol' pencil tracing style. My understanding is, these are no longer considered fun and instead tedious mapping chores more than anything else.

Most of them aren't very good for precisely the reason you bring up. This is just personal experience (and thus anecdotal evidence), but I could never make hedge maze style structures work properly. Now things like rooms closely resembling each other, corridors "leaving" one edge of the paper and "entering" from the other, the "M. C. Escher memorial stairs" (which go down forever, forming an endless loop), etc., are fun in a whimsical way when used sparingly (probably as features of a thematic "maze" dungeon level). Classical mazes, not really. Or I just don't know how to make fun out of them. Also note that things like the small room maze or the "whirlpool" in B1 don't belong to this category.

3. Lastly, do you think dungeons (meaning maps of any interior space; buildings, towers, sewers, etc.) should put more priority on non-linear style or on logical construction by the in-game designers?

There is no conclusive answer to this but I will take a shot anyway. Generally, it is no big deal if the structure in question is small. If it is larger, it is fun to insert something like that, especially to break up the monotony. Most real buildings aren't really that fun - they are built for utility, not adventuring. Severs, in particular, are rectangular grids. Nobody likes to explore such a place because it is no fun after a while. In a game, enjoyable play is the most important reason for a thing being the way it is. If we emulate reality, we should emulate its quirky manifestations like castles built by eccentrics and people who thought it would be a good idea to have a secret passage going from their drawing room to the garden.

 

[The Shaman]

Moving on from philosophical points, on to application - after all, that makes theory meaningful.

Here's another suggestion that might work on grodog's list: the dungeon design offers meaningful clues to the intent of the builders with a practical benefit to the adventurers if they figure out what's going on.

In one wizard's dungeon, the wizard installed a series of traps to waylay anyone who slipped past his orcish guards. This raises the classic conundrum: how do the guards avoid the traps? Rather than hidden toggles, I made the design of the corridors the means of avoiding the traps: doors used by the inhabitants to move around the dungeon were always located at the ends of straight hallways, whereas doors placed immediately around a bend in a hallway were traps designed to catch the unwary.

The wizard instructed the orcs not to use the doors around the corners and to keep the hallways swept to prevent dust on the floor from giving away the traps. With this in mind, I placed additional clues beyond the architectural detail of the corridor shape: wandering orcs with brooms, a dead orc with a broom in his hand who tripped one of the traps (and which the wizard's thief henchman had not yet reset), and a couple of trapped passages with thick dust build-up due to a fat, lazy orc who could often be found asleep on a pile of straw in a forgotten storeroom.

The adventurers could therefore avoid most of the traps in the dungeon by looking closely at their map.

Someone may have mentioned this already, but dungeons that are more than stone-walled corridors and rooms, that offer tactical advantages and challenges, are more interesting to players (and more fun for GMs to design). A series of descending caverns linked by an underground river and waterfalls instead of corridors, or water-worn passages inside a glacier come to mind - these environments create new challenges and provide the players an opportunity to use their characters' spells and abilities in novel ways.

 

[T. Foster]

Also, I had a few questions for Melan, if he's still reading.

I'm not Melan, of course, but this is a topic I have a lot of interest in, have given a lot of thought to, and have a lot of opinions on, so if you don't mind I'm going to "answer" (which is, to say, use as springboards for semi-related musing) these questions as well:

1. What do you think about the old mapping techniques on 8 1/2" x 11" grid paper? Many dungeons used to play a metagame of finding secret rooms or ensuring a level was fully mapped by filliing in all the spaces on a single sheet of paper. The Prince in T1-4 is probably the most famous case.

I don't see this as 'metagaming' (in the standard, pejorative sense) but rather as "good play" (and, incidentally, pretty much the only reason to ever go to the bother trying to draw an accurate map instead of a "trailing" map. This brings up the point of whether mapping in a dungeon is considered purely a player-level phenomenon or if it also has an in-game character-level component. It seems to me most people view/play it as the former, but I've always prefered the latter -- if the players are drawing a map, then one of the characters must also be doing so (and have the proper equipment, light, etc.). If the "mapper" character dies and no one recovers his body, or loses his equipment to a fireball, or just doesn't show up for the session, then his maps aren't available and the players must either draw new maps or rely on their memories. This is the same reason I'll never draw on the players' map (I might sketch out the shape of an oddly-shaped room on a piece of scratch paper if the players seem confused by the verbal description, but that's it) -- the accuracy of their map is their concern, not mine. In most cases, there's no need for a player map to be particularly accurate or to scale, and trying to do so is a waste of time. However, there are a few circumstances where having an accurate map can be, if not necessary, at least helpful -- determining if you've circled back to a previously-explored location or are exploring new territory, finding your way back if you become lost or misdirected (if you get caught on the wrong side of a one-way door or sliding wall, knowing where you are and where the exit is can be valuable, even if you can't get there directly; likewise if you're teleported to a new, unfamiliar location comparing your new map to your old one might help you to regain your bearings more quickly, if you're able to spot familiar features), and, last but not least, being able to deduce locations of secret rooms. If you've got an accurate map that is more-or-less completely filled in but has a conspicuous blank spot or two in the middle, deducing that there might be secret locations there that are worth searching for isn't 'metagaming' at all, it's the reward for careful mapping. The group that chose not to make a careful map and relied instead on a trailing map (or just their memories) likely had an quicker/easier time of it, both on a player level and a character level, but the opportunity cost is that they're more likely to miss the "easter eggs" of secret/hidden rooms.

2. What do you think of classical mazes in games? The kind kids used to buy in paperbacks for long trips. The ol' pencil tracing style. My understanding is, these are no longer considered fun and instead tedious mapping chores more than anything else.

As a player I love exploring mazes -- it's, in some sense, my very favorite part of the game. And as I DM I love designing them. But, alas, I realize that I seem to be in the distinct minority, and that most players apparently find them terribly frustrating and boring (the same way I feel about riddles and math-based puzzles, I suppose). Therefore in my dungeon-designs I try to split the difference by including maze-like areas but making them "optional" -- the players (unless they're exceptionally dim) will be able to recognize it for what it is and have the choice to explore it or not -- it's never "mandatory" for them to explore a maze-like area (in order to find the stairway to the next level, or the great macguffin necessary to defeat the BBEG, or whatever). It's a trade-off, like the mapping situation above -- players who choose to explore the maze will get "easter eggs" in the form of extra treasure (or knowledge, or whatever); those who don't will miss out on the easter eggs but they won't have had the hassle of exploring the maze, so they have to decide which they prefer.

3. Lastly, do you think dungeons (meaning maps of any interior space; buildings, towers, sewers, etc.) should put more priority on non-linear style or on logical construction by the in-game designers?

I don't see that the two necessarily have to be opposed. As a general principle, however, when in-game logic/realism comes into conflict with fun-condusive game-play and it's not possible (or feasible) to resolve it in a manner that satisfies both, I always favor the latter. Players, I find, will almost always excuse something that "doesn't make sense" as long as they're having fun, whereas if they're bored and not having fun, the fact that the dungeon is appropriately ventilated and has sufficient kitchen and bathroom facilities isn't going to make them enjoy themselves any more. Which isn't to say I'm in favor of totally random, arbitrary, and wilfully non-sensical dungeon design because in fact I do think a dungeon that follows some discernible logical patterns and "makes sense" (at least on the big-picture level) tends to be more satisfying to players than something that feels completely arbitrary, like it was rolled off the random tables in the back of the 1E DMG, I just think that keeping the players at the table engaged and entertained should always be the first priority, and in-game logic/realism should only be a consideration so long as it's compatible with that. (Plus that, the fact that the sort of "megadungeons" we're mostly talking about in this thread tend to have been built either by insane wizards or mysterious pre-human races excuses a lot of "illogical" design -- yes, a rational person would never design his home in such a way that you have to traverse thousands of feet of corridors and dodge assorted traps to get from the kitchen to the library, but who can say but that from the perspective of the Mad Archmage Zagig or the Serpent Men of Yuan or whoever originally built the dungeon that just such an arrangement wouldn't have been perfectly sensible. As long as the players are engaged and having fun, these sorts of flimsy excuses are usually sufficient to explain away seemingly illogical design-choices; if you've got players who consistently complain about the logic and realism of the designs and refuse to accept these sorts of explanations, there are likely bigger, player-level issues at stake which the realism-complaints are serving as cover for).

 

[grodog]

use of challenges that allow PCs to learn the dungeon environment from their successes and mistakes; some good examples of this occur in Scott Casper's Greyhawk Castle gaming fiction written @ http://kinazar.com/SouthProv/viewtopic.php?t=336&start=0 (down at the moment, but will likely be up again soon)

Scott's forum is back up, and I particularly recommend the 2nd chapter episode, beginning at the bottom of page 1 in the forum link above, continuing through page 3. It's a good (fictional) illustration of PCs leveraging familiarity with the environment's challenges (the secret doors in the pit trap, the type of trap in the room beyond the secret door, etc.).

 

[Victim]

In my opinion, exploration is usually BS. Only if order of operation is significant does non-linearity matter. But if the PCs can go routes A, B, and C, then they will eventually go through all of them. So if the order in which they proceed is unimportant then A->B->C = B->C->A = B->A->C, etc. At that point, the dungeon may as well be linear.

Exploration also constrains player choices in another way. Normally, splitting the party isn't viable, so the group should only be going one way at a time. Then effectively only one decision matters, so the rest of the players sit around bored.

I think secret areas and hidden treasure as a reward quickly lead to stupid play. If you can find a gem inside a monster's belly, why not cut open every monster? Why not strip out the dungeon and hammer at every surface to find secret doors? In my opinion, it was a stupid annoyance to run against every wall holding down the open doors button in Wolfenstein 3-D, and it's stupid in DnD.

Using blank spots in a map to find secret doors only is extreme metagaming IMHO, since those blank areas probably need to be structural supports. Plus the dungeon engineers weren't limited to a sheet graph paper, and could easily put the secret room off the edge of the map.

Logical dungeons are useful to players since the context can provide additional clues about which way to proceed, making exploration decisions more informed and thus more meaningful.

 

[T. Foster]

I think secret areas and hidden treasure as a reward quickly lead to stupid play. If you can find a gem inside a monster's belly, why not cut open every monster?

Because it would be bad tactical play -- if you're taking the time to fully dissect every monster then you're wasting time (on both a player level and character level), exposing yourself to many more wandering monster checks (thus, presumably, meaning more monsters you'll have to dissect, leading to more time wasted, and more wandering moster checks...), and using up your resources in a way that isn't likely to return significant rewards (better to use them accomplishing things that you know are going to get you treasure). Sure, if you chose to you could spend an entire session in one room killing every monster that came in and dissecting it in hopes of finding a gem in its gullet, not exploring anything and not accomplishing any goals (whether DM or self-imposed), but that would be a very lame and unsatisfying sort of play (on both a player-level and character-level) and I can't imagine why anyone would actually choose it.

Why not strip out the dungeon and hammer at every surface to find secret doors?

Again, because doing so would be bad play -- a completely inefficient use of time (real- or game-) and resources. If given the choice to go down to level 3 to fight some level-appropriate monsters and gain some level-appropriate treasures and you choose instead to stay on level one and spend 2 game-weeks (and perhaps as much as a full game-session or two) strip-mining the entire place on the off-chance that you might find a hidden "easter egg" treasure of a few hundred SP (which you'll only get 1/3 XP value for anyway, since it's a dungeon level 1 treasure and you're 3rd level) that's certainly your prerogative, but's it's very bad play, both on the character-advancement level and on the players-having-an-interesting-time-at-the-table level, and it's not the DM's fault, it's yours.

In my opinion, it was a stupid annoyance to run against every wall holding down the open doors button in Wolfenstein 3-D, and it's stupid in DnD.

I agree that this would be stupid and annoying, which is why you shouldn't do it (at least in D&D -- can't speak for Wolfenstein 3-D). Time and resource management and an ability to set goals and stick to them without becoming distracted by red-herrings and minutiate (and an insatiable desire to accurately map every inch of the level, or to receover every copper piece of potential treasure, are both unquestionably minutiae) are among the key elements of skillful D&D play (see Gygax's essay on "Successful Adventuring" in the back of the 1E PH), and the way the DM tests these skills is by presenting opportunities for the players to make good and bad decisions in these areas -- presenting possible distractions, red herrings, and fool's tasks that offer little reward for much effort and waste time and resources, thus preventing them from being able to accomplish the more worthwhile tasks they originally set out to. The lure of potential treasures that can only be recovered with much time and effort, so much that it probably outweighs the value of the treasure and precludes the recovery of other treasure, is every bit as much a test of player skill as any monster, trap, or puzzle. And the appropriate response to such tests isn't to blame the DM for including them and condemn them as bad design but rather to recognize them for what they are and not be taken in by them.

On a slightly different subject from the same post:

Using blank spots in a map to find secret doors only is extreme metagaming IMHO, since those blank areas probably need to be structural supports. Plus the dungeon engineers weren't limited to a sheet graph paper, and could easily put the secret room off the edge of the map.

Who says that every blank spot is a secret room or that the dungeon doesn't have necessary structural supports even with a hidden room here and there? Or that there aren't secret rooms off the edge of the map? (Perhaps the DM mapped the level on a 5 squares/inch grid and made it look like a "complete" level at 4 squares/inch but has secret rooms in the 'extra' squares off the 'edge' of the 4x4 sheet -- this is actually a clever trick which I think I might use! ). And I still fail to see how this represents metagaming in any way if we assume that the characters, as well as the players, are making maps of the level.

 

[RFisher]

Regarding the "if the players don't find it, it was a waste of the DM's time" idea:

1. It may be found by other players--or just other PCs played by the same players--in the future. It's really cool when you find something that has been secreted in a dungeon for a long time that many other players/characters have missed.

2. If the dungeon isn't one that is going to see reuse, then the "missed" portions can be lifted & put into the next dungeon you design.

So, I don't find them a waste of time. Besides...

3. Lots of stuff in my dungeons that the PCs may never discover only cost me a minute or two. That's a small price to pay for the enjoyment both the player & I have should a PC discover it.

...and...

4. Sometimes I get an evil enjoyment from knowing how close the PCs came to something really juicy yet missed.

That's my experience.

 

[Dr Simon]

To return to the original post, it's an interesting way to analyse adventure flow, and I don't you think you need to restrict it specifically to dungeon style play, which is really where the geography defines many of the choices rather than slightly less tangible concerns like you might get in a city-based adventure.

I have a few ponderables:

It is possible to design a fairly branched map, yet include events or some other form of direction to influence the choice. The module commonly hailed as the exemplar of rail-roading, DL1, for example, actually has a *fairly* open dungeon structure. There are two branches to descending into Xak Tsaroth (the lift or the long way), and several unconnected encounter areas once there that would resemble a Caves of Chaos type diagram in Melan's scheme.

BUT: The lift is pretty cool, and everyone I know has *always* taken that. There is a 'cut scene' that efectively shows the players where the dragon that they are looking for is living, thus herding them towards that encounter area instead of others. Even here, though, they can go in the front door or through a secret passage.

Secondly, how important is it for a dungeon-based adventure (as this is the specific topic) to have a climactic encounter? The Caves of Chaos do not have such a thing. Arguably it is the Temple of Chaos in area K, through virtue of probably being the most leaderlike of all the figures. However, clearing that room does not end the adventure. One could also argue that each cave area has its own mini-climax in the form of the chief of that particular tribe.

Does one need such an event? It can be satisfying to achieve an end-point, but by its very nature it requires a bottleneck to be set up so that the final encounter *is* reached. Is that, to merge with another thread, rail-roading? To use the DL1 example again, all roads eventually lead to the dragon, if not physically then by virtue of clues, hints or plainly being dead-ends. But one can still approach the dragon through different routes. (I'll leave aside the narrative nature of the climax as a different discussion).

On the other hand, take S3 Expedition to the Barrier Peaks. It has a fairly branching structure, hidden areas, areas with coded access, various fun encounters, things to play with, but... absolutely no form of narrative structure whatsoever. The opponents do not get noticeably tougher, there is no 'boss' monster, no sense of climactic achievement. It is a fun adventure to play, for about 3/4 of its length, but the last part begins to feel flat becaus there is no point - the 'puzzle' of it being a science fiction setting is by this point solved but nothing new is added to the mix.

 

[grodog]

Melan---

I was flipping through Rob Kuntz's new Maure Castle dungeon levels in Dungeons 112 and 124 over the weekend, and thinking about your mapping schematics. I think both The Statuary and Chambers of Antiquities levels would be useful discussion points, though the maps are substantially smaller and less complex than the levels from the era of Castle Greyhawk and El Raja Key: Chambers of Antiquities in particular appears relatively simplistic in comparison to The Statuary, but it has pretty discrete areas that are only reachable from a few locations (probably more like B2 in flow).

For reference, the Dungeon 124 Maure Castle maps are available on Paizo's site at
http://paizo.com/dungeonissues/124/DA124_Supplement_LRes.pdf (I guess there wasn't an online supplement for issue 112?).

Anyway, I thought it might be interesting for you to analyze these new MC 3.x maps that are made in the old style, to see if they differ substiantially from the modules you've already discussed. I think it would be interesting to see a visual representation of the dungeon from the Realm of the Slime God adventure, since of the three Jaquays modules you mentioned, it's the smallest (that said, I'm not trying to make more work for you than you've got time for, either ). I'm also going to give this a try too, and it would be interesting to compare how we assess the same maps