James over at Grognardia has published an interesting essay “Schrödinger’s Dungeon” which has garnered a lot of replies in the few short hours it’s been up. Much as I like and admire James’ work for the most part, I fear he has missed the mark on this particular subject.
I launched this very blog with the lament that James mentions in his first paragraph; we are all the poorer that Gygax didn’t have a chance (whether in the later years of his life, or while he was still with TSR) to publish some version of his own Castle Greyhawk. I stand by that assessment.
One of James’ key passages is this:
When most people think of a “dungeon,” they expect a set of maps with a key describing rooms and their contents. A megadungeon, by its very nature, can’t be detailed in the same way. It’s a lot more “impressionistic” and relies heavily on ad hoc adjudication by the referee, as the players explore it. Not all of the megadungeon’s rooms are inhabited at any given time — this is important — and many of their inhabitants might change, depending on player action, referee whim, or the luck of random rolls. Likewise, even the geography of the megadungeon might change, as the referee adds new sections, closes off old ones, or otherwise alters what the characters have experienced to date.
Which is, of course, correct, but which does not address the central point of the Lament of the Old-Timers. We didn’t have any model for that process! It was only discerned after years of careful reading-between-the-lines in various disparate sources; The Dragon, fanzines, hints in modules and rulebooks, etc. When I, and others, lament that Gygax hadn’t published his own version of the Castle, I think it’s implicit that we would have expected that such advice would be sprinkled throughout such a product, implicitly and explicitly. If nothing else, than by example.
Rather, when you look at the modules that TSR did, in fact, publish in the earliest years, the lessons were exactly the opposite of what James defines (and I, in large part, agree with) as the parameters of a “classic” megadungeon. Steading of the Hill Giant Chief et al weren’t much given to spontaneity (although lip-service is paid to the notion that the giants will not stand idly by while adventurers make numerous forays into their lairs). There was very much the idea of “come in, clean it out, and move on to greener pastures”, and it seems that Gygax himself was consciously leaning in that direction. Here’s what he wrote in the DMG (p. 91):
…but when it is all over the monsters will not magically reappear, nor will it be likely that some other creatures will move into the newly available quarters the next day.
It’s at odds with how we know his own Greyhawk campaign was run, but it’s still an interesting (if contradictory) statement.
But back to James’ argument; I think the real value in publishing a “complete” version of Castle Greyhawk (or Castle Blackmoor, or Maure Castle, or any of the various ones that dominated the early scene up in LG) would have been as an example of a starting point. Even the fabled Castle Greyhawk started as a map on paper with notes. Even if the map changed, and the notes most certainly changed based on various player activities (as well as reactions imagined by Gygax and later Kuntz), that starting point would still have been an invaluable resource to have. Most especially if it were accompanied by the briefest of introductory essays in the form of “you know, in the original campaign, none of this was lasting; once the player characters wiped out the kobolds on level 1, such-and-such happened; in your campaign it might be very different” might well have made all the difference in the world!
Instead, we had a steady stream of tournament modules (not that I’ve anything against them! G1-3 and D1-3 are among my favorite modules ever). They reinforced in the minds of tens of thousands of young dungeon masters (myself included) that dungeons were intended to be relatively limited in scope, have a particular theme, and as a rule end up with a fight against a Big Bad Guy to win access to the Treasure Room (or, in some cases, multiple Big Bad Guys and multiple Treasure Rooms). We learned by the only example we were given.
Imagine what would have happened if, instead of cranking out tournament modules, TSR had settled down to publish a large, if skeletal, Castle Greyhawk. A couple of lines per encounter. Encouragement to DMs to “take it and imagine the hell out of it” throughout.
James is absolutely correct when he states that a published module cannot possibly capture the ever-changing-based-on-player-actions nature of a megadungeon. However, I think that misses the point of the Lament of the Old-Timers. He says:
In every case, the changes are in response to play and it’s this quality of megadungeons that makes them hard to put into a published form.
And I must disagree. Any dungeon, no matter how small, can (and should) change in response to play. Whether it’s a three-room crypt or a ten-thousand-room megadungeon spanning twenty levels, the change-in-response-to-play aspect is constant. What we lacked, for years, was a model of how to properly set up such an enormous playing field in the scope of a single dungeon setting. How to get past questions of “dungeon ecology” (which Gygax admittedly didn’t give a fig about in the beginning, but begrudgingly came to realize as being at least something to take into consideration)? Factions within dungeons were, apparently, a staple of Gygax’s approach. A “still life” of them in action, at least as a starting point, would have been a great help in such a context. Ditto the “random zaniness” factor, noticeably lacking from the earliest module efforts of TSR.
James concludes with:
I simply don’t think such a thing would ever have been possible and any attempt to present a “Castle Greyhawk” trapped in amber would necessarily feel inadequate. That’s the nature of the beast and therefore I think the only way to experience a proper megadungeon is to build it yourself.
And I think here he lays the foundation of his own inconsistency. If I write my own megadungeon (which, incidentally, I have), it exists solely as a starting point. From the moment my players hit the corridors, it’s up to me, as the Dungeon Master, to alter and adapt, to change and manipulate, to reflect the actions of those players and how I imagine the inhabitants would respond. If someone else downloads Castle of the Mad Archmage and does the same, or if any of us had done so with a hypothetical “complete” Castle Greyhawk from 1982, what, exactly, has changed? Each of us would have taken (or will take) it in completely different directions, according to our own DMing style and the actions of our players.
Which, I think, is the whole point.