Adventure and Encounter Design

In discussing yesterday’s post about the goals of Echelon, I realized there are some unspoken assumptions on my part.

I mentioned how I didn’t mind ‘CR math’, the arithmetic used to combine Challenge Ratings to find Encounter Levels and thus estimate whether or not an encounter can be considered reasonable for a party of four iconic characters. This system, as written, suffers somewhat in a number of ways. I’ve found ways to work around most of these difficulties, that suit how I want the game to run, and no longer deal with the same problems.

The Challenge Rating system as presented in the Dungeon Master’s Guide runs into some complexity when it comes to determining experience point awards. Again, I typically use different means to determine experience point awards and thus don’t have to deal with those complexities.

It’s actually fairly simple.

Experience Awards and Advancement

I figured I’d start with the easiest part.

I typically don’t give experience points for encounters. I have learned that giving out experience points for fighting means that the PCs spend their time looking for fights, because that is the road to power. I’d rather see more goal-oriented adventuring, so instead I grant experience points based on the completion of story goals – successful completion ideally, but failure counts for something too.

Adventure Design

When I design an adventure I mark down a set of three to eight ‘story goals’ depending on the size of the adventure. There is usually one key goal, a couple of secondary goals, and some incidental goals. If reasonable I try to have about half of the goals relate to character-specific hooks. Each goal can have the following information

  • Brief (one-sentence) summary.
  • Problem statement
    • More detailed description of the goal.
    • Outline opposition and complications.
  • Engagement criteria (objective indicators that the goal was attempted or encountered).
  • Resolution criteria (objective indicators of the goal’s resolution).
    • Critical success (perfect victory).
    • Success (desired and expected result).
    • Mixed success (goal achieved, with negative side effects).
    • Failure (goal not achieved).
    • Critical Failure (goal totally not achieved, possibly permanently).
  • Resolution consequences (what happens if the goal is met, for each result).
    • XP awards.
    • Treasure and other rewards.
    • Story effects

In most cases there are multiple goals, which may conflict. I try to keep ‘mixed success’ to a minimum (it often indicates I didn’t make the goals detailed enough). I use the Challenge, Response, and Secret techniques to develop the goals – both to flesh out individual goals, and to explain relationships between them.

The awards and rewards for each goal are decided well before I get into detailed encounter development. If the adventure is a ‘mission’ then the director may actually describe the desired outcomes and priorities (“above all else, return with my daughter safe!” or “rescue my daughter if you can, but I must know if Duke Arlington is truly acting against me and if so, why”). For that matter, researching the adventure ahead of time may do much the same (“I know we want to retrieve the sword of vengeance, but I hear there are other treasures to be gained from this venture… and Tlarn is after one of them, so we might get lucky and be able to put paid to him, too”). The awards and rewards are generally proportional to the importance of the goals.

The expected difficulties are also roughly proportional to the goals they are associated with. I try to stay fairly consistent within a campaign, but I don’t have a fixed progression rate. I tend to work in units of ‘encounters’ for advancement. D&D 3.x expects 13.33 (40/3) between levels, I might have ten, five, or even fifty. Hard encounters are worth twice as much, very hard twice as much again, easy are half, and very easy are one quarter. At least, as far as planning, since you may be able to meet the goal without hitting all the encounters. I usually go about 25-50% over in the adventure design.

So, if I have an adventure with eight goals, I might have one primary goal, two secondary goals, and five incidental goals. The incidental goals might have two expected encounters each, the secondary goals three encounters each, and the primary goal five encounters (for a total of twenty-one). I might prepare another five to ten encounters around the edges of the relevant ones. These encounters tend to be easy to avoid (and may in fact be hard to get at – killing all the goblins might mean going out after the hunting parties that aren’t even in the lair at the time) and generally aren’t worth experience points. The extra encounters are not necessarily easy. A party of fourth-level PCs trying to evict the lizardfolk from the swamp might run into a CR8 nine-headed hydra – easily avoided if they don’t engage it closely, but a significant danger if they do… and possibly worth tackling anyway for the treasure it could be guarding, or even just the reputation they’ll gain from defeating it.

Note that the experience to be gained from achieving a goal is proportional to the goal’s importance, as are the expected encounters, but that there is otherwise no real relationship between experience points and encounters. Specifically, you don’t have to meet the encounter, you must achieve the goal.

Encounter Design

As mentioned above, I really only use Challenge Rating and Encounter Level to estimate how dangerous an encounter is likely to be. However, unlike the D&D 3.x approach I build each encounter on a character by character basis. I find single-monster encounters can be very difficult to balance.

  • If the single creature is significantly higher CR than average PC level, the PCs may find themselves outmatched (unable to affect the opponent, or unable to prevent being affected by the opponent – such as unable to hit it, while not able to avoid taking damage when it attacks).
  • If the single creature has a CR at all close to the average PC level, but there are many PCs, the ability of the PCs to focus fire and at so many more times than the opponent could result in a quick and easy kill.
  • If the PCs are not the ones the encounter is designed for, it can be awkward or difficult to adapt, especially on the fly.

Most encounters I develop have a primary opponent and secondary opponents. Each ‘opponent’ may be a single creature or multiple creatures working together (such as goblins riding worgs, or an ogre with a ‘pet dragon’). The primary opponent is typically the leader (of sorts) and the most powerful or difficult of the opponents, while the others are usually somewhat weaker – but still powerful or difficult enough that they cannot be readily ignored. I aim to have a number of opponents comparable to (if not equal to) the number of PCs and their cohorts.

Scaling for varying party sizes becomes very easy. Simply add or remove opponents until rough parity is met. For instance, an encounter might consist of a base of one ogre, a worg-riding goblin, and two pairs of goblin warriors (perhaps this was a meeting between emissaries of a goblin tribe and a neighboring ogre). We have a base of four opponents. If a player is unable to attend or a PC is otherwise unavailable, simply remove one of the opponents (a pair of goblins, I expect). If you have an extra player or two you might have two worg-riders and an extra pair of goblins.

Scaling for different-level parties is a little more effort. If the level difference is small you might just adjust the number of opponents as above (a higher-level party might be able to handle the extra flanking that comes from having more goblins, or the beating they could take from an extra ogre or two). I’d prefer to upgrade the opponents, though, such as bumping the monsters each a CR or two, or replacing them with more powerful monsters. I actually don’t have to do this one very often since I do the encounter development close to the time I expect to need it. Plot takes as long as I care to put into it, D&D 3.x mechanics can take quite a bit longer and might not even be needed if things don’t go as expected.

I have mentioned before that PCs are basically bullies. A balanced encounter for a four-member party is a single creature with a Challenge Rating equal to the average PC level. By CR math, each PC should be able to handle an opponent with a Challenge Rating four levels lower as a balanced encounter. It doesn’t necessarily work out this way if each is treated independently, but as a team this should be roughly so (the wizard might have a bad day of it if each PC is jumped by an orc barbarian, but if the martial characters can keep the orcs off him he might play a decisive role in the fight).

The baseline encounter could thus be a number of opponents equal to the number of PCs, each with a Challenge Rating (or Encounter Level, if more than one) four less than the PC level. This can also allow for mixed PC levels (the goblins go after the wimpy-looking guy while the ogres go after the tough guys, sort of thing). I try not to have each opponent be more than two creatures because it gets difficult to manage at play time and the focused fire effect can be overpowering (or the individual creatures are too weak to matter). Two working in conjunction is usually workable, four can fail.

Hard encounters usually have more powerful monsters (CR two less than PC level), very hard encounters yet more powerful (CR equal to PC level). They can also be done by doubling the number of opponents for each level, or doing both (twice as many creatures, with CR two points higher). For weaker encounters I usually start by halving the number of opponents before reducing their individual power.

I’ll try to draw up an example or two tomorrow.

No related content found.


  1. hadsil

    Reading comments at various forums, I find some DMs misunderstand what CR is supposed to mean. They don’t get it that a party facing opponents of equal CR means the party is expected to win using up only 20-25% of their resources. These DMs will then think the party is too powerful and want to nerf a character or two or 3E is broken.

  2. hadsil

    Oh, and opponents CR equals party does not mean the levels of the opponents equal the levels of the party. For a one-shot convention adventure I’ve had the party face their equals for the climactic battle. It’s a bloodbath. Fun, but bloody. That works for the one-shot convention adventures, not a regular campaign.

  3. Oh, I don’t think things are broken. I just think the PCs are bullies *g*

    At least, in that they tend to square off against things with a large advantage and are expected to win. I don’t question the need for this for most encounters, because otherwise PC attrition would be too high for the game to be playable.

  4. As I recall, though, a single creature of CR = party level is supposed to be a good encounter (20-25% resources consumed). *One* creature vs. four iconic characters, that is, not /n/ characters of CR = party level vs. /n/ PCs.

    IOW, if you want to have a figure for each PC, you probably want them to be about four levels/CR lower. If you want teams you’ll want each member to be a little lower yet (-6/-6 or -5/-7) but that’s getting pretty low.

Leave a Reply

Your email address will not be published. Required fields are marked *