D&D5e – Thoughts On Balancing Encounters

Last updated on August 25, 2017

In a recent video, prominent YouTuber and Dungeon Mastering evangelist Matt Colville (Twitter: @mattcolville) suggested that new Dungeon Masters should likely not use the DnD5e “Challenge Rating”, or CR, system as it was designed.

The video is here: https://youtu.be/xANZTijbrw8

I like Matt Colville’s video series, titled “Running the Game“, an awful lot. He has some smart insights. In many cases, he is articulating stuff I have always done, or have had tumbling in the back of my head but do not have words for. In some other cases, though, he says something I Spock-Eye at him about.

My take-away from this particular video was that he suggested Dungeon Masters use the “Challenge Rating”, or CR, system as a guideline to find a group of likely monsters to fight in the encounter, but do not use the suggested budgeting system. Instead, he advocated a sort of holistic “in the moment” decision, based on what amounted to gut instinct and experience, to dynamically balance the encounter at the moment the initiative dice are rolled.

My son and I have a running joke between us. It roughly amounts to:

Me: “You need to learn to make good descisions.”
Him: “How do I do that?”
Me: “By gaining experience.”
Him: “How do I do that?”
Me: “By making bad decisions.”

It is a paradox. When we are gaming, as Dungeon Masters, we are at least responsible for part of the “fun” of the night. We describe the world, and present its problems and challenges to the players.

For me at least, part of a good adventure is dramatic uncertainty. If we know the characters are going to triumph without concern or adversity there is not a lot of dramatic uncertainty. In other words, it is a boring story.

I have told players for games I have run in the past, either Dungeons & Dragons, Cyberpunk, Traveller, or whatever, that if they seem to be wading through a faceless horde of opponents that die by the half-dozen or so per attack the players make, that is the first hint the villian is just trying to waste their resources. Time is a resource, as much as potions, spell slots, shells in the automatic grenade launcher, or armor plates on the hull.

In other words, the only time it should not matter to the players that they are tied up in a battle, is if it matters to the villain that they are tied up in a battle.

So, if the intention is to create exciting and memorable play via uncertainty of outcomes — that is, dramatic uncertainty — then as Dungeon Masters (DMs) we collectively need some way to know what is too much or too little. My first reaction to Matt Colville’s advocacy of a holistic approach to encounter management was “for DMs who have been playing a long time, maybe.”

For new DMs who lack the experience with the system, I can see a lot of lousy encounters in danger of happening. No DM likes the sight of a bored group. At the same time, when a supposedly harrowing encounter suddenly turns into either a rout or massacre of the party, that can be no fun either.  Particularly for a new GM and new Players.

I, for one, do not like a D&D experience that feels like pen-and-paper Dark Souls.

So, is there a middle ground?  If we agree with Matt’s idea to not use the CR system, is there something that inexperienced DMs can use to calibrate against?  Something that will still keep what feels like a more holistic or gut-based approach to tuning an adventure?

Indirectly, there is. It’s the implicit behaviour of the rules, if you pry up enough floor boards.  It still needs some math, but it is not horrible. You can do most of it with pencil and paper.

First, some postulations:

1) We want to reduce the party, collectively to 1/3 of their starting health by the end of the fight. Forget maximum hit points, the question is where are they right NOW, when you are about to descibe the encounter just before rolling initiative.
1a) Hateful dice or bad decsions by the players might result in this margin being significantly eroded.
1b) Half the party hitting the floor by the end of a battle is acceptable & dramatic.

2) We want the party to reduce the monster collective health pool to zero.
2a) Consider a “bail out” threshold of 70% casualties for the monsters. Most of them are evil, and would prefer to live to fight another day.

3) Players hitting stuff is fun.
3a) Players getting hit by stuff adds to dramatic uncertainty.

4) Saving throws introduce chaos.

So, our party is made of four characters, 3rd level.

Able, AC 21 fighter, 21hp, avg 8hp per attack @ +7
Baker, AC 18 cleric, 18hp, avg 7hp per attack @ +4
Charlie, AC 18 rogue, 18hp, avg 10hp per attack @ +4
Danny, AC 14 mage, 15hp, avg 12hp per attack @ +6

Note the HP numbers are not the maxium for the character, they are the number at the start of the fight.

So now we can crunch some numbers, and make some pools. The group average AC is 18, the total health pool of the group is 72hp, average burst damage is 37hp, and average to hit bonus is +5.

Eight rounds of combat is a solid fight. A knock-down-and-drag-out seige is 16 rounds. What are you looking for? Lets go with 8 for this excercise; double the monster health pool and reduce monster damage per round by 35% for 16 rounds.

8 rounds times 37hp burst is a enemy health pool of 296hp. Now, factor enemy average AC. Since they have an average +5 to hit, they can generally hit AC 15 (10 + 5) half the time.

• If the players only hit 25% of the time, then cut the health pool by 75%; AC 20, 74hp.
• If the players only hit 50% of the time, then cut the health pool by 50%; AC 15, 148hp.
• If the players can hit 75% of the time, then cut the health pool by 25%; AC 10, 222hp.
• If the players can hit 100% of the time, keep the health pool as is; AC 5, 296hp.

Note that the first example may annoy the players because hitting stuff is fun. The last example may bore the players because they cannot miss.

Now, take the health pool of the party, 72hp, and multiply that by 0.67. To bring the group to 1/3 their starting health the monsters need to do 48hp. Divide that by 8 rounds, and you get an average damage of 6hp per round.  Its a low-level group, so it is low-level damage.

Lastly, take the average AC of the group, and subtract 10. 18 – 10 = 8. The monsters will need at least a +8 to hit to get 50% average damage done. A 50% hit rate feels kind of thrills and spills, so double the average damage per round to compensate for the misses.

So, you need a group of monsters that have a total of 222hp, AC15, does 12hp average damage per round, and has +8 to hit. Any ability the monsters has that will do more than 1/2 the HP of the lowest HP character should have some kind of saving throw. So, now, take a look at the monsters you planned to use, and compare. Adjust the monster head count based on the health and average damage of the creatures.

Many smaller monsters with many attacks is more dangerous than fewer big monsters with fewer total attacks. Because of the way D&D5e works, if there are more attacks than there are players, then things might go worse than expected.

Thematically, it feels more heroic to fight two Slag Giants that hurl anvils and swing railway ties than being nibbled to death by a slavering flock of Dire Ducks.

The fight will last an average of 8 rounds, depending on dice rolls. The players will take beating, but the enemy should run out of health shortly before the players do.

If you want a short, brutal fight, half the monster health pool and double the damage per round.  Warning;  a character will get laid out flat in one hit in this kind of fight. The characters will almost certainly bring out the howitzers and health potions.

Healing and Damage Reduction both do the same thing … they extend the length of the fight by effectively increasing the monster’s health pool.  Damage reduction worth 10% of the party’s average burst per round means the fight will last 10% longer.  By extension, a heal equal to 80% of the party’s average burst per round will add 10% onto the length of the fight.

Keep in mind that this in turn will add 10% onto the damage the party takes, total, because the fight lasts that much longer.  You only have a 30% margin, before acts of hateful dice, before you kill a character.

Note that most of these numbers rarely changes; the average AC of the group rarely changes. The average burst damage of the group rarely changes. The average bonus to hit rarely changes.

The only question is what the players current health is.

I can hear some readers saying “Hang on, Michel, that’s just a different CR system. I still have to do math.”

Correct. Matt Colville is doing math, too. He is essentially doing what I just spelled out to you. The difference is that he has seen enough fights and players at a wide variety of levels — keep in mind, he runs essentially the same set of modules and setting for each campaign, so he sees a lot of the same content — that a lot of the numbers are already “worked out” in his head.

In addition, I suspect his mental calculus is not “damage to 1/3” but “damage to zero”;  IE; he isn’t worried about “stopping the fight” before all the characters are nose down and making death saves.  My version is, because we are trying to calibrate for new DMs and new Players.  As you play through a dozen or more encounters, and you learn your players and their characters, you can push the margins of error more and more into dangerous territory.

That is what experience as a GM gives you. “Oh, I remember the last time a 3rd level group ran into a couple of Ogres … so, I’ll give one Ogre a Bracer of The Healing Waves, four charges for 3d4 each, and he will use one charge in the fight. He will use it when he drops to half health or the first time he takes a critical hit. That will also make some good loot for the party.”

For the new DM, however — or even DMs like myself who have been running games for 30+ years, some times having start position to calibrate from can be helpful.

The difference between the official challenge rating system verses the one I suggest here is when you do the work. The night before or the minute before. Both have their merits and minuses.

As Matt Colville says, there is no “right” answer. There is just the set of answers that works best for you.

You can build the system I described as a simple spreadsheet, by the way. I have included a link to an example one at the end of this blog post.

Alternate CR Calculator << download and try