Monday, 4 February 2013

The Megadungeon Paradox

As the blogosphere faces the possibility that all we'll have of Dwimmermount is a draft, critical evaluations of that draft are flooding in, from the dismissive to the constructive. One big part of the criticism - really just an extension of last summer's "9 rats, 2000 cp" beef -  is that the upper level maps are boring.

There's a reason for that, very much in keeping with the Phase 1 of the OSR that Dwimmermount represents. Phase 1 was all about copying the old stuff, with a "D&D is always right" mentality. Take a look at this paparazzi shot of Gary Gygax's Castle Greyhawk dungeon Level 1:


It's a hot mess of jammed-together, samey rooms and corridors, and from all accounts the keying was minimal. I'd rather explore even the upper levels of Joe Bloch's Castle of the Mad Archmage than this map - even though the Castle is another case of a Gygax homage intentionally putting the more generic stuff up top and the wilder level designs below. (Besides, the Castle has hidden depths to its design not evident from casual skimming - as I hope to demonstrate next post.)

But from experience of actual play, consider what this design approach means. Most explorers of megadungeons never get past the early levels - just as most campaigns never reach level 10. This means that most people's megadungeon experience is intentionally boring, with the real fun deferred for a lower level that never comes.

Phase 2 of the OSR means that enjoyment and amazement take precedence over carbon-copying the old school. The goal is not to reproduce the means, but the ends, of the class of 1974 - the sense of wonder and discovery now hard to recover from a set of expectations turned cliche.

This means that megadungeons should engage from the beginning, with early levels that reward exploration with variety - not the most awesome and dazzling stuff, certainly, but fun mapping, meaningful phenomena, intriguing hints that draw explorers deeper. I've tried to follow this principle with my first-level completion for the Castle - which I'm more and more inclined to release as-is this month - and with my new mega-project, one section of which is finished.

11 comments:

  1. "But from experience of actual play, consider what this design approach means. Most explorers of megadungeons never get past the early levels - just as most campaigns never reach level 10. This means that most people's megadungeon experience is intentionally boring, with the real fun deferred for a lower level that never comes."

    What's to keep you from taking a set of stairs down when you find them? Why does an entire level have to be mapped before moving on?

    From early accounts be EGG and OG on RPGnet, players would go as deep as they dared from the get-go.

    Dungeon play is not as linear as this point of yours implies.

    ReplyDelete
    Replies
    1. Part of the problem is that if a party goes to level 2 too early, they aren't good enough to meet the challenges there, and/or haven't found the Level 1 McGuffin that will help with Level 2.

      Delete
    2. You can have some level appropriate encounters on every level down, but with better treasure. The party can take the risk of going down lower levels to get better treasure vs an increased chance of encountering difficult monsters, which can be fled from.

      Delete
  2. They can run, sneak, hide, come up with creative ways to get around encounters that are too difficult.

    I've honestly never played or run a homebrewed dungeon-type environment where MacGuffins were needed to "clear" a level. That just smacks of linearity and makes the dungeon even more of a railroad than it already is.

    ReplyDelete
  3. Well, that's whats going on with Numenhalla for sure.

    ReplyDelete
  4. I know nothing of dwimmermount, but you've knocked down and non existent straw man when it comes to early dungeons being monotonous, challenge appropriate and linear. Have a look at Blackmoor, or ToTF, for a real example, or even the sample dungeon in the 3lbb's. I also see no evidence for that "phase 1" business in the OSR either. People have been innovative from the start, conversly, people also continue to clone.

    ReplyDelete
  5. I wrote a little piece today, and I don't think the early levels are boring of these old dungeons.

    I think that megadungeons are about slow-burn suspense, not action. I think that's what they are about, and some people find that boring.

    ReplyDelete
  6. I do think it's interesting to compare Dave's Blackmoor dungeon to Gary's Greyhawk. The first level of Blackmoor has 9 or so room but 10 different stairs down. There are even stairs down in the very first room.

    ReplyDelete
  7. I was actually mostly referring to the map design and layout of certain megadungeons, not what kind of monsters and treasure are found therein. Certainly there's room and precedent for out-of-level monsters and challenges, even in Castle Greyhawk (the defunct Axe and Hammer had a copy of Gygax's first level showing a hill giant on level one.) But what I don't agree with is the implicit idea in some designs that early levels should be intentionally made bland to force lower exploration. That should be driven by considerations of risk and reward, or by exhaustion of the easy pickings on an earlier level, not by getting tired of same-y rooms or encounters.

    ReplyDelete
  8. On what basis should we assume that Gary's maze-like map and minimal key led to a boring experience? One of the things about using your own creations as a referee is that a good part of the module probably remains in your head, accessible via a few jotted notes or keywords. I don't think this necessarily says anything about mega-dungeons as mega-dungeons, but rather about the difficulty of bottling a vision in a way that is useful to other referees.

    ReplyDelete
    Replies
    1. That's a valid question - see my following post where I end on the open question, whether it is better to key sparsely and put the fun in the hands of the GM, or to lead the GM there more explicitly. This depends, of course, on the individual GM, but there should be at least one product out there that shows how by example.

      Delete