Balancing Combat for Length in D&D 5e
Why and how to consider how many rounds your combats should last
I was running a game of Baldur’s Gate: Descent into Avernus when I first noticed myself balancing encounters by length. One combat, I decided to track how much damage the party was doing each round, and then use that damage to determine the hit point pool for the next encounter - basically, I realized that if I gave out hit points to enemies so that the total monster HP was three times the party’s damage per round, it would take the party about three rounds to defeat that set of monsters.
It didn’t feel like any kind of great insight at the time - just another math idea in a game full of math - but since that time I’ve noticed that when I don’t balance encounters by length they tend to get a bit out of hand, and I’ve also noticed that none of the advice I’ve read on encounter balance even brings up the idea of balancing encounters by length. A cursory search of the usual sources yields the following advice on combat length: “Don’t worry about it.”
It’s sort of common knowledge that three rounds is the “standard” combat encounter, and given how long combat rounds can take in 5e, it intuitively feels like a good idea to stick to that. And if you use 5e 2014’s CR system to create encounters, it typically creates encounters that take parties of 4-6 characters designed according to 5e RAW about 3 rounds to defeat. But there’s really no mechanism other than that which constrains the lengths of combats, and because the CR system tends to just fall apart at higher levels, often the result is combats which are over too soon, or which take too long.
So let’s start by getting a feel for how players experience combats of different length.
A Taxonomy of Combat Lengths
1 round: The “flash” combat. A combat which lasts for one round or less often feels anticlimactic. Some players may have rolled initiative without even getting the chance to act. Some players may set up their combat rollout (like casting Hex, Rage, Bladesong) only to get minimal or no use out of those features. Flash combats can enhance the story, but they are rarely satisfying as combats in themselves.
2 rounds: The “quick” combat. A quick combat is generally quite satisfying. Every player acted, some more than once, and because the combat ended quickly most players will feel like they contributed to a great victory. Quick combats make the players feel powerful. Quick combats often minimize damage to PCs, which means you can sprinkle them in liberally without slowing the game down too much or draining too many resources. Players will often experience these as "easy".
3 rounds: The “standard” combat. Because this is the “default” combat length, players will be primed to feel like it is the “normal” length of combat. It can serve as the baseline by which players judge combats of other lengths. It can have a range of outcomes depending on how hard the enemies hit, and might feel easy, medium, or hard.
4 rounds: The “extended” combat. An extended combat doesn't feel long enough to call "long", and will rarely feel like a slog, but extended combats will often feel more difficult than standard combats both because they are longer than average and because enemies have had more opportunities to damage the party. Sometimes combats can be extended because enemy reinforcements arrive, players have very bad dice luck, or one or more enemies have some tactical advantage (like hiding, cover, healing magic, etc.) that keeps them alive longer than regular foes. Extended combats therefore can be exciting, but should be used sparingly - if every combat is extended, the pace of the game will start to feel slower.
5 rounds: The “long” combat. A long combat is on the verge of feeling like a slog. Often, something has gone wrong, and one or more party members are doing healing magic to bring up fallen companions, slowing down the party's damage output and dragging the combat out further. These combats very much feel like a brush with death and can easily tip into a TPK. Some players enjoy the constant threat of dying in combat, but for other players, these near-death experience combats will stop them from feeling powerful, which can take some of the fun out of the game.
In other cases, a long combat is just a case of an enemy having too many hit points (or too much AC), and in those cases the combat starts to feel repetitive. The DM will be tempted to have the bad guys surrender, or handwave the combat with "you finish them off", because it is so obvious where the combat is going that playing it out offers no suspense and no chance of failure, both of which are crucial elements in making combat fun.
Depending on how the DM paces each round, long combats can also be a challenge for some players to focus on, and you may see fatigue and disengagement. For these reasons, it's best to use long combats sparingly.
6 or more rounds: The “slog” combat, or the “special” combat. There are a couple of ways to get to 6+ round combats. One is to just have things go wrong, as in 5 round combats, but just to stay wrong for longer.
However, a longer combat could also be an intentionally epic battle - perhaps a JRPG-style boss that, once defeated, regains hit points and assumes its "final form". This may be better thought of as two 3-round combats in direct sequence.
Similarly, an enemy might run away into an adjacent room where reinforcements are waiting. If the PCs follow the enemy into that room, they could provoke a second encounter right on top of the first one. This is in some ways like two 3-round combats in a row, but this can also start to feel sloggy anyway. Often it's better to break up the two encounters (don’t have the enemy run for help, don’t have neighboring enemies hear the sounds of battle) even if it sacrifices some game realism.
Waves of enemies may also result in longer combats - you may have an encounter with one wave at round 1, a second at round 5, and a final wave at round 10. In such an encounter, if the PCs defeat the first wave in 3 rounds, they have round 4 to recover and prepare for the next wave. That round of breathing room can "reset" player attention spans so the combat ends up feeling more dynamic rather than slower, and can also give players an incentive to use more resources to defeat each wave sooner rather than later in order to get that breathing room.
Special combats can be a lot of fun, and can bring needed variety to 5e combats - but they should be special, as in, not frequent. As for slog combats, many DMs and players hate them, and it’s probably best not to plan any, with the assumption that they’ll occasionally happen by accident anyway.
Designing for Length
Considering how these combats feel, I would recommend designing almost all combat encounters to last for three rounds. The outcome of this is that players will often expend limited resources to make the combats go faster - and this is great, because trading resources for the feeling of power and satisfaction is one of the fundamental processes that makes the game fun. Occasionally players will have bad luck or use poor tactics, and the combats will go a bit longer, and this is okay because it provides some contrast and shows the party what their limitations are. Rarely, something will go wrong, and you’ll get long and sloggy combats, and sometimes you can adjust on the fly to end these if they go very poorly, or sometimes you can just ride them out.
For special fights, including staged fights and some boss fights, you may want to break the combat up into stages, each of which individually should be designed to last for three rounds, and then have a plan to string them together without overtaxing the party.
The basic principle involved in designing for length is very simple: figure out how much damage your party does in a round and multiply it by the number of rounds you want your enemies to last. If your party does 100 damage in a round, give them enemies with a total of 300hp and your combat will last about three rounds.
This isn’t a hard and fast rule. The amount of damage your party does in a round is naturally going to vary quite a bit, based on resource expenditure, luck with dice, and party strategy. If the barbarian gets downed, you lose a lot of damage output in that round - the barbarian doesn’t attack, and neither does the healer who has to cure the barbarian’s wounds!
However, estimating your party’s DPR and then using it to set enemy health will give you a good chance of, on average, having combats that last about as long as you want them to. Again, aiming for three rounds will get you a nice spread of combats lasting from one to five rounds, and you can always recalibrate your combats if you find that you’ve over- or under-estimated your players.
This principle leaves open two questions: One, how do you find out your party’s damage per round? And two, how much offensive power should the enemies have?
Determining Damage per Round
Method 1: The Empirical Method
Run a combat with the party. Keep track of how much damage the party as a whole does in each full round of the combat. Average these numbers to get the party’s average DPR. You could even do this two or three times to get a better sense of the average DPR. The advantage of this method is that it involves very little math and very little theory, and automatically accounts for things like extra damage features and use of Area of Effect spells and features to damage multiple opponents. The downside is that it’s hard to define whether any given combat is typical, and also it obviously only works if you are running an ongoing campaign with the same group. You can’t really do this for one-shots.
Method 2: Average DPR Estimation
On average, a player character should hit about 65% of the time, or, about twice as often as they miss. In a three round combat this means they’ll feel effective in a minimum of two of the three rounds. You can deviate a bit from this number, but 5e and its system of bounded accuracy is meant to nudge encounters toward this ratio. So, to estimate a character’s DPR, take the average damage per hit and multiply it by .65. A fighter that does 2d6+3 damage with a greatsword hit does 10 damage per hit on average, or 6.5 DPR. If you find this math tricky or unappealing, try a DPR calculator.
Finding the average damage per hit can be complicated. Suppose you have a Warlock. Do you count Hex? Suppose you have a Rogue. Do you count Sneak Attack? For ongoing features that increase damage, you could multiply the damage by the fraction of encounters you expect to see it in. If the Warlock uses Hex half the time, take half the damage. If the Barbarian rages in 2 out of every 3 encounters, take 2/3rds of the rage bonus. If the Rogue uses Sneak Attack on every attack, count it for the full value of the Sneak Attack damage.
What about area of effect spells and features? Do you have to account for every Fireball, every breath weapon, every Spike Growth? What about control spells, like Hypnotic Pattern?
Generally, for these types of casters, I’d just calculate their DPR based on their most common cantrip, and anything they cast that uses a spell slot doesn’t go into their DPR. This way, using limited resources like spell slots pays off by shortening combats, but not every combat can be shortened in this way. At higher levels, when you can expect casters to use at least one big damage spell slot in every combat, you can estimate their damage output using benchmark spells for their class like Fireball or Spirit Guardians.
Remember, using this method assumes that the PCs hit with 65% of their attacks and effects. You could enforce that by setting the average enemy AC equal to 8 + the PCs’ to hit bonus, or if you have an enemy with fixed AC you can individually calculate DPR against that enemy. Similarly, the average enemy saving throw bonus should be your PCs’ spell save DC - 14 if you want them to fail 65% of the time. If your enemies end up with higher defenses than your DPR calculations accounted for, it will take longer to kill them, and battles may turn sloggy. If they have lower defenses, battles will be over sooner.
Determining Enemy Offense
Here’s where I’m tempted to say “Don’t worry about it.” If you pick 3-5 enemies out of the Monster Manual whose HP add up to three times the party’s DPR, you’re bound to get a set of enemies whose offensive power is reasonable for that encounter.
However, if you want to specifically calibrate encounters to feel easy, medium, or difficult, you could use the “balancing for length” method in reverse: basically, compare the enemy DPR to the party’s HP pool. Add up the party’s total hit points, divide by the number of rounds the party should last, and allocate attacks that do that much damage per round. Divide by three for a deadly fight, where the party and enemies are evenly matched. Divide by four for a tough fight, where the party is likely to prevail, but one or two PCs might get downed. Divide by six for a medium encounter, where the party will be damaged for about half their HP by the end of the fight. Divide by twelve for an easy encounter, where the party will lose about a quarter of their HP total.
Remember, each party member that gets downed reduces the party’s DPR and lengthens the combat. The same is true for PCs who are incapacitated or otherwise taken out of the combat. Aside from prolonging the combat, a PC who can do nothing on their turn but roll a saving throw is having less fun than a PC who gets to act, so use features that take PCs out of combat - either with conditions, or by knocking them unconscious with damage - sparingly. The DMG advice to avoid enemies that can one-shot a PC is good. Try to spread attacks around, unless it makes sense for the enemies to focus fire (if they’re particularly smart and disciplined). The exception is to hit the party tank a lot to make them feel powerful by resisting a lot of hits. If you do wish to have exceptionally deadly enemies, reduce their defensive stats - if your enemies are going to be downing party members, you can reduce HP by 20-40%) to counterbalance the reduction in party damage output from the downed PCs.
Conversely, also remember that when PCs take out enemies it reduces their damage output. Parties that focus fire can easily end up taking half as much damage as parties that spread damage around evenly.
If you know how long you expect your encounters to be and how much damage you expect your monsters to do in each encounter, you can also use this data to create an “encounter budget” by assigning encounters that will damage the party a certain amount over the course of a day. Taking into account HP, hit dice, and healing magic, parties can be expected to burn through up to 200% of their HP pool in a standard adventuring day - any more than that and they may need to use consumables or extraordinary measures. Again, this is something I don’t typically worry about - I usually let players set their own encounter budget by deciding whether to leave an area and rest or whether to press on, and if it seems like they’re resting too often I’ll have something interrupt them just to keep some tension and pressure on them.
Final Thoughts
While creating this post I had an insight which may seem obvious but which I hadn’t articulated before now. There are two basic knobs on monsters to adjust combat difficulty. The DM can turn up enemy defenses (AC, saving throw bonuses, HP), which has the effect of making combat last longer and feel slower, and making PCs feel less effective. Or, the DM can turn up enemy offenses (attack bonuses, damage) which has a variable effect. Up to a point, it makes combats feel more deadly without actually drawing them out more - you can have quick yet extremely deadly combats with low defense/high offense enemies. However, after that point, when the enemies start downing PCs, the party DPR plummets and a combat that would have taken 3 rounds suddenly takes 10. Dial the offense up further and it goes in the opposite direction - instead of squeaking out a victory in 10 rounds the party gets TPKed in 3 or 4.
Or, put another way - if your combats are going by too quickly, turn up the enemy defenses. If they feel too easy, turn up the enemy offenses.
And again, nothing in here is a hard and fast rule, and you don’t need to like, make a spreadsheet to calculate your party’s DPR (although I have). For most DMs running campaigns the empirical method works just fine, and you can further fine-tune your encounters using the principles in this post without having to make calculations at all.
Encounter Map
Since these posts need some art, I thought I’d just share an encounter map I made with Inkarnate for that Baldur’s Gate: Descent into Avernus campaign I mentioned above. I call it “Stygian Shipwreck”.
A valuable treasure lies just out of reach, in the hold of a sailing ship submerged in the memory-erasing waters of the river Styx. How will the party retrieve this treasure - and what foul creatures might they disturb in the process?
If you’d like to join one of my tables, I currently have three Descent into Avernus games running, as well as a selection of other games. Check out my GM profile at startplaying.games if you’re interested!