A common behavior (at least at every table I’ve ever been part of) is the clean dungeon sweep: making sure that every nook and cranny of a dungeon has been explored, every enemy fought and every goblin slain. Players often do this because they don’t want to miss any treasure or experience. In an idealized world, this turns out to be an inefficient strategy that actually slows players down. In practice there are external considerations that the GM controls.
To see this, we need to simplify things a bit and imagine an entire campaign as a single dungeon that is a long corridor with rooms along it and many side paths of varying length like so:
The players can start from the left and advance towards the right with or without detours down the side passages. As they advance to the right challenges get harder and rewards get better.
The typical adventuring party will then explore lots of side passages and make sure they sweep them clean of all encounters garnering loot and xp along the way like so:
But for the same input of time, the party could instead ignore all the side passages and simply travel to the right, only hitting encounters down the main corridor and only gathering loot and xp found there:
The argument for the first exploration strategy is that it ensures that no experience or treasure is missed and thus results in better geared and higher level characters. This is the heart of the fallacy. In fact, the second exploration strategy leads to better geared and higher level characters. The important shift in thinking is not to compare the two groups at the point when each reaches room X, but rather to compare the two groups when they have each invested X hours in the dungeon. Both groups have been exploring rooms at the same rate, have been encountering challenges at the same rate and acquiring treasure and other rewards at the same rate. This means that both have had about the same number of encounters and received about the same number of rewards. The challenges and rewards received both follow the same type of distribution (some a few levels up, some a few levels down, some common some rare) so both have had about the same number of tough and easy encounters and about the same number of good and bad rolls on loot tables. The big difference however is that the second group’s average encounter level and average treasure level has been climbing much faster so their rewards have been more valuable. Thus they have a greater net xp and better gear. If at any point, either group discovers that the challenges they are facing are too difficult, they can always backtrack a little and grind until they’re able to move forward but the grinding the faster group does is more valuable than a much greater volume of grinding the slower group is doing at lower levels.
Of course this assumes the above setup. In sandbox games, for example, there’s no guarantee that challenges scale with time. Players can diddle around the starter area as long as they want or go crashing headlong into dire places beyond their abilities. It also assumes an infinitely available source of material. If the party speed runs your current adventure, it assumes the next is waiting. If they speed run that one too, the next one is expected to be on tap. Though in theory this could require you to plan many session ahead, in practice it probably only requires one or two. It also assumes an endless run with no endgame. If there is some end destination at the point when both groups arrive, it would seem to be the case that the slower group is better geared and higher level. However, keep in mind that all of the old loot rewards are replaced by later loot rewards or made insignificant in magnitude eventually and that generally XP awards and scales grow at a non-linear rate so the relative impact of earlier to later grinding diminishes rapidly and the payoff required to have a noticeable impact on gameplay grows as time goes on. Net effect: even if we compare both groups when they reach room X versus at time X, the slower group’s additional gear that counts is only from recent grinding, not early grinding and the net impact of their extra experience is unlikely to be significant unless your system has a highly linear experience system.
So of course the question of the hour is: “Who cares?” If you and your group are happy with the pace of your game, then the answer should be “Not you.” If, on the other hand, your game is slogged down with clean sweeps, hunting red herrings and chasing down fleeing minions, then share the above with your players (it also speeds up video game runs significantly). The clean sweep is a hard habit to break. You have to turn off your little OCD voice that says “But the chest we missed probably has the best loot in it!” This is also useful when your game has somewhere you’d like to get to, but you don’t necessarily want to skip a lot of the intermediary parts. There is also a suite of things you, as GM can do to push the clean sweep by the wayside.
Handwave the ends of adventures: Once you’ve found the McGuffin, or unmasked the villain, call it and move on. Yes, there are rooms left to search, treasures to find, enemies to kill, lose ends to tie up, but that’s a waste of time. Skip it.
Level up as soon as possible: If the system isn’t too cumbersome, allow players to level up, spend xp, etc… mid session. This will require a running total of xp, but a spreadsheet on your phone will allow you to keep a tally with minimal fuss.
Tweak challenges on the fly: If your players level up mid session, or if the adventure is turning out too easy, adjust things upwards. GMs often use these on the fly adjustments to reduce the challenge when they’ve accidentally made things harder than desired, but don’t be afraid to shift things the other way. Just don’t forget to increase the rewards when you do so.
Bring more material: Another common reason for the clean sweep is because the evening’s adventure is over and it’s clean sweep or call it a night, Instead handwave the leftovers, take a 10 minute break and while the players level, pull out your notes for next session. This requires a little more prep, though mostly it’s front end. Once you have a session in reserve you can go back to making one adventure per session until you have to burn two at once, then you have to catch up.
Back when I was running Pathfinder, I would be thrilled if the PCs would find a way to shortcut to the important, plot related stuff in the dungeon.
It’s great for them to have choice and options to become more powerful if they really want to spend the time and all of that, but honestly, I’m thrilled with clues that tell you “you could just go down the hall to the left, there’s a big sign that says Boss Monster, can’t miss it.”
The worst thing professional dungeon designers for RPGs can do is to create dungeons with the “only one solution” option in hand, whether it be that the boss monster is only vulnerable to a weapon in the dungeon, or there is a special lock that the PCs must have to be able to get to the Boss Monster.
That kind of design only makes players paranoid that the adventure is setting them up to fail if they don’t explore every single inch of the place.
That’s not to say a dungeon can never have a special weapon or special key element to it, but even then, you should have big huge clues, going in “You need A and you really should have B, everything else is just gravy before you pound on the boss.”
Excellent point. There are definitely adventure designs that encourage the clean sweep and those that discourage it. Contrast the “collect the 7 keys, use them to unlock the sarcophagus where the wraith lord lies scheming” playstyle to “there’s a bloody smear on the floor leading directly to the ogre and screams coming from his lair.”
Matt, this was a great article and good food for thought. One thing I’ve been thinking about a lot is those “filler” encounters, your “side rooms.” In a sandbox, exploration game they can be anything. But in a more storyline campaign, I don’t want them to seem to be just fillers.
And as your article said, this can be applied to the whole campaign as well. The random orc (or Klingon, or Vampire) attack on the side of the road can be fine, but it is much more meaningful if there is some link to the greater story (if you are running that sorta game).
Something I didn’t get into above, but if you abstract your current game, you can overlay the direct linear route over whatever route the PCs actually took (minus any dead ends or backtracks). So if you provide two hooks, one about the evil vizier trying to overthrow the duke, and the other about a horde of orcs rampaging out west, whichever bait they take and gets developed further is the main trail.
In the macro context, it’s very hard NOT to take only the main trail. You’re not going to develop both of those ideas into complete level 1-6 adventure paths and then after they finish one, you just go ahead and run the other, the fact that it’s way too low level be damned. You’d at least bump it up a few levels to make it an appropriate challenge.
Video games are a great place to see branching on the macro scale. They have their “main quest line” and then all sorts of side quests, and if you want to complete a level 2 side quest when you’re level 50, by all means help yourself.
Micro scale however has a lot more play because you DO prepare a lot of same level material and players can spend a lot of time chasing down same or lower level content with little payoff.
This article has me thinking again about the nature of XP.
In a classless XP-oriented game system like Savage Worlds one can sort of make a convincing argument that since the points are spent arbitrarily to increase whatever aspect of a character a player wants that XP are a generic “solve the problem” prize.
But in a strongly-classed game system like Pathfinder or D&D where the XP a fighter earns are really not the same as the XP a cleric or bard does as they “pay for” different advance styles one could (were one so inclined) to make a point that while a rogue might get XP rewards for avoiding a problem (as in a combat), a fighter should really only be rewarded for confronting it. I mean, if my Dwarf fighter is going to use his next bunch of XP to obtain Cleave shouldn’t some of that XP have come from situations in which Cleave would have been vital?
This, of course, is an old argument and one reason why XP-driven advance systems became less popular for such a long time in the 80s. In the end most people wisely choose not to sweat it.
Your conclusions are interesting, Matthew. Not sure I understand every nuance of what you are talking about, and will have to re-read to make sure I really am on your page. Which isn’t a bad thing.
Thanks for provoking thought.
Hmmm. At first, I was thinking some of the completionist ideals comes from video games, where you generally want to explore every nook and cranny in a dungeon, but then maybe that came from the old grognard days of D&D.
I suspect that loot’s a big driver of the clean sweep. Sure, if you follow the main plot thread, you’ll get the big bad’s loot, but you’ll miss the encrusted sceptre in room A12 and the photon driver that’s at the bottom of the well.
It’s like a lot of things: you can quit as soon as you do the hard thing (kill the boss), but there’s nothing to make sure that you get a minimum of the reward. So the reward (XP and loot) seems related to the time you invest searching every room, but the difficulty is never harder than the big bad. Because he’s the big bad!
Matthew’s right about it in the modern sense. If you’re not shackled to a module, you can move loot ahead, just like you move plots and monsters. If the players “fall behind” by taking the direct route, that just ups the GM’s reward budget until they’re back in balance.
Well, I get what your saying but my point was more that because of human nature players THINK they’re going to fall behind if they DON’T clean sweep, but It’s actually more likely that they DO fall behind. So in theory you shouldn’t have to move anything forward (although if they miss something you spend a lot of time on as opposed to just rolled on a table, I wouldn’t give anyone grief over moving it forward).
I’m not sure what started it. I suspect it’s more OCD human nature than anything else. You’re right it’s prevalent in video games which are littered with secrets where they have no business being.
However note that video games are also the place where completionism makes (usually) the least sense. Unlike the table top all the material is already there so there’s nothing stopping you from fast forwarding to the next bit and the sword you miss in dungeon 3 is definitely less powerful than the one you can buy in town 5.
Just from a tactical standpoint, you want to clear the sides so you don’t get ten rooms in, then have to fight your way out when all the nasties you left behind come out to see what all the commotion is about. You always want the way out/your back clear or things get interesting really quickly.
I think what you are describing here is kinda solved perfectly by ‘The Five Room Dungeon’.
Those five rooms should have the really cool stuff from the traditional 25 room complex without the cruft and forgettable. Even if there is a 25 room complex map, just detail the five most important rooms and leave the others to verbal description.