A good encounter isn't just "enemies with the right numbers." It's a situation with enough shape that the party has real decisions to make, not just attack rolls to resolve.
Start with the fight's purpose
Before picking any monsters, decide what the encounter is for:
- A resource drain before a bigger fight (doesn't need to be hard, just needs to cost something)
- A real threat the party could plausibly lose
- A puzzle fight where the terrain or enemy behavior matters more than raw numbers
- A set piece meant to be memorable more than dangerous
The purpose changes what "well designed" means. A resource-drain encounter that's too easy is fine; a set-piece encounter that's too easy is a missed opportunity.
Use difficulty guidance as a starting point, not a contract
Difficulty budgeting systems (whether the older CR-based math or the 2024 guidance) are useful for avoiding an accidental TPK, but they're estimates, not guarantees. Action economy - how many actions the enemies get relative to the party - usually matters more than raw CR. Three weaker enemies are often more dangerous than one enemy with equivalent total CR, because they get more turns and can split the party's attention.
Design the space, not just the stat blocks
The terrain often does more work than the monsters:
- Give players cover and elevation to make positioning matter.
- Add one interactive element - a chandelier to drop, a bridge to collapse, a lever that changes the room.
- Consider enemy objectives beyond "kill the party" - protecting something, escaping with something, buying time for a ritual.
An encounter in an empty stone room is rarely as interesting as the same monsters in a room with something to interact with.
Write a short enemy behavior note
For each encounter, jot down one line per enemy (or group) describing how they actually fight: do they focus the obvious spellcaster, do they flee below half health, do they fight to the death because they're mind-controlled? This takes thirty seconds to write and saves you from making it up under pressure while five players are staring at you.
Know when to end it early
Not every combat needs to go to zero HP. If the outcome is clear and the remaining rounds would just be mechanical cleanup, it's fine to narrate the rest: the last two bandits flee, the ogre is clearly beaten and surrenders. Extending a decided fight rarely adds tension, it just adds dice rolls.
Once the encounter is built, ArcForge's combat tracker handles initiative, HP, and conditions so you can focus on running the fight instead of managing the math.