## 2014-07-24

### In Which My Girlfriend Gets Hooked

So I've been trying to get my girlfriend Isabelle into D&D for what, 15 years now? The first go-round was as part of a long-running 3E D&D campaign with rotating DMs when it first came out and we were newly dating, so she was willing to humor me a bit more, but it really didn't take for her and was always a bit confusing and oblique. I've had numerous one-off games over the years, and the best match I'd gotten up to now was the solo-thief adventure, module O1 (The Gem and the Staff).

But then this Jul-4 comes around and she rather hesitantly agrees to join me at Paul S.'s house where we'll plan to play D&D for 3 days straight, purely old-school style, using Paul Jaquays' mega-dungeonish Caverns of Thracia (1979). And to our great mutual surprise, she is now totally hooked. She got up every morning that weekend increasingly hungry for what the party would explore next, where to find more treasure, how to possibly find and strike down the Minotaur King. We wrapped up Sunday afternoon and she was still jonesing for more. We've been talking for the last week (as I write this) about when we can play more, how we can jointly write dungeons and adventures together, etc., etc.

One of the great things is to be able to see the game through fresh eyes. My girlfriend doesn't have any back-in-the-day nostalgia. She first slogged through 3E and did not find it to her liking. (As a side note, the new 5E rules basically look to me like 3E with some of the parts moved around, and I was over that circa 2007.) But as we've reverted back to the most old-school stuff, and really gotten the first chance at a proper "mega-dungeon" style run, it's completely converted her. Here's a few observations that she's made independently in the last week since that game:

The large but limited dungeon environment felt manageable for the first time. Here it seems clear that she's locked onto the strengths of the mega-dungeon design structure. She said that prior campaign games seemed endless, without any apparent limits, and therefore somewhat pointless. But this environment excited her in that each foray allowed the players to focus on one subsection of the caverns and have a real possibility of exploring it in full, defeating it and solving its puzzles, and sacking its treasures and experience. So it felt like the players could set clear goals and actually "win" for the first time; but there were more subsections to be chosen and won over in later gaming sessions, and thus the excitement built on itself. And those sub-sections seemed rational and coherent in ways that other games did not (a nod to Jaquays' design, I think). When Isabelle said this, I felt like I was hearing Gygax's words on successful play from the back of the PHB all over again.

Having an advance sketch of the environment intensified interest. Two completely insane accidents served as an experiment which I would have never devised on my own. One is that Paul's character walked up to a statue of Apollo and cried out, "Oh Apollo, blessed healer, touch me and make me fly" -- to which I gave the old 1% chance that a god responds to an entreaty, and then couldn't believe my eyes when I leaned around the DM's screen and saw the percentile dice publicly come up the requisite "00"! So Paul's character took the opportunity to fly around the entire "outdoor" level and map out the gardens, orchards, and classical palace from above. On the next session, the party used this information to guide an assault the palace and just happened to catch the officers of the guard by surprise in the 2nd room, kill them, and take sketchy maps of almost the entire dungeon complex. Once I turned these over, all the later sessions started with players huddling over the maps and picking locations that seemed interesting, mysterious, or promising, and working out the most strategic paths of attack and areas to avoid or defend against, etc. The whole proceeding took on an extra level of strategic thoughtfulness and commitment, and the players clearly had their destinies in their own hands in a way that I'd not seen before. Isabelle & I have since been talking about what other devices could be used to give players similar advance, partial knowledge as to options about where they can explore in the future. (To which I'm thinking of some classic adventures that start with multiple obvious entrances for selection, like G1, G2, B2, etc.)

The locations seemed rational and not random. Apparently this was Isabelle's first foray into a dungeon that had a recognizable king, officers, guards, servants, a kitchen, dining area, etc., etc. To me this seems old-hat (again: see any of G1, G2, B2, etc., "Gygaxian naturalism"), but it had somehow escaped me that she'd never had a chance to experience that. The fact that players could quasi-correctly guess what the next few rooms contained made the game again more concrete, immediate, manageable, winnable, and rewarding.

One thing that I'm now personally wrestling with -- and feel free to say that I'm late to the party as usual -- is that this kind of design pattern is totally dislike what is presented in any of the classic D&D rules as a adventure-design protocol. That is: Make a map, place a few specially-crafted encounters in about 1/6 of the rooms, another 1/6 with traps, 1/3 of the rooms with random monsters, and another 1/3 or more empty (see: OD&D Vol-3, Dungeon Geomorphs/Monsters & Treasure, Moldvay Basic D&D, DMG Appendix A, etc.) I've been trying to follow that protocol, with frankly little success, for a number of years. But almost none of the classic published D&D adventures follow that design: practically every room has something to interact with in it, and nearby rooms are coordinated together in clans or supporting design or usage. Jaquays Caverns of Thracia has over 126 numbered areas in it (several with sub-locations lettered A-H or so), and not a single one is an empty room. (In fact: Area #85 makes a gag out of this by being labelled "An... Empty Room", and then when the author tries to inform you of this, he comically breaks down and can't resist putting in a set of giant watching eyes on the wall that put a curse on the players if they stay within.)

This is really bothering me at the moment, and I'm feeling like I've been blind (unlike those giant eyes) to what really hooks people into a good dungeon adventure. Perhaps we might say that while Gygax & Co. greatly advanced their dungeon designs from their initial mid-70's creations, the advice on dungeon construction stayed locked at that early state of OD&D in much later rulebooks. My own concern is that without random-table supports, I may not be personally creative enough to come up something interesting in every single room of an expansive mega-dungeon.

But Paul Jaquays did, and thanks to his creative genius it seems like my girlfriend is finally hooked on D&D. More interesting stuff to come later, I hope.

1. >> She said that prior campaign games seemed endless, without any apparent limits, and therefore somewhat pointless. But this environment excited her in that each foray allowed the players to focus on one subsection of the caverns and have a real possibility of exploring it in full<<

That's an interesting observation. I have found that some new players freeze up as I tried to explain that they can do anything and go anywhere. If you've ever taken up a new sport it can be useful for someone to say 'for today just stay here and keep defending' to restrict your focus to something manageable.

>>players huddling over the maps and picking locations that seemed interesting, mysterious, or promising, and working out the most strategic paths of attack and areas to avoid or defend against, etc. The whole proceeding took on an extra level of strategic thoughtfulness and commitment<<

I completely agree. I love maps and my players love scrutinising them so maps have become a most prized kind of treasure IMC, and I usually find a way for the party to get hold of them. Since these places don't exist, maps in D&D are both the reality and the abstraction.

==

There is an over-reliance on the thoughtless machinery of random tables in the wider community particularly among those who want to publish their material. I think CoT is probably the best D&D supplement and everyone could learn something in it about design (or thinking).

1. I think I agree with that in a couple of ways. The OSR trumpeting of more-tables-is-better has confused me for some time; my goal is always to strip things down more, and get faster play at the table (ideally from memory). I do use one table on my screen for critical hits details, but why you'd want an ever-larger sheaf of paper to page through at the table mystifies me.

That said, you're helping crystalize the use of the "maps" results in random treasure tables in a way I hadn't seen before. Particularly as a kid, I'd always try to skip or avoid those results because they weren't interesting to me. But maybe that's a golden spot to input more player agency, and err on more information in those maps rather than less.

2. OK, hold onto your hat, I have several disconnected thoughts:

1. You're welcome. Sounds like though it was my birthday, you're the one that got the best present.

2. You now owe me. Please try to convince Jenn that D&D can be just as much if not more fun than WFRPG.

3. I really think you would enjoy Michael Curtis' Stonehell, despite your opinion of the one page format. I wish I had thought to have you peruse it while you were here. Maybe next time, or maybe you'll find it worth the $15 to get your own copy from Lulu. Or maybe we just need to find a venue where I can run it for you. 4. On having sketchy maps ahead of time, I totally agree. It may be a bit cliched at this point, but I think it's such a great mechanism to really grab the players and get them into the game. I'm really curious to hear if you guys can come up with other ideas that evoke a similar level of engagement. 5. I totally agree on the adventure design problem. I've definitely had times where I stocked my own dungeon using those rules, and had it come out kind of flat. That said, my current process is a very slight twist on that. I start off with either my own idea of 1-3 major inhabitants, or use tables to generate them (Matt Finch's Adventure Design Deskbooks are great for this), then start going through the rooms putting in specific rooms that I think are required given that power and it's goals. You could argue that this is the "few specially-crafted encounters", only often this ends up filling half the rooms or more. Then I go to the tables, but I use the results on those tables as inspiration to add further specially crafted rooms influenced by the nearby denizens. If I roll something that doesn't immediately inspire me, I re-roll. Sometimes one roll will lead to several ideas which I run with and use to fill all the rooms I can. I also often treat the "empty" result as "empty save for some possibly confounding elements that have little to do monsters or treasure." This whole process ultimately started out by the book, but clearly jumped track at some point. 6. Oh, and finally speaking of adventure design, I recommend you check out Tony Dowler's "How to Host a Dungeon" (link: http://planet-thirteen.com/Dungeon.aspx). It also is a great starting point for inspiration when creating dungeons, and is practically a game in of itself. I once played this with BigFella with one of us (me) running the tables and rules, and the other (BigFella) doing the drawing, and both of us interpreting results. It was pretty fun, and I imagine you and Isabelle would enjoy it and make something pretty cool as a result. Note, there's a free version at the above link that's perfectly playable for a trial, but I did eventually fork over the$5 for the PDF which has enough additions to make it well worth the cost.

1. Great comments! You're right, it was a total treat for us. :-) I'm constantly wanting to rope Jenn into a game... I'm glad she could at least sit in the room with us, I was hoping that will have a long-term effect...

Actually Kent's point above about the predominance of treasure maps in the old tables, which I always used to skip out on, is a pretty good point. Isabelle's first stab at the issue was to get the PC's drugged and then have ghostly visions at times as they explore the underworld, yowza. Or ancient forgotten cave paintings of the place. I suggested some riddles but her take on that is that it's cliched. But still.

I agree, I need to upgrade my design skills, and I think that old advice is certainly not what created the G/D series, for example. I think you're way out in the lead on me on that. I did check out "How to Host a Dungeon" in the past after you & BigFella recommended it... it seemed a bit more complicated than I wanted, but I don't think I actually played it out. However, I'll make a note to get Stonehell, I was on the cusp of getting it when it first came out, your recommendation puts it over the threshold.

2. I'll second a lot of what Paul said. This is, however, mostly a jumble of my thoughts on the points raised.

First off I'm glad to hear that Isabelle's caught the bug for D&D. Get a creative person engaged and you can be guaranteed to get some very interesting stuff out of it. I really look forward to what the both of you can cook up together.

(I guess I always thought she was having a good enough time with the old GA campaigns, I still laugh about her one crack about Schubert's elf child character being so excited about being sold to the circus. I blame 3rd ed, which I also have come to see as something that needs to eventually be gotten over.)

I'll give another thumbs up for "How to Host a Dungeon". I remember that Helga's Heroes session that Paul and I played it very fondly, and still have the notes somewhere for the Volcano Lair of the Burning Lich. (Or the ancient dwarven calamity known as "The Great Quack". Never pick up a duck in a dungeon.)

On thought about HtHaD, though, is it's really more of a dungeon contextualizer than a generator of map/key, in my opinion. I think one positive result is that it establishes some factions, which I think is a neat thing to do in a large dungeon format. I've seen that approach in several megadungeon projects and I think it's really cool.

I've been stocking dungeons on the fly a lot for the campaign I've been running at work. I've been using the stocking table and sticking with the results I get, but I generally start with a central idea or theme and fill out key accordingly. (For example, I've recently been keying an extensive lair for Morgus the Malevolent, a name that Paul would recognize from past Creepy Crawls, and for whatever quixotic reason I decided that this dungeon should be brought to you by the letter M, so there's mummies and morlocks and manticores and much much more! :))

I find that empty rooms are kind of rare, at least as far as the dice I'm using say. Unique rooms are what I get far to many of, but they do push me to challenge myself. I've taken recently to using the Unusual Effects table in Labyrinth Lord to give me an idea of what makes a room unique.

As for empty rooms, I generally look at them as "rest zones", areas where the party can regroup and maybe even hole up to recharge (I know this flies in the face of what Paul calls "Goldilocks Syndrome", where the party is bedding down in somebody's "house", so to speak, but I tend to get more irritated with the "Home Base Yo Yo" model of play, where the party gets a little scuffed up and decides to withdraw. Of course I punish them by re-stocking the dungeon, but I do find it kinda irritating.

I gotta give a shout out to Dyson Logos and his wonderful maps. Maps have always been a mild stumbling block for me because my creative juices just flow better when I have something to pour them into rather than starting from a blank page. I'll also note that the style in which he does them make them idea for handouts and the like. I've come to welcome "Treasure Map" coming up on the treasure tables because I can usually print out a fragment or the entirety of the Dyson map and key it up in time for when its needed.

I guess I'll wrap up with what I always say about randomized tables in gaming. Sometimes they give you poetry, and sometimes they elbow trumpet rutabaga.

Random tables are good for the rough cuts, the big chisel for the block of marble. Eventually, though, you need to grab the wheel and polish the final product by hand.

3. Good stuff, B.J. :-) I guess I should sit down and actually play through a session of HTHAD and see if it's fruitful or inspiring to see it in play. At first blush it was higher-level results than I'd expected.

Interesting to think about dialing in the proportion of "empty" rooms, because old rules vary anywhere from 1/3, to over half, to as much as 80% (which is ludicrous). Admittedly I lean more towards Paul's philosophy on resting in-dungeon. I tend to be fond of Gygax's note in the original rules that the side-rooms and dead-ends are there to "trap players being pursued by monsters".

Also interesting to think of using the random tables for the big stuff and then manually detailing it (like HTHAD), as opposed to manually developing a theme and set pieces and then randomly filling in the corners (like in the old rules). Perhaps that's a better philosophy -- maybe use it to generate a theme for a half-dozen chunk of rooms at once?

3. Dan - great post and very disappointed I missed a chance to play Caverns of Thracia - it's a classic.

I totally agree with your thinking on Empty rooms. I've replaced them with Hint rooms in my random tables. Hint rooms are nominally empty of treasure or opponents (barring wandering monsters) but contain at least one incidental clue to the history of the dungeon, the nature of the inhabitants, or indications of threats and opportunities. Things such as fresh orc graffiti from the nearby lair that patrols this area, an already triggered trap that reveals how others of this type are concealed, a damaged mosaic of the original builders of the dungeon holding a glowing hammer which is secreted somewhere on this level, or deeply gouged door and an old campfire showing that this is a secure place to camp even from wandering beasts.

I look forward to reading future blog posts as your thinking on this topic matures.

1. Thanks for posting, Adam! That's a really nifty idea for the empty/unoccupied rooms. Hints, I like that.

4. A couple thoughts. I think the rational dungeon (including maps sanitized and ready to show players) is an awesome target to shoot for. When you're playing every Friday, though, it becomes less and less plausible to provide that level of prepared location for players. I think the industry has done a pretty lousy job of providing tools to help with that. It seems the choices offered are the two extremes of go completely random or buy a pre-made module.

Here is my attempt at something in-between:
http://recedingrules.blogspot.com/2012/04/dwarven-outpost-kit.html

(I'm currently revising this so you don't need to make physical stencils)

The other thought is that for new or young DMs, working with multiple intelligent factions in a dungeon is exponentially harder than just a bunch of caves with beasts or mindless undead. Again, another place tools for a DM would be helpful.

1. I agree with that, and in fact one of my disputes with the old B1 module is how hyper-detailed it is, which as a model for a new DM is kind of discouraging (or least hard to live up to). That's a good point that those book guidelines are maybe not bad for newbie DMs (like EGG & Co. were at one point), but there's no advancement of the art for more graduated DMs. Certainly when I use random tables I move things around to taste and try to make sense of them, but to date even that hasn't been great. Love the 3D look from your stencils by the way!

5. I should clarify that I don't think the complete absence of randomness should be the goal, because it helps keep the DM from being put in an adversarial position and it makes it fun for the DM too, by providing surprises that make for an emergent story.

But, yes, completely random results can be chaotic and seem senseless. Better random results filtered through the DM and built on at least a skeleton of a rational location.

6. This is awesome. I love reading this stuff. I also desperately want to get my hands on a copy (digital or otherwise) of the original Caverns of Thracia adventure. I can't tell you how many times its been recommended to me and I've never heard of anyone playing it and not having fun.

@Kent Like you I have noticed in the past that not all players take to the whole open world thing quickly. Lord knows I don't. I find most people need some parameters, framing, or context to work with before they really start to be more independent and free wheeling. Then you get the occasional player who takes to the freedom and runs like a madman and you spend more time playing catch up than leading them.

One thing I would caution against is the use of the word "interesting " when thinking about or building dungeons/locations. I try to think of it as "each location/room needs something of note" instead of "each location/room needs something interesting" because I find that using the word interesting tends to lead us in the direction of the crazy, fantastical, or weird. While not necessarily bad, I don't think we want every room to be that way, unless you are in the castle of the God of Madness. Including the mundane helps to make the exceptional stuff standout more and creates a sense of normalcy that players can identify with.

I also find starting with broad ideas and then adding one specific detail can help to create unique experiences for a location. For example, if the players walk into a room and you describe it as a dining area (a fairly broad concept) than that can have wide variance depending on the context of the room. Are you in a castle? A mansion? A fort? A cave? Most players will envision the dining area of a palace very different from the dining area of a fort or goblin-infested cave. That's good. The one specific detail gives the players something to latch onto that can provide information about the location or its inhabitants. Telling players that they enter a kitchen with a live fire and bubbling pot immediately tells them that someone either just left the room or is about to come back in. This allows you to use multiple rooms of the same type and yet still make them unique. Just think about the bedrooms in your house.

The other thing I try to think about is what the rooms or locations say about their inhabitants and does that match up with the inhabitants you've chosen or the idea you have in mind. Another example, if the players enter a fort manned by the Knights of Boombaloo, a notoriously strict order of knights, and they notice that things are messy, disorganized, and sloppy then that creates a dissonance with what the players would expect from a supposedly strict and organized order of knights. Now, this could be done intentionally (perhaps the knights' reputation is overblown or these are really bandits posing as the knights so they can sack nearby traders) but it is something to keep in mind.

1. Yeah, I highly recommend Caverns of Thracia (I think I got it from RPGNow), it's the best quasi-megadungeon publication I've seen to date. Good point about shooting for "interesting" in every room, I'm with you on being a bit more restrained and not crazy lunatic in every single room.

One thing on the details, a bit of a tangent: the maps in Caverns of Thracia always show which way the doors open (like a little blueprint), which is great because it always comes up when PCs are trying to bash in or barricade doors -- with the standard Gygaxian "boxes" I'm always making that stuff up on the fly.

7. @Telecanter I agree with you here. Randomness by itself is silly but randomness with a theme and a skeleton can be hugely helpful.

Using random tables has helped me to design better dungeons and locations by not having to worry about coming up with every last detail on my own. By having a theme or idea in mind with a basic skeleton I can use targeted and appropriate random tables to populate the non-critical stuff. Having tables for random castle furnishings or random fungi can help with populating content for your castle or cave complex.

8. An interesting note: Underworld & Wilderness Adventures does not mention empty rooms. It says "unoccupied". I don't recall if the instructions in the geomorph sets or in the DMG changed this to "empty", but it's my conclusion that there shouldn't be hardly any empty rooms. In an area occupied by intelligent inhabitants, the rooms are store rooms or support areas of different types. In other dungeons, they show signs of previous use. Either way, there may be things of interest other than treasure.

1. Well, you're right, and that's a possibly useful interpretation (although Vol-3 doesn't give any other guidance, it just doesn't address those non-occupied rooms at all). Other works do use the word "empty": Dungeon Geomorphs ("Roughly one-third of the rooms should remain empty", e.g., 3 of 8 in the sample key, p. 1), DMG Appendix A (Table V, 1-12 in 20, p. 171), Moldvay Basic (p. B52), etc.

9. The DMG may say empty, but the other choices are trap, treasure, monster or stairs. I think empty means no monster, trap, treasure or stairs.

Then again I have never really minded empty rooms, at least they go fast. Rooms with tons of content, but none worth discovering (no value, no clues, nothing interesting), I think are worse.

1. Yeah, see, I kind of lean that way myself. Although lately I've been saying at the start of a game to new players "some areas will be totally empty, don't assume secret doors, etc." to avoid sometimes endless searches for "something must be here".

2. The subject of "hidden stuff" has come up often as we've gamed over the years, from the "pixel bitching" search for secret doors in "Tomb of Horrors" to arguments over the unrealism of searching every 10' block of hallway or room for traps.

I tend to favor contextual clues for hidden things. Scorch marks on the floor from unseen traps. Unusual wear patterns on the floor. Information blurted when you lean on the wretched kobold janitors. Etc. Make it more about the player's deduction skills instead of a dice roll. (Plus when you get right down to it, as a GM I want them to find stuff, instead of having them pass up all the content hidden behind that secret door. I've had some experience with secrets too well kept and it's frustrating.)

Empty rooms in Labyrinth Lord have a chance of containing treasure, I always try to hide stuff pretty well in that context, rather than just leaving out free treasure like honor system breath mints by a restaurant cash register. But even then I'd probably put around clues, or include it in a treasure map. That's worked out for me pretty well.

On a related note, I've come to appreciate traps that are obvious but begging to be sprung. So instead of just bad luck being the culprit its player cussedness or greed that dooms 'em.

3. Yeah, again, I find myself a bit weak at coming up with fresh clues all the time like that. Certainly it's an attractive option. In my house rules I'm pretty liberal about giving bonuses to the d6 search checks (from race, Intelligence, clues, etc.), and I'm pretty comfortable with how that's been working. In the game the other week the players seemed attracted to anything simply decorating the wall -- a relief, a tapestry, etc. -- and thus found about 3 secret doors that way (with my standard rolls at that point).

There was also an example of a completely telegraphed trap (deep rumbling when they approached bejeweled statues) that they were unable to resist. You're right, that was fascinating and kind of delightfully horror-movie to watch happen.

4. A couple thoughts on clues.

A: Perhaps a random table would be in order. With stuff like "unusual wear on surroundings" "dead end trail" or "chalk sign from previous adventurer" on it. Hrmmm.

B: Regarding what was tipping off the players in your Caverns of Thracia run, I think it just shows to go you that description = significance in tabletop rpgs. I think that's why the 100' of nondescript wall in the depths of the "Tomb of Horrors" pisses me off so much.

5. Yeah, I'm totally down with the description = importance observation, that seems clear to me at the moment.

I wish I could re-run Tomb of Horrors at some point, being sure to make the initial hidden riddle available to players, and then running it as an extended puzzle-solving exercise. I think that particular secret door in the long hall was where I finally decided to terminate the event, because it finally dawned on me too how awful that was going to be? I think?

10. Pffft! You've got more than enough imagination. Heck, Gygaxian naturalism makes things easier for you.

Here's a trick: find natural groupings of locations and then randomize what's in that entire group. These rooms in the NW corner: goblins. Ok, they'll want guard rooms at the entrance points, their chief will take the best room for himself, they'll want a kitchen and a larder... It's a home, so treat it as such.

These rooms here came up trap, so the monsters don't go there because TRAP! So you put nasty traps at the entrances to that zone with some untouched treasures beyond.

This sort of thing is well within your powers. Go forth and (rationally) populate! ;)

1. Thanks for the vote of confidence!! Good advice indeed. :-)

11. By the way I meant to say congratulations on the girl friend getting hooked.

I must say if I could chose between winning a million dollars, or having my wife get hooked on D&D, I think I would pick D&D.

12. I'm not too familiar with classic modules (something I should really rectify), so I can't comment on how well they might adhere to those probabilities. However, I always thought "empty" just meant no traps or monsters, not literally empty. I see I'm not the first to mention that, but I wanted to emphasize how important that is to consider

Also, I don't think a pre-fab dungeon necessarily needs to follow those guidelines. That'd somewhat defeat the purpose of using one and reduce its utility. They should probably have a super-majority of "empty" rooms since that's a gameplay feature, but I wouldn't worry about monsters, traps and treasure

1. Here's the quickest example in reach: Keep on the Borderlands, Cave A, that has 5 rooms inside. 4 rooms have monsters (3 groups of 1 kobolds, 1 pack of rats) and 1 room is a locked food storage room. So that's just 20% of rooms sans monsters, and I think that's the general theme for all the rest of the module and other published dungeons. That's like the total opposite of the 33%, 60%, or 80% empty-of-monsters rate recommended in the old books.

Perhaps my greater grief (as obvious as it should be, perhaps) is that none of those guidelines ever suggest the "subset of closely connected rooms with monsters of the same type supporting each other" which is an obvious and strong design element in the published stuff.

13. Added note on "manageability": I was tickled to find Dave Wesely (developer of Strategos N and Braunstein in Arneson's early gaming circle) suggesting the exact same word in Peterson's "Playing at the World", p. 68, footnote 122:

Dave Wesely has advanced the very plausible suggestion that the dungeon's appeal lay in its finite scope, its manageability -- something his Braunstein sessions lacked. There were simply less choices and complexity when what lay before you was a branch in a tunnel or a room full of monsters. It was thus easier for referees to run fantasy adventure games in a dungeon and easier for players to decide what to do.