Are the rules in the DMG for calculating encounter difficulty inaccurate?

Question

I recently DM'd a module. It seems the characters were easily able to get through encounters that the DMG considered “Deadly,” even without factoring in the “encounter multipliers” for multiple monsters. After the first encounter, I realized the party seemed to be beyond the “very strong” party indicated by the module, so I beefed up the encounters. After running the module, and calculating the XP, I was incredulous as to how the party got through it all rather unscathed, so I double checked the DMG encounter experience rules.

Death Giant

Huge Undead

CR 16,
253 HP AC 18
DC 18
Immunities: necrotic, (fire or as per base giant)
Str 32(+12), Dex 13 (+1), Con 20 (+6), Int 12 (+1), Wis 20 (+5), Cha 25 (+7)

3 x +10 (4d6+18) sword attacks
Circle of Death (recharge 5-6, can be used in lieu of one attack on same round),
Unholy Aura, giving them AC 22 (concentration), takes action to cast.
Flame Strike, can be use in lew of 1 attack on same round
Reaction: DC 18 Con or 1d6 constitution drain to melee attacker that hits it

*If you go down to 0 hp within 10 ft of Death Giant, you instantly die, and require Resurrection, not Raise Dead.

This was more or less based off suggested stats for CR 16 creatures in DMG—offense was cr 19 (120 avg dmg over 3 rounds), and defense was 13, average=CR 16.

(Update: After playtest, this monster is great btw. For a stronger party, or you just want to use one or two, consider adding 50hp or +2 attack and still call it cr 16. )

The party consisted of:

• 16 Barbarian
• 16 Paladin
• 16 Barbarian
• 16 Cleric
• 13 Rogue
• 13 Paladin

According to the DMG, I calculated the encounter XP threshholds as:

• Easy: 8600 = 1600+1600+1600+1600+1100+1100

• Medium: 17200 = 3200+3200+3200+3200+2200+2200

• Hard: 26000=4800+4800+4800+4800+3400+3400

• Deadly: 39000=7200+7200+7200+7200+5100+5100

• Long Rest Threshold: 64000=16000+16000+16000+16000+13,500+13,500

The Combats

The cleric casted Heroe's Feast previous, so they all had advantage on Wisdom saves, which are my #1 way to beat the party, with Mass Suggestion, Hypnotic Pattern, etc. But they had no problems getting DC 20 saves with their advantage.

So the party killed 3 of the death giants, with two 15 cleric support enemies, a 12 Wizard casting Fireballs. They underestimated the 30 skeletal archers the first two rounds so that got me in 150 dmg there.

Then 3 death giants with three support casters and a supposed "legendary" vampire.

In the next combat, versus “legendary” daemon, I just couldn't do enough. I added a Goristo, a pair of Glebrazu, and several vrocks.

Eventually I had the gnome archmage come in flying the Chronotyrn (18th level sorcerer/timestop casting bird that can cast and concentrate on two spells at a time), and casted Time Stop --> Mirror Image, Force Cage (I could get two of them), and then Prismatic Wall (as a sphere), but as they came out of the sphere to cast their spells (and retreat back in), I realized both enemies would be dead in a round or two(from readied actions), and they already blew their slots. Meteor Storm seemed not worth it. The chronotyryn gated in an Elder Storm Elemental which did a 8d6 thunder + 21d6 lightning 60' blast. Most people took 1/4 damage.

Then the cleric made his divine intervention roll on one shot, and I allowed the remaining demons to be banished(so just the wizards and elemental remaining.)

The fourth combat against a CR 24 daemon lord was admittedly cheesed somewhat, as I rushed it.

Observations

The PC’s had resistance to just about everything, advantage on just about everything, immune to poison and fear, some had resistance to all spells, resistance to fire and cold, two had cloaks of displacement (although not much use against the multiple attacks of the giants, they certainly helped.)

So totaling up the string of 4 encounters, with no rests at all, the XP total was 346,400.

So I am aware of the following mistakes on my side:

• I should have had the wizard use his Wish to dispel the Heroes’ Feast.
• I should have made use of the spell Maze, as everyone in the party had low intelligence.
• I should have added Dispel Magic and “Cast a spell” to the legendary action options of all enemies, and will do so from now on.
• To mess up the PC’s readied actions for the gnome and Chronotyryn coming out of Prismatic Sphere to cast a spell and retreat, I should have used Major Image or Programmed Illusion, to create an illusion of the gnome coming out of the sphere.
• I should give all higher level casters several 8th and 9th level scrolls—it would make sense in game that 18th level wizards would have such scrolls.
• I should have added a new spell to the game, Mass Dispel, which does a Dispel Magic on everyone in a 30’ radius: (did 5e really get rid of any way to cast a Mass Dispel?--I had two wizards cast dispel on them, but just so happens randomly it was the 13th level ones).
• I should have realized casters are pretty nerfed in 5e, they aren't the challenge I thought from previous editions,
• Use sorcerer's with scrolls instead of wizards(for the metamagic and damage potential.)

Initial Misconceptions about encounter difficulty

I also realize my initial misconception: Before I read this section of the DMG, my assumption was a “Deadly” encounter had the probability of TPK, but here it just says it “could be” lethal for a couple players.

Hard. A hard encounter could go badly for the adventurers. Weaker characters might get taken out of the fight, and there's a slim chance that one or more characters might die.

Deadly. A deadly encounter could be lethal for one or more player characters. Survival often requires good tactics and quick thinking, and the party risks defeat.

The two 13th level players did go down multiple times, requiring a raise dead and resurrection from the cleric, but only once did one 16th barbarian reach 0 hp.

If I raised the DC's arbitrarily, they will complain. If I just had 7 5th level wizards coming in and casting Dispel Magic, it would be cheesy. In Pathfinder, I could get away with nudging up enemy DC's a number of ways, increasing size categories, etc., but in 5e it is like sacrosanct not to do that. In Pathfinder the wizard could have had a number of spells going at once.

In my mind, these were four “hard” encounters. According to the DMG, these were four encounters way beyond “deadly (as per DMG definition),” in fact each encounter was beyond the parties “adventuring day” XP.

Questions I ask myself are: Was I doing something wrong, and not playing the monsters to their potential, was the party way beyond their abilities, or was it the fact the party was not balanced and had mostly melee classes, or is there something wrong with the DMG encounter calculations? Should I not have allowed the cleric to cast the Heroes' Feast from the get-go? Did the 5e idea of a "deadly" encounter change from previous editions?

Question

Core Question: Given that the PC's gained 346,400 (5x adventuring day, and close to 10x a deadly encounter), each encounter was twice the deadly XP, and 50% more than the adventuring day xp, is there something wrong with the calculation method in the DMG?

Answer 1

Reasons why it was easy

There are a number of reasons why your encounters were a lot easier than you might have expected. Generally, these reflect deviations from the DMG's encounter guidelines.

1. Party Size: The DMG's guidelines assume a party of 3-4 adventurers. You have 6 characters, which is almost double the recommended number. For a single monster fight, the fight is literally half as difficult (DMG 83):

   If the party contains six or more characters, use the next lowest multiplier on the table. Use a multiplier of 0.5 for a single monster.

2. Action Economy: One of the biggest deals/issues/things with encounters in 5e is that the action economy is super important. It looks like most of your fights only involved a handful of monsters. Especially with a large group, small numbers of monsters are easily overwhelmed. I think it's telling that your group of 30 skeletal archers were actually fairly effective, even though they were so outmatched by CR. In my experience, fights are more balanced if the number of enemies is similar to the players. In particular, more enemies means that powerful single-target spells are weaker. Note that the DMG tells you to ignore relatively weak monsters (DMG 82):

   When making this calculation, don't count any monsters whose challenge rating is significantly below the average challenge rating of the other monsters in the group unless you think the weak monsters significantly contribute to the difficulty of the encounter.

3. PC-classed NPCs don't hold up: If you look at the end of the MM and at VGtM, you'll see NPCs that are statted out like PCs, but are significantly different. Generally, they have more HP and don't hit as hard. This is because enemies built like PCs tend to be glass cannons, and they will go down pretty quickly. A general rule of thumb is that PC-classed NPCs have a CR that's about half of their level.

4. Number of encounters: The DMG (pg. 85) assumes that parties will face 6-8 encounters a day:

   Assuming typical adventuring conditions and average luck, most adventuring parties can handle about six to eight medium or hard encounters in a day.

   It looks like your players faced off against 4 encounters that day. That's only half the recommended maximum.

5. HP isn't the only resource: HP is a useful metric for determining whether a party is hurting, but it's not the only one. What about spells/abilities/healing expended? Perhaps if they had fought a few more encounters that day, they would have taken a lot more damage.

Making it Harder

I don't really have an answer to a lot of these things; I too struggle with encounters that are too easy. Still, there are some things you can do:

1. Use the DMG's guidelines: DMG 85 has suggestions for making an encounter more difficult, which mostly boil down to unfavorable terrain and difficult conditions. Adding extra externalities is always a good way to spice up an encounter--you don't mention any of these in your question.

2. More Actions: The MM's "boss monsters" have legendary actions, which allow them to act multiple times per turn. This technique is there in order to address some of the action economy issues, which are discussed above.

3. Smarter enemies: You already address this in your question, so I won't say more.

Was it actually too easy?

PCs generally win their fights. After all, if they all die, then the campaign is either over or completely different. It seems like if you had the level 13 characters go down multiple times and the level 16 barbarian go down once, the fights were appropriately difficult. You say that your players really enjoyed the session, and that they felt like the fights were difficult, so it sounds like you did a good job, and accomplished your goal!

Conclusions

In conclusion, it looks like your day didn't quite follow the DMG. The DMG seems to favor longer adventuring days that grind at resources, whereas you had a few large fights. Moreover, the guidelines are just that--5e gives a lot of leeway to the DM for adjusting the game.

Still, in my experience, the DMG's guidelines are indeed quite lenient. I have also had the same issues that you've had, with my players destroying encounters that were designed to be overwhelmingly difficult, and ending up with a lot more XP than the DMG says that they should. I haven't been able to find any developer commentary on intended difficulty, so I'll just say that you're not the only one who feels like they're too easy.

Answer 2

Combat is won by tactics, not CR

There are no guidelines for an encounter going from Deadly to TPK. So even at thrice the Deadly threshold, your party can still survive. It depends on the decisions of the party during combat, and the counter-tactics of the DM/monsters -- and not the XP threshold -- as to if the PCs will win the encounter.

A closer look

If I am reverse-engineering your encounters correctly, then

• Your first encounter presented the party with an XP budget of 146,400.

• Your second encounter was at an XP budget of 82,500.

• Your fourth encounter, at 31,000.

• Your third encounter was at 86,500.

  • You didn't specify how many Vrocks you put in or the CR of the Gnome Archmage, but the total XP you presented minus the XP budgets of the 1st, 2nd, and 4th encounters leaves us with this number.

The first encounter is when the party is strongest

You gave your PCs their toughest fight when they had their full resources. Even if this is ~5x the Deadly threshold, because your players are at a high enough level, they can weather this storm without suffering a TPK.

As you yourself noted, a Deadly encounter does not equate to a chance of a TPK.

Deadly. A deadly encounter could be lethal for one or more player characters. Survival often requires good tactics and quick thinking, and the party risks defeat.

In general, the higher level the party, the more resilient they are while at full resources against encounters that are far past their Deadly threshold.

Your second and third encounters were half as difficult as the first

The XP budgets of both the 2nd and 3rd fights fell very drastically from the first one. This means the party had twice as much breathing room in these battles. And again, even if the fights were at ~2x their Deadly threshold, this does not guarantee a TPK. It only means there will be a party death, and the PCs may lose if they do not use good tactics.

Divine Intervention was too powerful in your use case

Divine Intervention can do whatever the DM wants, but it is most appropriate to replicate a high level Cleric spell with it. Banishment of all the demons with this use has drastically adjusted the XP budget of your third encounter, removing at least 52,350 XP from the budget without spending resources (1 Goristro, 2 Glabrezus, 3 Vrocks).

The Gnome Archmage was effectively at Offensive CR 0

In terms of the actual game, it seems like your Gnome Archmage did not deal significant damage by himself. This means his Offensive CR, whatever it was in theory, was reduced to 0 in practice.

CR is fluid: theoretical CR does not equate to practical CR. For example, the damage calculations of a Goristro is based on the idea that it will take the most damaging options on all of its first three rounds of combat, which is the Gore attack plus Charge attack. This is the only way it reaches Offensive CR 13. If in its first three turns, it fails to do this, it is not able to perform up to its expected Offensive CR. And so in practice, it had a lower Offensive CR -- and thus, a lower overall CR -- after all.

And so the same can be said for your Gnome Archmage. Having failed to deal significant damage to the party, his effective Offensive CR in practice was 0.

The final encounter was within the Hard threshold for the party

Battling a single CR 24 monster with a full party of 6, this fight was within the Hard threshold, not Deadly. Given how your party seems to be geared towards using good combat tactics, this one was in the bag for them.

The flip side: your party's resource expenditure

You did not tell us how your party fought, but it's important for you to consider how often they used non-rechargeable resources over these four encounters.

For example, if the party got through the first encounter without expending a single non-rechargeable resource (unlikely, but just for the sake of example), then they will be going into the second fight while essentially still being at full power.

If you were able to take note of how many spell slots, Rages, etc. the party expended per encounter, you will be able to track their attrition rate.

However, your party is not dependent on rests. And so their resources are actually preserved and continue to be useful throughout this adventuring day. That is, the guideline on the DMG for 6-8 Medium encounters has the assumption that these encounters will provide enough attrition to the party as each combat progresses. But your party is attrition-resistant.

The abundance of potions

You say in this comment that some party members had 50 potions of healing. This makes the party incredibly attrition-resistant, since they have access to such a large amount of healing between encounters. They can engorge themselves in healing even without rests.

The Rogue

Consider you have 1 Rogue in the party, which is arguably the least rest-dependent class in the game. They can score a Sneak Attack every round without expending resources, and they have Cunning Action, Evasion, Uncanny Dodge, etc which can be used every round as well.

The Barbarians

Then consider you have 2 Barbarians. These guys are dependent on their Rage, but at 16th level, they have 5 uses of it. You presented only 4 encounters. And so your Barbarians were not dependent on a rest for this adventuring day. The rest of their combat features can be used without expending resources -- Extra Attack, Reckless Attack, Danger Sense, Brutal Critical, etc.

The Paladins

Next, consider you have 2 Paladins. While they are dependent on their spell slots for Divine Smite, they can choose to use it after they know the attack is successful. This means they can ration out their spell slots in a very cost-effective way, making them very efficient spell slot consumers. On top of that, they have Improved Divine Smite, which reduces their need to use spell slots for Divine Smite. They also have Lay on Hands, which means between the two of them, they can heal up to 160 points of damage between encounters. This reduces the pressure on the Cleric for healing, reduces the need for them to use potions (an expendable resource), and reduces the need for short rests. And finally, they have their ever-present Paladin Auras, which provides significant, resource-free defense boosts to the entire party.

The Cleric

So we come down to the only rest-dependent member of this party: the Cleric. If this Cleric is smart, they will (and, as it seems, they did) save their spell slots for support-type and resurrection-type spells. At level 13, they can cast Revivify all day long. And as long as the Cleric doesn't fall, none of the party deaths can be truly lethal. Even in the presence of the Death Giants, the Clerics can defy death with Revivify or Resurrection.

You pulled your punches

Upon your own admission, in both the question and in this comment, you pulled your punches. This means the party did not receive the full brunt of power and damage that your monsters could have dealt.

As I noted in an above point, the effective CR of monsters changes when in actual combat. If they do not do the most damaging options available to them, the true CR of the fight is different -- and lower -- as to what you would expect.

If you started pulling your punches in the first fight, it goes a long way towards explaining how they survived that one. Note that if the Death Giant takes a single round to deal no damage by casting Unholy Aura, their Offensive CR drops from CR 19 to CR 12 (ave damage drops from 120 to 80). The true XP budget of your fight was different from its theoretical XP budget due to this, and instead of something like ~5x their Deadly threshold, it could have been ~3x instead.

And again, for a party that seems to be this tactical, they can survive this. A Deadly encounter does not guarantee a TPK, it only grants you a good chance to kill a PC.

Conclusion: It makes sense

So when we consider that your first encounter was the toughest (and also when the party was strongest), the next three encounters were significantly easier while the party did not get significantly weaker, and that some monsters, especially in the 3rd fight, did not operate at their full Offensive CR, it makes sense why the party survived this adventuring day.

When the PCs weathered the first fight you gave them, it proved their tactical and combative ability. The next three fights had to be tougher, not weaker, to give them a challenge.

The DMG Guidelines are accurate

In my experience, as long as I follow my CR calculations precisely, the DMG guidelines do give an accurate measure of difficulty. I just need to be aware when something I planned goes wrong (ex., monster X is unable to do action Y) because that changes the XP budget a lot.

What this encourages me to do is to have the first three turns of my monsters planned out, and then I'll calculate the true CR of that encounter based on that (and not based on the most damaging options available to that monster).

So I base my encounters on the true CR of the fight I expect, not the theoretical one. And under this style, I find that, yes, Hard encounters are Hard (and may even threaten a TPK), and a Deadly encounter gives the players ideas about they're all going to die.

I recommend a lot of swing

PCs are much squishier than monsters, so when I make custom monsters/villains, I aim for a high variance between the Defensive CR and Offensive CR (ie, glass cannon bad guys, or meat shields with little damage). I find it helps a lot in adding the sense of danger/urgency to have the enemy deal a ton of damage, but the players can wipe out the enemy fairly easily. Conversely, it scares the players a lot when they can't hit or harm the bad guy, even if those bad guys are not dealing that much damage back to them.

When calculating based on the true CR, do not go overboard

If you more strictly follow the DMG guidelines when using the monsters to their fullest extent, your encounters will kill the party. Try and reign the XP budget in, if you start designing encounters this way.

Answer 3

While the above response is excellent and thorough, I would like to add that CR's are extremely subjective by nature and cannot account for all circumstances, they are loose guidelines at BEST. Consider an encounter with a group of kobolds. The CR can be calculated but if it's the 14th encounter and the party has been on a ship in a storm and hasn't had good rest and no spell refresh suddenly the CR looks different...

OR the kobolds are dug into their homes, fighting from elevated positions behind cover with difficult terrain to cross is certainly different than the same exact kobolds met and fought on an open field.

I usually factor in a multiplier to my CR based on circumstances of that specific fight. To be sure as characters level, circumstantial challenges become less and less important as characters have more and more ways to overcome the challenge.

One technique to use is to throw the opening combat as a quality illusion in hopes that the party will blow a lot of buffs and resources, only to wait the next hour before the real battle. All the short term buffs/limited usage effects will be down and wasted.

And remember the bad guys, if you are in their base, have had a very long time to plan and prep for attackers, so it's no cheesy to have lots of traps, dead ends, etc...

All that being said, I think players tend to be more competent than some generic CR, so I have seen this be difficult to get CR's to align with the party as well, in most systems that seems true. It almost seems like the CR's are geared towards a group that makes no effort to optimize themselves and is just playing the most vanilla characters as possible.