A Unified Theory of 5e Combat Design

This article is the first in a series about D&D 5e combat design theory. Subscribe to the Dragna’s Den Substack now to get future articles in your inbox!

Dungeons & Dragons 5th Edition is, at its core, a resource management game. As much as veteran players and DMs enjoy discussing tactics and strategy, the outcome of each combat is generally determined by the number of fireballs and hit points each side has available at the start of battle.

The result? If a party is “fresh” from a long and/or short rest, the first encounter they face will usually be significantly easier. This issue has led to widespread debates in the 5e community about the “standard adventuring day”—a concept suggesting that, over the course of a single day, a Dungeon Master should run enough encounters to nearly (though not entirely) exhaust the PCs’ resources.

The Dungeon Master’s Guide suggests that DMs should run six to eight encounters per adventuring day. Many DMs, however, prefer to run only one or two encounters per adventuring day, and compensate by making each encounter increasingly deadly—a decision that often creates high-variance encounters where the PCs are far more likely to outright lose.

However, in addition to the “resource-management” game, one additional factor plays a role: the “action economy.” In short: the side that can take the greatest number of actions per round usually wins. As the members of a particular side are knocked out of the battle, that side is able to muster up a decreasing amount of actions each round. In other words: a side that begins to lose tends to continue losing, creating a “death spiral.”

As a consequence, “solo boss battles”—combat encounters involving a single powerful monster—are challenging to run. If the boss is too weak, the PCs will overrun it without a significant expenditure of resources. If the boss is strong enough to survive, though, it’s likely to KO any PC in one or two hits, beginning a player-side death spiral.

Taking these two principles into consideration, the outcome of any particular combat encounter is generally determined by five factors:

• Resource Level: The fraction of their daily resources that the PCs are able to wield in that encounter, including Ability Resources (including spell slots, short-rest features like Action Surge, and resource pools like ki) and Health Resources (including hit points and hit dice).

• Party Efficiency: The rate at which the PCs can deplete the monsters’ total pool of hit points, which increases if the PCs have devoted lots of Ability Resources to offense.

• Monster Efficiency: The rate at which the monsters can deplete the party’s total pool of hit points, which increases if the PCs have few Health Resources remaining and decreases if the PCs have devoted lots of Ability Resources to defense.

• Party Lethality: The rate at which a single PC can deplete a single monster’s hit points (thereby starting a monster death spiral), considering the party’s level, the monster’s defensive CR (i.e., its AC, saving throws, and hit points), and the ratio of PCs to monsters.¹

• Monster Lethality: The rate at which a single monster can deplete a single PC’s hit points (thereby starting a PC death spiral), considering the party’s level, the monster’s offensive CR (i.e., its attack bonus, spell-save DC, and average damage), and the ratio of monsters to PCs.²

In other words:

• If you’re a PC, you want to have lots of resources available and outnumber the monsters. To win faster, you can start a monster death spiral by killing individual monsters as quickly as possible.³

• If you’re a monster, you want to outnumber the PCs. To win faster, you can start a PC death spiral by KOing individual PCs as quickly as possible.

Put together, I believe these principles represent a unified theory of 5e combat. This theory, in turn, allows us to do the following:

• It allows us to create and calibrate encounters that meaningfully challenge our players without risking TPKs.

• It allows us to understand why short rests matter, and how to structure the adventuring day around them.

• It allows us to create adventuring days that significantly drain our PCs’ resources without risking complete exhaustion.

• It allows us to have open conversations with our players about their expectations and adjust our encounters to update those expectations.

It also allows us to do a number of other fascinating things, which I’ll cover in subsequent articles in this series:

• It allows us to understand why solo boss battles fail, and how D&D 4th Edition can help us improve them.

• It allows us to understand why magical healing is usually suboptimal in combat, why hit-point regeneration is a uniquely powerful ability, and how to calibrate encounters with regenerating monsters.

• It allows us to understand why support characters can’t win without allies to take advantage of their support, why casters have to spend spell slots to stay competitive, and why fragile parties struggle more than tanky ones.

• It allows us to understand why some caster archetypes are intrinsically better than martial archetypes—even at low levels—and how modern D&D culture naturally ameliorates this.

• It gives us ways to restructure 5th Edition to preserve its spirit while replacing its resource-management core with tactics, strategy, and encounter-based storytelling.

How? Let’s dive in.

I. How to Build an Encounter

In a previous post, I proposed a new framework for evaluating the difficulty of a 5e combat encounter:

1. As per the Basic Rules, we calculate our party’s Total XP Threshold for an Easy encounter of their level. 

For example, an Easy encounter at 4th level has an XP Threshold of 125XP per PC. For a party of five PCs, that’s a Total XP Threshold of 125 XP • 5 PCs = 625 XP. 

(You can do this automatically on Kobold Fight Club.)

2. Next, as per the Basic Rules, we calculate the Total Adjusted XP for the monsters we plan to use. 

For example, a bugbear is worth 200 XP. For an encounter with five bugbears, that’s a total XP amount of 200 XP • 5 bugbears = 1,000 XP. Multiplied by the multiplier for an encounter with 3-6 monsters, we get a Total Adjusted XP amount of 1,00 XP • 2 = 2,000 XP.

(You can also do this automatically on Kobold Fight Club.)

3. Finally, we divide the Total Adjusted XP by the Total XP Threshold to calculate our Encounter Difficulty. 

For example, if we divide our Total Adjusted XP amount of 2,000 XP by our Total XP Threshold of 625 XP, we get an Encounter Difficulty of 3.2.

So what the heck does this “Encounter Difficulty” mean? As I’ve discussed previously, we can use our Encounter Difficulty to change 5e’s “Easy / Medium / Hard / Deadly” difficulty tiers into something slightly more sensible:

• An encounter with an Encounter Difficulty of 1 is a Light encounter. The PCs will need to use few to no resources to overcome it.

• An encounter with an Encounter Difficulty of 2 is a Moderate encounter. The PCs will need to use some minor amount of resources to overcome it.

• An encounter with an Encounter Difficulty of 3 is a Taxing encounter. The PCs will need to use a substantial amount of resources to overcome it.

• An encounter with an Encounter Difficulty of 4 or 5 is an Exhausting encounter. The PCs will need to nearly exhaust their currently available resources to overcome it.

• An encounter with an Encounter Difficulty of 6 or 7 is a Nightmare encounter. The PCs will need to nearly exhaust their currently available resources and take a strategic, efficient approach to overcome it.

• An encounter with an Encounter Difficulty of 8 or greater is an Impossible encounter. The PCs will be unable to overcome it at all.

These numbers may seem arbitrary—but as it turns out, they’re not. After using a spreadsheet to simulate a hypothetical combat encounter between a variable number of bugbears and a party of either five fourth-level fighters⁴ or five fourth-level bards,⁵ I found that we can use this Encounter Difficulty to roughly predict how many resources a party must expend to defeat it.

Of course, this was a simulation conducted under optimal conditions, so expect spherical cows in vacuums to abound. Nonetheless, I found that, under the given conditions—if we assume that the PCs spend approximately one-third of their spell slots and/or abilities in each encounter⁶—the ratio of the percent of resources expended to the Encounter Difficulty is always roughly the same.

The ratio in question: 6.5 percent.⁷

In other words, each time your Encounter Difficulty increases by 1, your PCs will need to spend approximately an additional 6.5 percent of their daily resources to overcome it.⁸

If we apply this ratio to our encounter difficulties, we find:

• A Light encounter will force the PCs to expend approximately 7 percent of their daily resources.

• A Moderate encounter will force the PCs to expend approximately 13 percent of their daily resources.

• A Taxing encounter will force the PCs to expend approximately 21 percent of their daily resources.

• An Exhausting encounter will force the PCs to expend approximately 26 percent of their daily resources.

• A Nightmare encounter will force the PCs to expend approximately 39 percent of their daily resources.

• An Impossible encounter will force the PCs to expend approximately 52 percent of their daily resources.

(Here, “daily resources” includes: (1) hit points, (2) spell slots, (3) short-rest features and resource pools, assuming two short rests per day, and (4) fifty percent of the PCs’ total hit dice, assuming at least one short rest per day. A PC with four 1st-level spell slots, 20 hit points, and 4 hit dice may, upon spending 25 percent of their resources, have three 1st-level slots, 15 hit points, and 3 hit dice remaining.)

II. Why Short Rests Matter

Page 84 of the Dungeon Master’s Guide suggests that every adventuring day should include two short rests—”about one-third and two-thirds of the way through the day.” Let’s assume that all members of the party regain all of their abilities and use the maximum number of hit dice possible every time they take a short rest.

Let’s call the time between two rests an “adventuring sequence.” If the PCs take two short rests a day, they’ll have three adventuring sequences:

1. one after they wake up from their long rest,

2. one after they finish their first short rest, and

3. one after they finish their second short rest.

Imagine you’re a 4th-level fighter. You get back your Action Surge and Second Wind—all of your expendable abilities—whenever you finish a short or long rest. If you get two short rests a day, you’ll be able to use your Action Surge and Second Wind three times each—once in the morning, once in the afternoon (after your first short rest), and once in the evening (after your second short rest).

During a single adventuring sequence, then, you’ll be able to use your Action Surge and Second Wind once, representing an expenditure of 33 percent of your projected daily Ability Resources. But what happens if you come face-to-face with an encounter that demands 50 percent of your projected daily Ability Resources?

You can’t use Action Surge or Second Wind a second time—each one has a strict one-per-rest policy! As such, if a single adventuring sequence asks you to spend more than 33 percent of your projected daily Ability Resources, you’ll hit a wall you just can’t climb. (You can try to compensate by dipping into your projected daily health resources—your hit points—but because you can’t convert hit dice into hit points without a short rest, you’ll quickly run out of gas there, too.)

Accordingly, we should avoid creating adventuring sequences that expect our players to spend more than 33 percent of their daily resources. Moreover, because our players can’t convert hit dice—which represent about one-third of a PC’s daily health resources—into hit points until they finish an adventuring sequence, we should err on the side of caution and try to limit our adventuring sequences to 25 percent instead.

A single adventuring sequence, then, might contain:

• Four Light encounters (4 • 7 percent = 28 percent),

• Two Moderate encounters and one Light encounter (2 • 13 percent plus 1 • 7 percent = 33 percent),

• One Taxing encounter and one Moderate encounter (1 • 21 percent plus 1 • 13 percent = 34 percent), or

• One Exhausting encounter and one Light encounter (1 • 26 percent plus 1 • 7 percent = 33 percent).

What about Nightmare and Impossible encounters?

That’s where long-rest classes come in.

A Nightmare encounter asks your PCs to spend 39 percent of their daily resources to defeat it. A short-rest class like a fighter or warlock can’t do this—but a long-rest class like a wizard or bard can by dipping into high-level spell slots to make up the difference. 

This is why Nightmare encounters require strategy and efficiency: The wizard has to recognize that they’re facing a Nightmare encounter, and immediately increase their output to make up the slack.

Why? Remember: the PCs are on a timer. If the combat goes on for too long, at least one PC will get knocked unconscious—and the death spiral begins. A wizard who immediately uses 45 percent of their own resources can usually increase the party’s damage output to compensate for the fighter’s 33 percent. However, if the wizard waits too long, they can’t ratchet their output up to 60 percent to make up for lost time.

What about Impossible encounters? An Impossible encounter asks your party to spend 52 percent of their daily resources to overcome it. If our hardworking wizard wants to compensate for the fighter’s 33 percent expenditure, they'll need to increase their per-encounter output to 71 percent of their own daily resources.

Now, that sounds doable—in theory. But in practice, if you’re a fifth-level wizard with two 3rd-level slots, three 2nd-level slots, and four 1st-level slots, it’ll take you five rounds to get there—and by then, a death spiral is very likely to have already begun.

Even if you’re a party composed entirely of long-rest casters, an Impossible encounter is still nearly insurmountable. A long-rest caster regains all of their spell slots and hit points on a long rest, but still requires a short rest to convert hit dice into hit points and regain minor short-rest features (e.g., a cleric’s channel divinity or a wizard’s arcane recovery). Even a long-rest caster, then, is a short-rest character in disguise.

Imagine a character who can convert hit dice into hit points as a free action, who can take any number of actions and/or bonus actions per turn, and who can use their rechargeable abilities three times per long rest instead of once per short rest. A party composed entirely of such characters would, hypothetically, be able to tackle an Impossible encounter.

Unfortunately, no such characters exist.

III. How to Build an Adventuring Day

Let’s sum up what we’ve learned so far:

• We can calculate a combat’s Encounter Difficulty by dividing its Total Adjusted XP by the party’s Total XP Threshold.

• We can multiply this Encounter Difficulty by 6.5 percent to calculate the rough percentage of our party’s daily resources that it will consume.

• We should avoid running Nightmare encounters without providing our players with the ability and incentive to properly strategize and prepare, and we should avoid running Impossible encounters altogether.

Moreover, because 5e combat is far messier than the ideal circumstances we’ve worked with so far, we should give ourselves a buffer, just to be safe. We also want to make sure, however, that every adventuring day meaningfully challenges our players. Accordingly, we should try to make sure that every adventuring day requires our players to expend at least 40 percent of their daily resources and no more than 80 percent.

Additionally, we should structure our encounters to avoid requiring our players to expend more than 25 percent of their daily resources on any given adventuring sequence within that day. If you want to run an adventuring day that consumes 75 percent of your players’ resources, you had better make sure they can easily access at least two short rests!

Notably, players are naturally risk-averse, and will therefore always attempt to take a short rest whenever doing so is reasonably possible. An “adventuring sequence", therefore, is a “series of encounters in which taking a short rest is not an option.”

For example, an adventuring sequence might be defined by:

1. A lack of time (e.g., because the party is racing to rescue a prisoner before their execution);

2. A lack of space (e.g., because the adventuring sequence involves multiple waves of monsters in the same area); or

3. A lack of safety (e.g., because the party cannot remain in place without being discovered or attacked).

Imagine a quest that promises the PCs a substantial reward if they defeat the monsters in the quest-giver’s house at any point within the next week. If there are monsters in the kitchen, monsters in the living room, and monsters in the foyer, you might think that the act of clearing out the house always constitutes a single adventuring sequence—and you’d be wrong.

If the monsters in the kitchen and living room patrol the foyer once every fifteen minutes (and vice-versa), it’s true that we’ve compressed our encounters into a single adventuring sequence. But if the monsters in the kitchen never leave the kitchen, the monsters in the living room never leave the living room, and the monsters in the foyer never leave the foyer, you don’t have one adventuring sequence—you have three.

Short rests naturally expand to fill any narrative beat without tension or stakes. If, after killing the monsters in the kitchen, there’s no danger to remaining in the kitchen and no reason to immediately leave it, your PCs will take a short rest almost every time. Good combat design, therefore, naturally incorporates narrative tension in order to ensure that the party takes short rests only when they’re supposed to.

IV. Why Player Expectations Matter

Let’s take a moment to flash back to the problem with Nightmare encounters: If our players don’t realize quickly enough that they’re in a Nightmare-level fight, they’ll enter a death spiral before they can spend enough resources to make up for it.

What about the inverse? What happens if a player thinks they’re in a Nightmare encounter, but they’re really in a Moderate or Light one?

Then, we’d expect to see that player go “nova”—digging deep into their resource piggy bank and spending way more than they can afford to. Instead of spending 10 percent of their daily resources on this low-stakes encounter, they’ve spent 60 percent—and they’re rapidly running out of gas.

This is the danger of player expectations. A player who fails to realize an encounter’s lethality will fail to spend enough resources. A player who fails to realize an encounter’s simplicity will spend too many resources.

In the first case, if I’ve prepared a three-encounter adventuring day and my players treat it like a six-encounter adventuring day, then they’re liable to fritter away precious time casting cantrips instead of fireballs and saving Action Surges for an emergency that will never arrive. For martial classes, this isn’t a huge problem. However, because killing monsters sooner stops them from dealing damage later, any caster class that’s capable of “going nova” is better off casting their high-leveled spells first and their cantrips second.⁹

In the second case, if I’ve prepared a three-encounter adventuring day, and my players treat it like a one-encounter adventuring day, then they’re liable to blow most of their resources on the first encounter, spend the remainder on the second, and show up to the third encounter—probably the climactic finale—without any gas left in the tank. This is a real problem, and it affects more groups than you’d think.

Why? If a DM sets an expectation that most sessions will include only a single Nightmare encounter, then the players can’t treat any encounter like a Moderate, Taxing, or Exhausting encounter without risking total defeat. If the DM decides to switch things up and run an adventuring day with five Moderate encounters instead, the players are exceptionally likely to expend all of their resources on the early encounters and have nothing left for the later ones. 

By contrast, if the DM starts by running an adventuring day with six to eight encounters per day, and then suddenly stages a session with a single Nightmare encounter, it’s exceptionally likely that the players will fail to switch gears in time.

In both cases, this happens because the players didn’t know that the DM had changed the rules. If you’ve been taught so far that every encounter requires your maximum effort, you’re going to be left out in the cold when the DM reveals that the session’s first fight was just a warm-up.

Fortunately, there are ways to set and update player expectations. Setting them is easy: During Session Zero or at some other time between sessions, tell your players what kind of adventuring day you prefer. Are you going to prepare lots of monster-infested dungeon crawls? Or do you prefer single-combat adventuring days with isolated, flashy set-pieces?

Frequently, though, you’ll wind up in a campaign with lots of both, making it difficult for the players to know what to expect. Fortunately, DMs have many tools available to update their players’ expectations at the start of a particular encounter:

• Monster Size. The bigger the monster, the tougher it is. While not always true, players will almost always assume that a Small or Tiny monster is relatively weak, and a Large or Huge monster is relatively strong. (This continues even at higher tiers, though players might upgrade their expectations and assume that Small or Medium monsters are relatively weak and Huge or Gargantuan monsters are relatively strong.)

• Monster Notoriety. A named monster is more likely to be powerful than an anonymous one. An encounter with a regional boss monster like King Grol, lord of the Cragmaw goblins, is very likely to be Taxing or Exhausting. An encounter with a campaign-defining monster like the Tarrasque is almost certain to be a Nightmare. 

• Monster Behavior. In general, players assume that monsters that feel threatening are more likely to actually be threatening. A revenant knight who swears undying vengeance upon the PCs or a slavering ghast intent on devouring their flesh will likely feel more threatening than a cackling goblin.

• Environmental Posture. The less mundane an environment feels, the more threatened the PCs will feel. By contrast, the more mundane an environment feels, the safer they’ll feel. A battle within a wine merchant’s shop feels far less dangerous than a battle in a dormant volcanic caldera, even if the monsters are exactly the same.

• Encounter Complexity. In general, the PCs’ lives are the only thing at stake in any given encounter, and the only challenge they face is reducing the enemy’s hit points to zero. However, you can make an encounter feel more significant—and therefore more dangerous—by adding additional combat mechanics or narrative stakes. If the PCs have been tasked with rescuing a hostage, defending a target, or completing a puzzle, they now face two tasks (e.g., surviving and rescuing) instead of just one. As a result, they’ll naturally work harder—especially if they’re also facing a round-by-round countdown.

In Conclusion

For an eight-year-old role-playing game, 5th Edition has an astonishingly underdeveloped metagame. While resources like The Monsters Know have attempted to improve DMs’ access to tactics and strategy, few contributors have taken aim at the resource-management concept that lies at 5e’s core.

Instead, the vast majority of DM creativity has been spent on refining and advancing the content of combat. What subclasses or spells will the players choose, and what magic items will they find? What monster statblocks will the DM create, and what dungeons will they live in?

It’s telling, I think, that the bulk of player-side energy has focused on the concept of “build optimization.” In its own way, the marketplace of ideas on the players’ side of the DM screen has organically created a natural, if clumsy, complement to the ideas I’ve tried to lay out here. 

It was players, after all, and not DMs, who pioneered the concept of using healing word to soak up monster damage. It was players, and not DMs, who realized that it’s often better cast hypnotic pattern than fireball. It was players, and not DMs, who first realized that pairing a hexblade’s armor proficiencies with a sorcerer’s damage output maximizes Player Efficiency while minimizing Monster Efficiency—though not, of course, in as many words. (I plan to discuss the twin concepts of party composition and character-build optimization in a future article in this series.)

Time and time again, however, the past eight years of gameplay have highlighted a DMing community fraught with anxiety—over creating encounters that are too easy, encounters that are too hard, or adventuring days that are worth less than the sum of their parts. This imbalance has likely frustrated players, as well. After all, when facing an uncertain DM, an optimized build can feel less like “innovating” and more like “bullying.”

The historic refrain of 5th Edition DMs is simple: “Encounter-building is an art, not a science.” But this belief is misguided. Encounter-building is a science. It’s just been missing the proper tools.

Those tools, I believe, are here.

---

This article is the first in a series on 5th Edition combat design theory. Future posts will discuss problems with solo boss-monster encounters (and how to fix them), the relationship between magical healing and combat outcomes, the impact of party composition and character-build optimization, and potential paths to removing 5e’s reliance on the resource-management game. Thank you for reading, and make sure to subscribe to receive future articles in your inbox!

---

Special thanks to Twi, PunchingPotato, Booyagh, Flutes, vanguard2k, Humanfarmerman, and Bjarke the Bard for their feedback and suggestions regarding this article. Cover image credit to Wizards of the Coast.

DragnaCarta is a veteran DM, the author of the popular “Curse of Strahd: Reloaded” campaign guide, a guest writer for FlutesLoot.com, and the former Dungeon Master for the actual play series “Curse of Strahd: Twice Bitten.”

To support Dragna’s work and access perks like DMing workshops, homebrew content, personal campaign advice, and a supportive Discord community, you can click here to join his Patreon.

To get Dragna’s hot takes and commentary, you can click here to follow him on Twitter.

1. If there are more PCs than monsters, some PCs will have to attack the same monster, making it easier to start a monster death spiral. If there are fewer PCs than monsters, some monsters won't get attacked at all, making it harder to continue a monster death spiral once one begins.

2. The same principle as Footnote 1 applies, but in reverse.

3. When it comes to death spirals, I’m a fan of drawing a comparison to Orson Scott Card’s novel Ender’s Game. In Ender’s Game, soldiers can win battles by seizing the enemy’s headquarters. Traditionally, the act of seizing the enemy’s headquarters is a ceremonial ritual completed once all enemies have been defeated. However, the protagonist of the book realizes that, because seizing the enemy’s headquarters results in victory, he can bypass the actual “battle” by immediately seizing the enemy’s headquarters instead.

Death spirals in 5e are exactly the same—each encounter effectively ends when a death spiral begins. A party or monster can ignore the “battle” and win the encounter by focus-firing individual enemies as quickly as possible, thereby triggering a death spiral for the opposing side.

4. Here, I used the Clockwork Mod Dice Calculator to calculate the average damage each combatant would deal per turn, given the following assumptions:

1. Each fighter has +4 to STR and +2 to CON.

2. Each fighter uses a one-handed longsword, chain mail, and a shield.

3. Each fighter has chosen the Champion subclass and the Defense fighting style.

4. All attacks occur simultaneously.

5. Each fighter and each bugbear attacks a single other bugbear or fighter, respectively, with no focused fire.

6. Each fighter uses their Second Wind and Action Surge on the second or third round of combat.

5. Here, I again I used the Clockwork Mod Dice Calculator and made the following assumptions:

1. Each bard has +4 to CHA, +1 to CON, and +2 to DEX.

2. Each bard uses studded leather armor.

3. Each bard has chosen the College of Eloquence subclass.

4. Each bard casts shatter, thunderwave, and thunderclap sequentially, hitting an average number of targets given by p. 249 of the Dungeon Master’s Guide.

5. All attacks and spells occur simultaneously.

6. Both sides attempt to spread out the damage as much as possible, with no focused fire.

6. This actually matters a lot, because it turns out that Health Resources and Ability Resources aren’t created equal!

Imagine a 5th-level wizard with 25 hit points at AC 12 with a single fireball spell (representing around 17 percent of their ability resources, including arcane recovery) fighting a single bugbear.

If the wizard only casts fire bolt (+7 to hit, 2d10 fire damage), they’ll deal an average of 7 points of damage every turn, which will take approximately four rounds to deplete the bugbear’s 27 hit points. Assuming both sides act simultaneously, the bugbear also gets to attack four times, dealing an average total of 29 points of damage to the wizard before it dies, representing approximately 60 percent of the wizard’s total Health Resources.

However, if the wizard casts fireball first (DC 15, 8d6 fire damage) and switches to fire bolt afterward, they’ll deal an average of 22 damage on the first round and an average of 7 damage on the second round, which is enough to kill the bugbear. During this time, the bugbear only gets to attack twice, dealing an average of 14 points of damage, or 30 percent of the wizard’s total Health Resources.

The takeaway: by spending only 17 percent of their Ability Resources, the wizard has saved a staggering 30 percent of their Health Resources!

7. This number is actually the approximate average of two other numbers: 7.16 (for the fighter simulation) and 5.28 (for the bard simulation).

These two numbers diverge because the simulated bards spent all of their high-impact spell slots at the beginning of the encounter, ending the fight before too many hit points could be lost. In contrast, the fighters lost a steady amount of hit points each turn, and the impact of their early Action Surge was much smaller than that created by the bards’ shatter and thunderwave spells.

8. Because fragile party members can make death spirals too easy to trigger, this formula can break down when used for a party full of “glass cannons.” To compensate, find the fraction of the party that both:

• has a d6 hit die or a negative Constitution modifier, and

• is unable to raise their AC above 10 plus their proficiency bonus.

Then, multiply this result by two and add it to the Encounter Modifier.

9.  If they cast their cantrips first and their high-leveled spells later, there’s a high probability that their high-leveled spells will “overkill” the enemy, wasting a huge amount of excess damage. Plus, as discussed in Footnote 6, a player who refuses to spend Ability Resources inevitably winds up spending a disproportionate amount of Health Resources instead.

The best player-side strategy, then, is to cast high-leveled spells until your enemies are at or below half-health, and then either:

• continue casting high-leveled spells if you’re also at or below half health, or

• switch to cantrips otherwise.