Improving Encounter Economy and Design

A-Z 2016 "I"Developing encounters for the Pathfinder Roleplaying Game can be simplified, and easily made more flexible and scalable. I will show how below, contrasting with the PRD Designing Encounters text.

Sample Encounter Design

In the sample encounter design in the PRD, the party consists of two fourth-level PCs and four fifth-level PCs.

  • Step 1: Determine APL. Calculate the Average Party Level (add all PC levels, divide by the number of PCs, round up; subtract one if the party is three or fewer PCs, add one if the party is six or more PCs.
  • Step 2: Determine CR. Choose a Challenge Rating by assigning a CR equal to the APL plus or minus a certain amount. Subtracting one is an ‘easy’ encounter, no change is an ‘average’ encounter, adding one is a ‘challenging’ encounter, adding two is a ‘hard’ encounter, and adding three is an ‘epic’ encounter.
  • Step 3: Build the Encounter. Determine the experience points to be gained from the encounter. Assign monsters, traps, and other challenges until that XP budget is spent. Adjust the CR for elements that improve or worsen things for the PCs.

This can be done more easily, but I’ll need to share some background first.

Encounter Design by CR and APL

Realize that the ‘average’ encounter is actually pretty easy. PCs are expected to win their encounters, spending 20-25% of their expendable resources (hit points, ammunition, and magic, mostly). A group of four PCs against a single ‘level-appropriate’ monster are basically ganging up, four on one, against a creature nominally as powerful as any of them. It is by no means a fair fight, nor is it intended to be.

If each fight were a 50-50 proposition, the PCs would have only a 6% chance of surviving four fights, and a 3% chance of surviving five. Instead, they are expected to come through four encounters in a day with some wear and tear, bleeding and bruised, while five they should expect to be reaching their limits — possibly with some casualties.

An encounter with a CR four points above the APL is beyond ‘epic’, but actually in principle more or less a fair fight, where either side stands about the same chance of winning. In the D&D 3.x era this was marked with ‘caution, at this point you can expect at least some PC deaths in the fight’.

A group of four or five third-level PCs against a CR 7 hill giant is a beyond-epic fight, and either side will stand a chance of winning if the giant can engage. With the giant’s high attack bonus and damage (greatclub +14/+8 (2d8+10 before Power Attack), 2 slams +13 (1d8+7), ranged rock +6 (1d8+10)), PCs are going to get hurt. And possibly dead; 2d8+10 is 19 points of damage on average, and many of the squishier PCs can’t take that. A couple of rocks sent at the wizard or sorcerer stands a good chance of doing him in. The other PCs might still manage to kill the giant — action economy means there are probably still three or four attacks against the giant for each of his, though his 85 hit points will keep him around for a while — but there’s a good chance some PCs will die. The hill giant is not only a match for the party as a whole, but individually greatly outpowers any of them.

An encounter with a like number of creatures with the CR equal to the APL is similarly expected to be a fair fight — which is to say, unhealthy for the PCs. In principle each opposing figure is a match for any one of the PCs

Addressing this is pretty simple.

Scale the Encounter by PC

Rather than working out the APL (with math and adjustments) and trying to find a combination of monsters and other threats that balances both the average power of the PCs and the number of PCs (and thus the action economy), simply have one threat per PC. That threat should be of a CR about four lower than the PC’s level.

So, given the party of two fourth-level PCs and four fifth-level PCs, half a dozen CR 1 threats should be enough to occupy them briefly, if intelligently applied. Second-level NPCs (with elite class levels, not NPC class levels) are CR 1, so a similarly-constructed lower-level party should be fine, or ghouls, small elementals, wolves, or similar. It may seem kind of light, but that’s mostly because the monsters have little that really threatens the PCs.

On the other hand, a CR 5 monster such as a basilisk isn’t such a great threat either at this point, as long as you’re not ambushed at close range. Stay more than 30 feet (or better, 70 — the basilisk has a 20-foot move and a 30-foot attack) and pepper it with arrows until it drops. AC 17 at this point is easily defeated, touch AC is only 9 (acid arrow is a go-to, and the +4 Reflex save is pretty easy to beat too). About all the basilisk has going for it is 52 hit points and a gaze attack.

In discussion on Facebook it was mentioned that knowing that “30 feet (or better, 70 — the basilisk has a 20-foot move and a 30-foot attack)” smacks of metagaming.

I mentioned the specific values to illustrate how easy it is. I assume a fair bit of competence and professional knowledge among the PCs, especially about level-appropriate stuff and at fifth level. In this case the fighter (who has no ranks in Knowledge) probably has spoken with enough of his colleagues and picked up enough minor knowledge about monsters that he knows to stay a bowshot away and fill the basilisk with arrows.

Those classes with knowledge-oriented abilities (Knowledge skill, bardic knowledge, etc.) might know the specifics and details (more precise ranges, what exactly the powers are, and so on). There still is benefit to the knowledge-oriented abilities, they go beyond what I expect as basic competence at the adventuring trade.

For a ‘standard encounter’, this seems to work well enough.

Increasing Encounter Challenge Rating

The above advice gives you an encounter that is more or less ‘standard’ for the party. If you have more PCs or fewer, it still works out pretty well: if the party were just a two fourth-level PCs, a pair of CR 1/2 or CR 1 threats should be workable. If the party consists of ten fifth-level PCs, a pack of wolves — especially if it initially goes after a perceived weak PC — could be an exciting encounter, even though ultimately it shouldn’t be a big threat.

To increase the encounter rating, you have a few choices.

Simplest is to choose higher-CR threats. If you choose threats with CR three less than the PC levels, you’re looking at the equivalent of an “APL+1” encounter, a “challenging” encounter. Using threats with CR two less than PC levels is roughly an “APL+2” encounter.

Another option is to add more threats. If the PCs are faced with a squad of orcs, give them a leader two CR higher for every four or five orcs. This makes the encounter CR one higher (four or five CR 1 orcs is an average CR 5 encounter, adding a CR 3 leader moves this toward CR 6).

For a major set piece battle, I do both. I’ll usually increase the base CRs to the PC levels minus two, add a suitable number of mid-grade threats (CR based on PC level), then perhaps a boss figure two levels higher. For the party described earlier (two fourth-level PCs and four fifth-level PCs) this might mean two CR 2 threats, four CR 3 threats, one CR 4, one CR 5, and one CR 6 or 7 threat. Altogether this is approximately CR 9 or 10. However, no one figure is excessively dangerous (highest is only one or two higher than the PCs), but none are so weak they can be ignored. This is probably a good major battle, such as at the climax of an adventure. If the PCs come in rested and prepared (and not surprised, tricked[1], or unlucky) it should be manageable, but if they’re exhausted or near the end of the day it could be very touch and go.

[1] In an adventure I played years ago, we were well-forewarned of a ‘witch’. We prepared to deal with a sorcerer or wizard… turned out it was a succubus, completely unaffected by most of our precautions. That was a Bad Day… even before our DM, student of the old school, had her successfully call her boyfriend, a balor. We abandoned the treasure, the other enemies, and noped our way out of there as fast as we possibly could.

The halfling paladin — in armor — was completely unable to keep up.

Me: “I grab him and carry him under my arm.”

DM: “What’s your Strength?”

Me: “Enhanced by terror.”

DM: “… okay, that sounds fair.”

Notes on APL-Based Challenge Ratings

In my experience “APL+1” is closer to being an actually ‘average’ encounter. Since the original guidelines were written, many classes have enjoyed some power inflation, some supplements make an effort to ‘denerf’ certain classes, and new classes have been added. “APL+2” encounters are often easily manageable.

Further discussion suggests that the ‘average’ encounter CR can probably scale as the levels climb. Having ‘APL+1’ as an ‘average’ encounter a low level (up to sixth), ‘APL+2’ in the middle levels (seventh through thirteenth), and ‘APL+3’ at the higher levels (fourteenth and up) might be more accurate, as more resources become available. I include this as an observation from others, but I don’t dispute it.

Note on Sandbox Encounter Design

I run a very sandboxy game. Not everything is based on or around PC levels, there are places that are definitely too dangerous for PCs at their current level.

However, I do try to provide enough information for the PCs to know this. I tend to assign CR by region, more or less (there will be variation within each region). The Dark Forest is generally CR 5: almost all the common folk fear it, but the PCs can be expected to handle it. The Forbidding Hills are mostly CR 9 or 10. The PCs might make a run in there for specific purpose and reward, but be properly cautious. The Sky-Spear Mountains are a mostly CR 15 region and the PCs stay right away unless there is an incredible need.

Not all creatures stay at home. From time to time there might be an incursion out of the Forbidding Hills, or something might come out of the Sky-Spear Mountains… but those are mostly special cases. Mostly.

Closing Comments

I find the encounter design principles that base the encounters on ‘Average Party Level’ can be challenging to apply. The math involved has many steps with intermediate rounding, and it is easy to make gross errors because of it. The effects of action economy — the ability for PCs to focus fire on small numbers of enemies — can be devastating, and the effect of powerful creatures on PCs can be similarly disastrous.

To create an ‘average’ encounter for the PCs, add about one threat per PC, of a CR equal to four less than that PC’s level. The threat can be of any kind (monsters are the easiest to consider, but other options such as traps and other hazards can be good too). Don’t assume that each PC will be targeted by one and only one of the threats, or that the threats a PC faces are the ones you expect. This is, after all, a team sport.

To increase or decrease the encounter difficulty, either increase or decrease the CR of the individual threats, or add or subtract new threats.

By design encounters like this you give yourself a great deal of flexibility in scaling. More PCs means adding more threats, rather than looking for new ones.

I did not write here about other things that can further change the encounter. Changing the encounter objective can make it easier or harder to succeed in the encounter (if you just have to get past the threat you might not need to fight at all. But not fighting is crazy talk…). I didn’t even touch on circumstances and changing the arena, the terrain and other features of the encounter location that can drastically change the results.

I also didn’t discuss multi-phase encounters… something I’ll write about for tomorrow’s post. (Near the end, as it happens; I had a lot more material to cover than I expected.)


  1. Pingback: JRPG-Inspired Encounter Design | In My Campaign - Thoughts on RPG design and play

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to Top