Classic dungeons: What makes them great?

Doug McCrae

Legend
Certain dungeons - The Temple Of Elemental Evil, the giants' lairs from the G series, the Caves of Chaos from Keep on The Borderlands - are widely regarded as classic. In an attempt to improve my own dungeons I'm trying to analyse what makes them great.

1) Variety of monster type while remaining within a certain theme. One doesn't want all orcs, but otoh one doesn't want a total zoo. The four elements in ToEE are a classic example of a theme. That said, much of Gary's monster placement in the 1e era seems to be completely random.
2) Variety of encounter type - monsters, traps, NPC interaction.
3) Variety of encounter difficulty. For example, the feasting giants in G1 are expected to be far too numerous to be fought conventionally, forcing the PCs to be stealthy, tricky or strategically brilliant.
4) Size. Not too big, not too small. But how big is too big? How small is too small?
5) Raison d'etre - the dungeon needs to have some reason for its existence. It can't just be a vast unexplained monster-infested hole.
6) A reason for going? PCs need a reason to go to the dungeon. I'm not sure if this required, maybe general treasure-seeking is enough, without it even being a specific treasure.
7) The nature of the place. Gary seemed to really love temples - ToEE, Hidden Shrine of Tamoachan, Forgotten Temple of Tharizdun. Coincidence? Or do temples to strange gods make for better dungeons?
8) Internal conflict. A popular element is a conflict between two or more factions. This is the case in ToEE, Sunless Citadel and the Conan classic tale, Red Nails. This adds both interest and an element of roleplaying as PCs can join one or other faction. Or play each side against the other, Yojimbo style.
 
Last edited:

log in or register to remove this ad

I'm running my group through B4 - The Lost City, which takes place in a pyramid in the middle of the desert. I decided to run it after reading a new introduction adventure to the module in the January 2007 issue of Dungeon (Mask of Dreams). I liked the story aspects of this adventure (going to a party where the guests start wigging out, then getting rounded up by slavers) and it gave the players a really good reason for going on the adventure - saving the guests from the slavers.

I like the B4 module for reasons 2 through 8. Who wouldn't like tromping around a pyramid? I've taken out some of the variety of monster types. The pyramid is a base for the evil priests of Zargon, and I thought more rooms should be devoted to their needs. I left in some levels untouched, like the undead level. I streamlined some of #8 - internal conflict within the dungeon. The original had this, but I've moved the adventure about 40 years after the original module, so the Zargonian's have control and the other groups have been pushed into the background.

Having fun running B4, I bought some other classics like Scourge of the Slave Lords and Desert of Desolation, but I'm not as satisfied with them. I think it's because I purchased the 2E campaign adventures, which run 130 pages each, rather than the shorter originals. These bigger modules have way too much information in them! I'd have a much tougher time running these I think, and I'll probably go back and just purchase the 1E versions instead.
 

The thing I always point is how many of the classics came out of tournament adventures. And thus, they're really tough, in the sense that they're set up to kill most parties that come through in an attempt to distinguish who the winner of the tournament should be. That's a lot of what people perceive as DM-vs-players emphasis in the 1E era.

But also there's a great sense of accomplishment if you do get through them.
 

SPOILERS BELOW


DIFFICULTY. I do think that many of the classics were very difficult. There was much less fretting over whether something "was an appropriate threat." As a player, you felt a sense of accomplishment, if you merely survived Tomb of Horrors, the Giants/Drow series etc.

DYNAMISM. On a related note, because they were generally short and because the statblocks weren't as unwieldy, many of the classics were more dynamic than modern modules -- the Delve format is probably a great example of how some 3e products degenerate into pure set piece battles. Back to the classics, think for example, of the humanoids in Forgotten Temple of Tharizdun, they have a strategy, they move from one area to the other if a noisy incursion is made, etc. Think of Strahd in the original Ravenloft, he may have different motivations, the whole is set up for him to make guerilla attacks in a variety of directions. Think of the smugglers in Saltmarsh, especially on the ship. If the PCs do this, they do that. If they attack from another direction, they do something else. If the PCs try to bluff their way on board they do something else. If the PCs are too stupid to figure things out, they simply sail away. Note, the well regarded 3e module Red Hand of Doom manages to be somewhat dynamic but I cant think of too many other 3e modules like that.


SHORT and EASY TO MASTER/ALTER. The classics are generally short. It is much easier for a DM to master a module and then add creative touches if it is 8, 16 or 32 pages than if it is 192 pages and all of the stat blocks are gigantic and all of the components of such stat block interrelate to each other. It was easy to move Nosnra or Strahd and others around because their stats fit on an index card. That is no longer true with modern villains.

LOVE of PLAYING Vs. RULEMAKING. To me, in some intangible way the classics reflected authors who loved PLAYING the game (or perhaps killing those who played the game). There are still some good module writers out there (e.g. mona, jacobs, logue) but it seems like all of the fuss is over designing rules, not creating cool places to play in. Why are the Hill Giants having a drunken feast in Steading? Because its cool to sneak around or to have a brawl in a feast hall with drunken giants! Why is there an Ice Chasm in the successor module, because at least pc or npc needs to get thrown off a cliff and eaten by an ice worm! Why is the ToH a death trap, because Gygax's players were getting cocky and they needed a serious challenge. Why is the Iron Golem in Maure Castle so horrible? I dont know but it creates a serious challenge for PCs to circumvent him and/or defeat/escape him. Is it "fair" within the "rules", no and who cares, as long as the Classic DM does this sparingly. Is it "fair" that Erik Mona put a beetle swarm in a first level dungeon -- this is literally against the CR rules for D&D (I consider Whispering Cairn to be a 3e classic in the vein of the old classics)? No, but he does it sparingly and it serves to make the PCs worry, scout, prepare, consider running.

The biggest thing I've noticed both in reading about and playing old time or new classics and the bulk of 3e stuff, is that sense of danger -- oh s--t, we may die.

By the way, I dont really consider KoTB to be a classic certainly not by comparison to Hommlet, Saltmarsh and the Cult of the Reptile God.
 

Doug McCrae said:
Certain dungeons - The Temple Of Elemental Evil, the giants' lairs from the G series, the Caves of Chaos from Keep on The Borderlands - are widely regarded as classic. In an attempt to improve my own dungeons I'm trying to analyse what makes them great.

When you talk about what makes a dungeon great, largely you are talking about the map. To talk about the encounters beyond the level of a map and a simple key, you have to talk about what makes the adventure great which is a related but slightly different thing.

The problem is that when people talk about 'a great adventure' they don't usual mean anything but that it is fun, and this is of only tangential help to us if we want to design good dungeons. In the case of something like 'Keep on the Borderlands', this boils down to little more than nostalgia. The 'Caves of Chaos' are not a great dungeon, and get such a blanket pass largely because they are often peoples first experience with a great game. In fact, they are in many ways a very poorly designed dungeon (in other ways, a very good on).

My basic rule of thumb when designing a dungeon is 'Map = Story'. In a dungeon based session, the map is your story outline and should reflect this. The rooms in a dungeon are chapters, and the corridors between them are the connections between chapters. The different ways that the players choose to travel between these connections represent the different ways that the story can unfold.

A dungeon map should therefore be designed according to the narrative to be played out, or if it is part of a larger story, then according to the role in the story that the dungeon will play out. Story and map are the same, so every adventure is a dungeon, even if the designer takes pains to make it not look like a dungeon. Likewise, every dungeon needs a story.

Older dungeons by and large where not deliberately designed with story in mind. This results in a story that can be traversed without any real dramatic tension or larger meaning emerging during the process. Random story walks can be alot of fun, and they have thier place. They maximize the ability of the PC's to create thier own story, but they tend to become tiresome very quickly once the newness of the gaming experience wears off. 'Keep on the Borderlands' is remembered fondly only because its a low level module typically played by newbies. A random story walk can be good after an extremely linear section, to regenerate the player's sense of being in control. But if that's all you offer, most experienced gamers are going to want more meat.

The reverse problem is more often seen in modern dungeons. In an attempt to tightly entwine the map with the story, the dungeon is laid out in an extremely linear fashion. The result is a strong sense of drama, but very little player choice in how the story plays out. These can quickly turn into site-seeing tours, which are either joyless slogs or at best like having an interesting story read to you.

Classic dungeons get the right mix.

If you look at 'ToEE', its got lots of things going for it. It has a very high degree of player choice, and yet, the map can be broken into subsections each having its own theme, structure, and rising action. And the map in total has the same structure. With very few clues from the text, mostly involving under what conditions the various doors can be opened, the basic story of the module can be recreated from the map alone. I'm not saying that the module is perfect, but it is a great dungeon.

On the other hand, if you look at 'Caves of Chaos', you mostly get chaos. There are themed sections, but very little in the way of a story. Most players go through the dungeon with no clue about what is going on and with very little non-combat interaction with the inhabitants. It's very basic dungeoneering - kill things and take thier stuff. Nothing wrong with that, but you can kill things, take thier stuff, and be part of an exciting story at the same time. There are alot of things about CoC that has always bothered me. One of them is that if you don't steer the players, they are very likely to get in over thier heads in a hurry. Any door but the kobolds or the goblins is likely to TPK a party of first levels almost immediately. There is nothing necessarily wrong with challenges outside of the characters current abilities and a strong starting hook is a great thing (both play important roles in certain stories), but such challenges need to be structured quite carefully. Another problem is that in a story sometimes you want to have a moment where you unveil something. If the 'reader' looks ahead and sees what's coming, the big moment is lost. The converse problem is that sometimes you want to foreshadow what's ahead so that the players are anticipating it. 'Caves of Chaos' potentially has both problems. Not only is the mechanations of the temple of chaos hidden from the players, but when its revealed it may not even be a climatic point in the 'story'.

On the plus side though, the small sections of 'Caves of Chaos' are very good dungeons indeed. They are about 6-10 rooms, which is as much as you want to bite off in a single chunk without being so small as to have no mystery to them. They aren't linear, but they always have strong starting hooks, strong themes, and strong climatic fights with sub-bosses plus unique features to explore (an armory, a jail, an enchanted chapel, a secret room, etc.). What's needed IMO is not a complete rewrite, but to organize the Caves as a whole according to the excellent design of its parts. I've actually done this, and with very little work or addition on my part, it turns what is in my opinion a potentially pedestrian, frustrating, and forgettable dungeon crawl (if its not your first time) into the true classic it should be and I think EGG intended. Simply, I break the dungeon into 3 parts (levels) - sections suitable for 1st level characters, sections for 2nd level characters, and sections for 3rd level characters. As you go deeper into the dungeon, the action 'rises'. I then add to this structure, components of a meta narrative. Thus, passing through from chapter one to chapter two involves passing through a small arena/commons complex (shared by the various demihuman tribes), where the players can encounter thier first clues about tribal rivalries, perhaps have thier first in dungeon RP encounters, and first get glimpses of the secret temple deeper within. Passing through from chapter two to chapter three involves passing through a lesser temple complex, where the characters can get the first foreshadowing of the minotaur 'boss monster' and face off against a temple priest for the first time. Finally, after negotiating the minotaur maze, the players can enter into the climatic fight against the temple cult itself and thwart thier evil plans once and for all. By which time, the players actually care about killing this particular foe and taking his stuff.

Alot of that could probably be handled with extremely good DMing, but as an introductory module why demand of the DM so much when something as 'simple' as a good map would do the trick?

In my opinion, if you want to look at well-designed D&D dungeons, you need to look at 'Tomb of Horrors', 'Ravenloft', and the modern classic 'The Whispering Cairn'. Ravenloft may be the best map for a D&D module ever. It's simply too good to talk about meaningfully, because it doesn't have any problems with the possible exception of the catacombs being too large. If you can make a map that intriguing, I'll forgive you for being a little 'long winded' in places. People tend to either love or hate the content of 'Tomb of Horrors', but as a map its a triumph of making what's basically linear map seem non-linear. After playing the module, and then picking it up to DM it, I was shocked at how linear the map was. It doesn't seem like it when you are playing it because of the illusion of choice that it gives you, the few alternative paths, the uniqueness of the individual locations, and a couple oppurtunities to backtrack. Plus, it's the definitive 'find key and open door to access new level' module. Finally, 'The Whispering Cairn' may be the best introductory adventure ever written, and a good part of that is owed to the extraordinarily deep understanding Erik Mona displays in his dungeon design. It's an amazing peice of work, and every DM ought to pick it up just to study the design.

In my opinion though, one of the areas that cRPG's have ultimately excelled the pen and paper games that inspired them is in dungeon design. I think this is perhaps due to the fact that in a cRPG or one its offspring - first person shooters, adventure games, etc. - you've got very few tools for controlling the story but the map. The map is about all the computer can be made responcible for. If you can't design your adventure as a flow chart (basically a map), then you can't design it for the computer at all. Maybe the most inspired dungeon design of all time doesn't look like a dungeon at all and doesn't have combat in the tradiational sense. That 'dungeon' is the flowchart for the game Grim Fandango. It's an amazing peice of work just as a map, much less all the witty narrative and artwork that happens in the places on that map.
 

Doug McCrae said:
Certain dungeons - The Temple Of Elemental Evil, the giants' lairs from the G series, the Caves of Chaos from Keep on The Borderlands - are widely regarded as classic. In an attempt to improve my own dungeons I'm trying to analyse what makes them great.
I think the main thing that makes a great dungeon (not sure that all those you list above qualify, but that's a different thread) is the presence of player choices which have significant impact on success or failure and adequate data available to the players from their surroundings to allow those significant choices to be at least partially informed (educated guessing is fun and tension-provoking, wild stabs in the dark tend to get frustrating and boring after a while). I very much like Melan's "circuit diagram" analyses of linear vs. complex dungeons he posted a few months ago because it illustrates perfectly that linear dungeons really do limit the availability of "significant choices" the players can engage in when exploring a complex. The more non-linear the dungeon the more significant each choice the players make can be.
 

Doug McCrae said:
In an attempt to improve my own dungeons I'm trying to analyse what makes them great.

One major element you aren't going to be able to intentionally replicate is the shared experience. Many of these modules would merely be good to very good if it wasn't for the fact that so many people played them, and discussed playing them.

"When I stuck my hand in the mouth on the wall..."

"Really? When we first encountered the kuo-toa we just called them 'fishies.'"

"We had no idea what this black elf thing was. The elf in our party told us they were long extinct."

Look at the list, it is dominated by adventures from the days when a module release was an event. Two of them were the first two official 3E adventures, there is a shared experience there, even if it was slightly lesser than the AD&D stuff. After that you only have a couple of adventures in the top 30.

Don't get me wrong, these clearly had things that made them stand out from the pack. I'm not sure how many would reach "classic," though, if they had as much selection as we have today.
 


Doug McCrae said:
Certain dungeons - The Temple Of Elemental Evil, the giants' lairs from the G series, the Caves of Chaos from Keep on The Borderlands - are widely regarded as classic. In an attempt to improve my own dungeons I'm trying to analyse what makes them great.

I think a lot of your points, Doug, are valid, but they are still just design elements for potential greatness, so to speak: many people love the GDQ modules, for example, but many have also loathed them, since they weren't DM'd well, or weren't played well, or some combination thereof.

The experience of playing a great module is the three-fold collision between a great module, a great DM experience, and a great player experience. If any two of those experiences fall down, IMO, the whole likely suffers in the end: even with great players and a great module, a mediocre DM can spell boredom; even with the best DM playing with engaged players, a crappy module may well still suck; a great module and a great DM with players who are uninspired may result in a bland session.

An analogy: one of the key concerns that Gary Gygax has voiced about publishing the original Castle Greyhawk materials (keys and maps) is that, by themselves, they still would not enable a group of players and a DM to recreate the magic that was the crucible from which the game sprang:

Gary Gygax in revised TLG blog entry on 6 August 2006 said:
http://trolldens.blogspot.com/2006/08/for-those-of-you-who-may-or-may-not.html

The Castle and Dungeons were and remain a dynamic creation in progress. As any good GM knows, role playing games are fluid and the creative juices never stop flowing. Gary believes a revised and coherant set of new levels that have the same feel and spirit as was in the original versions of the ever-changing Castle dungeons makes ultimate sense for Game Masters and players alike. Of course the new versions of the dungeon levels will be packed with the old, often fanous or infamous ecounters that were in all the vestions of the work, as well as some new and unexpected ones as well. After all, the old school does not mean stasis and lack of improvement.

Speaking of that, as Gary has continually pointed out, the original levels had encounter notes taking up one line on a page. He created most details on the spot, and Rob learned to do the same. Any version of the original levels would be revised, because the encounter material from them would be unusable by most of the GMs that acquired it, mainly because they would have no idea how Gary and Rob would have managed the scant information, spun it into a richly detailed scenario.

Thus, it's the interaction between adventure, DM, and players that made Castle Greyhawk so memorable (in addition to the brilliance of the design, DMs, and players, I'm sure!).

I'm not trying to say that the adventure design isn't important, it's just not the only factor in a great gaming experience, and not even the most important one IMO.
 

Some very good points made above.

Echoing some points made above, I think it has to do more with the ability for the DM and PC's to create the "story" out of a skeleton module.

Thinking about my DnD initiation with the Blue Book Basic Rules and B1 "In Search of the Unknown" what struck me most, and stays with me today, is that the game is about creating a series of events based on situations that may or may not be intentionally connected. Randomness, or letting "fate" decide monster placement, dungeon design, etc seemed to be integral to the game.

The classic modules I enjoyed (thinking Hommlet and D1-D2 here) had a combination of good monster placement (by chance or design through playtesting?) and monster motivations that, for the most part, remained simple or thumbnail sketched, allowing for expansion by the DM in his own campaign world.

I think what made a module "classic" was that the batter used for baking during that period had a better combination of ingredients for some reason.
 

Remove ads

Top