Duergar Tactics: Mordenkainen’s Duergar, Part 2

Moar duergar! The duergar mind master is the last of the CR 2 duergar, the one with the ability contour of a spellcaster but no actual spells. What it does have is Mind Mastery, a feature with a 60-foot range which requires an Intelligence saving throw to resist. More to the point, it targets one creature within 60 feet and requires a DC 12 Intelligence save to resist.

This feature, frankly, is terrible. Even a level 1 PC who’s dumped Intelligence still has a 40 percent chance of succeeding on this saving throw. It’s a straight-up waste of an action in any circumstance save one: as part of an ambush. In this instance, a hidden mind master can use Mind Mastery against a target without giving away its position or even its presence if it fails, since Mind Mastery is technically neither an attack nor a spell. If it succeeds, it gets to force an opponent to sucker-punch one of their own allies—or, depending on the local terrain, walk directly into a chasm or a river of lava or something. With Intelligence 15, a mind master is smart enough to know not to bother using this feature in open combat.

So forget treating it as a spellcaster; we’ll pretend that its Intelligence is nothing special after all and it’s just another shock trooper, using Dexterity for offense as well as defense.

In this case, its Multiattack is its core offensive feature, allowing it to make either two melee attacks or one melee attack along with one use of Mind Mastery. Since we’ve established that Mind Mastery is awful, it’s going to stick to the two melee attacks. Note the phrasing: Since Multiattack says “two melee attacks,” not “two Mind-Poison Dagger attacks,” other melee attacks can be substituted, including unarmed attacks, attacks with improvised or seized weapons, grappling, and shoving. However, in most cases, we’re just talking about the Mind-Poison Dagger, which deals 3d6 psychic damage on top of the normal weapon damage. Used twice in a single turn, it’s pretty vicious.

The mind master has Invisibility on a 4–6 recharge, which is unusually generous: it means that, on average, a mind master will be able to use it every other turn. How does a strategy in which the mind master turns invisible on odd turns and Multiattacks on even turns, with advantage on the first attack, stack up against simply Multiattacking each round? Once again assuming a target with AC 15, the mind master has a 55 percent chance to hit normally and an 80 percent chance to hit with advantage on the roll. With damage averaging 16 per hit, that’s 9 expected damage per normal attack, 13 expected damage per attack with advantage. Over two turns, two Multiattacks can be expected to deal 35 damage, while Invisibility followed by one Multiattack can be expected to deal only 22—not nearly good enough. The conclusion to be drawn here is that the duergar mind master turns invisible only when it needs to turn invisible to defend itself (such as when it’s being double- or triple-teamed) or when it has just finished off an opponent and is preparing to relocate, not merely to gain advantage on a single attack.

In lieu of the standard duergar’s Enlarge, the mind master, hilariously, has Reduce—with a handy clause in the description of its Mind-Poison Dagger which ensures that it still does the same amount of damage despite having shrunk to the size of a toothpick. Thanks to this guarantee, Reduce is nearly all upside: although the mind master isn’t going to be bending any bars or lifting any gates while minimized, it gets to Hide as a bonus action with +5 on its Dexterity (Stealth) check, and its Armor Class increase from 14 to 19. Most importantly, it gains these benefits at no penalty to its speed, so it can scamper around, Hiding behind rocks, and pop out to stab confounded foes with its Mind-Poison Dagger. This bonus Hide action, not Invisibility, is the mind master’s strongest source of advantage. Best of all, when it Reduces itself, it can use that bonus action as part of the same turn, so it doesn’t even give up much in the way of time.

The mind master turns invisible and skedaddles when seriously wounded (reduced to 15 hp or fewer). If its Invisibility happens to be on cooldown when its hit points drop this low, it Dashes away if it’s Reduced and Disengages before retreating if it’s not.

The duergar warlord, like the common duergar, is a brute, but a much tougher brute. It’s better armored and has nearly three times as many hit points, but more important, it gets to both Enlarge and attack as part of its Multiattack. Thus, it has no reason not to Enlarge, regardless of the difficulty of the battle. Although the hammer/javelin attacks are listed first, the order of the Multiattack isn’t strictly specified, and using Enlarge before attacking is a no-brainer. After doing this, the duergar warlord combines its melee attacks with Call to Attack on subsequent rounds.

Because Call to Attack consumes the reactions of other duergar, you might think the duergar warlord has to be careful not to use it on allies who might have to make opportunity attacks. It’s true that if you want to maximize the effect of Call to Attack, you want to throw the reaction attacks to duergar who aren’t going to have to make OAs—but nothing is lost by using a reaction for a Call to Attack attack instead of an opportunity attack. The attack simply occurs sooner. That being said, the foes you most want to wear down with as many attacks as possible are the tough front-liners, and they’re also the ones least likely to run from a fight, so using Call to Attack against them is a good bet.

The Psychic-Attuned Hammer is a significantly better weapon than the javelin. Favor the hammer, unless the warlord can’t actually reach an enemy to hit them with it. In that case, huck some javelins—or turn invisible. The duergar warlord’s Invisibility, like the mind master’s, has a fast 4–6 recharge. But it uses this feature to appear in surprising places, not simply to gain advantage on a single hammer attack; throwing three javelins offers greater benefit than the increase in expected damage from having advantage on one hammer attack.

When the duergar warlord itself clearly has no need to make any opportunity attack, it can use its reaction on a Scouring Instruction instead. This damages the ally the warlord uses it on, but the warlord is lawful evil; do you think it cares? The greatest courtesy it affords its chosen minion is not to use it when said minion is 4 or fewer hit points away from being seriously wounded.

As for the warlord itself, it retreats when it’s seriously wounded (reduced to 30 hp or fewer), using the Dodge action if it’s engaged in melee with only one opponent and the Dash action if engaged with more than one. (Disengaging is pointless—duergar are too slow to avoid being chased down. Better to take one hit per opponent now than two or three hits on each of your opponents’ turns.)

With its extraordinary Strength and Constitution and bargain-basement Dexterity, the duergar despot isn’t just a brute, it’s a juggernaut, and its tactics have nothing in common with those of regular duergar.

It’s happy to engage in melee with multiple foes at a time, because its Multiattack includes two Iron Fist attacks, and each of those attacks has the potential to knock a foe back up to 30 feet. But note that this is up to 30 feet. “Up to” includes zero. Against a single opponent, the most powerful Multiattack combo is Iron Fist–Stomping Foot–Stomping Foot–Iron Fist, with the first Iron Fist attack knocking the target prone (on a failed saving throw) but not away. Thus, the prone target remains within reach of Stomping Foot, on which the duergar despot now has advantage. This sequence also works if the opponent is knocked away only a short distance—say, 5 or 10 feet. Although the duergar despot can knock a foe 30 feet back, it’s going to limit its knockback to 25 feet—and then use its movement to stay in that foe’s face, so that they provoke an opportunity attack if they try to get away—unless that foe is either a particularly problematic melee opponent or so far beneath the duergar despot, in its estimation, that it’s just taking out the trash. Against two melee opponents, it uses Iron Fist once and Stomping Foot once against each.

Two Iron Fists and two Stomping Feet are a huge combo, dealing an expected 50 damage against AC 19 (if you’re tanking against a duergar despot, you probably don’t have AC 15 anymore). But Flame Jet is an area effect that according to the Targets in Area of Effect table in chapter 8 of the Dungeon Master’s Guide can be expected to flambé three opponents, for an expected total of 40 fire damage (assuming a 50/50 likelihood of success or failure on the Dexterity save). Now, usually when a monster has a Multiattack and an area-effect ability, the area-effect ability consumes a full action and must be compared and contrasted with Multiattack head-to-head. But in the duergar despot’s case, its Multiattack allows it to replace any or all of its melee attacks with Flame Jet!

The implications are staggering. Flame Jet no longer has to be rationed down to instances in which it can strike three or more opponents. Against one opponent, it’s better than Stomping Foot against a non-prone target. Against two, it’s better than Stomping Foot against any target, prone or no. Against three or more, it’s a damage hose that the despot has no need to shut off.

And it’s not limited to targets it can hose down from its current position. Once per day, it can cast misty step, and woe to foes who slip up and arrange themselves three-in-a-row, in any direction, even once: the duergar despot will bamf to just the right spot to incinerate them, placing itself adjacent to the nearest one, so that they get whomped if they try to scramble away.

As for the duergar despot’s other spells—well, they’re less clear-cut than misty step. Mage hand and minor illusion are limited in their usefulness during combat. Stinking cloud has potential against four or more enemies who aren’t lined up properly for a Flame Jet, but the save DC is embarrassingly low, making it a suitable use of an action only when (a) its targets are trapped in a confined space and can’t simply walk out of the cloud, (b) the duergar despot can’t yet get close enough to beat on them directly, and (c) the duergar despot doesn’t need its allies to march into the cloud and help it out, because they’re not immune to the cloud’s poison. (Duergar constructs are immune to the poison, however, so if there are any of those on hand, they can attack targets in the cloud, no problem.)

A lot of Dungeon Masters are unsure how generous or stingy to be with counterspell, vis-à-vis how readily a caster using it can tell what spell is being cast and decide whether it’s one that needs to be countered. Here’s the house rule I’ve settled on:

  • If an opponent is casting a spell you know, you always recognize it immediately.
  • If an opponent is casting a spell you don’t know, you recognize it if you succeed on an Intelligence (Arcana) check with a DC equal to 8 + twice the spell’s base level (that is, since scorching ray is a 2nd-level spell, the DC to recognize it is 12, even if it’s being cast using a 4th-level spell slot). If you’re not proficient in Arcana, you have disadvantage on this check. No roll is required if passive Intelligence (Arcana) is sufficient to meet or beat the DC.

Applying this rule to the duergar despot, it makes its skill checks to recognize opponents’ spells before they take effect at +2, with disadvantage. Since it’s immune to being charmed, exhausted, frightened, paralyzed or poisoned, it doesn’t bother countering spells that inflict any of these conditions. (It doesn’t care whether its minions are affected by them. That’s lawful evil for you.) It also doesn’t bother countering 1st-level spells, and its chances of countering spells of 4th level or higher are so poor that it doesn’t bother trying to counter them, either. Finally, it takes its chances on spells that call for Constitution or Wisdom saving throws, because its save modifiers for these are very good. Its Dex save mod, on the other hand, is dismal, so damaging or debilitating 2nd- and 3rd-level spells that are resisted with Dex saves are at the top of its list of spells to snuff out. These include call lightning, conjure barrage, fireball, lightning bolt, sleet storm and web.

Psychic Engine is a passive feature that deals psychic damage to adjacent creatures when the duergar despot takes a critical hit or is reduced to 0 hp. This trait merely gives the duergar despot an incentive to stay in melee reach of as many enemies as possible. It might also give it incentive to fight to the death, but . . . nah. Its Wisdom is high enough that it would rather live to tyrannize another day. When it’s seriously wounded (reduced to 47 hp or fewer), it withdraws, using Iron Fist attacks to knock pursuers back 30 feet if there are any within melee reach and Dodge if there aren’t. In the former case, to round out its Multiattack, it uses Flame Jet opportunistically.

Duergar may be accompanied into battle by one or both of two types of constructs. The duergar hammerer is an unlucky duergar imprisoned inside a device whose function is to destroy objects and structures, and that’s what it does, without subtlety or complexity. Similarly, the duergar screamer is a duergar imprisoned inside a device that amplifies its anguish to rock-piercing intensity. Basically a rolling thunderwave spell with a melee attack attached, it rolls directly toward clustered enemies and tries to blast its way through them. If opponents aren’t clustered, it rolls toward one at random. Beyond these simple directives, neither hammerers nor screamers possess any target selection criteria, any self-preservation impulse, any ability to adapt or any independent judgment whatsoever.

Next: chitines and choldriths.

7 thoughts on “Duergar Tactics: Mordenkainen’s Duergar, Part 2

  1. Long time reader, first time poster. These guides are an incredible, indispensable resource for my d&d sessions – my players have just delved into the Underdark for the first time and met their first Choldrith, I will be fascinated to see how you suggest running them. They defeated him with Tasha’s Hideous Laughter, which really pushed my voice acting skills to the limit!
    Thanks for writing these, and keep up the good work!

  2. I should point out that the Despot’s psionic save DC seems to be miscalculated. Their proficiency is 4 and their intelligence is 2 so it should be 14 rather than 12. Still kind of low, but not so low as to dismiss the use of its’ save spells entirely. Errata may fix that. (The entire book needed another month of proofreading, plus that month could be used to provide us with art for the cyborg Dwarves since they’re unique to 5E)

    I feel you’re doing a slight disservice to Hammerers and Screamers: Yes they are mindless constructs, but they’re designed to follow orders. They warrant a short writeup of what tactics their commanders might have them employ.

    1. I agree with the need to proofread longer and find more art as the Molydeus is one of the most interesting monsters in MToF and there is no art for it, I would of loved to of seen a new representation of ManWolfSnake as my players dubbed him after I searched some art online.

  3. I think you forget on thing, or maybe you don’t forget and just make the assumption. A good DM will take the stat block, and alter it to what the DM deems is appropriate his party’s abilities. If the statblock says DC 12, and the DM thinks its weak (and it most certainly is), the DM can make it DC 15, or DC 25, or whatever.

    I would prefer if you did these excellent writeups from the view of making suggestions to things inexperienced DM could alter to improve the challenges to something more reasonable.

    1. Such changes aren’t made in a vacuum though; it can be harder for the inexperienced GM to have the discipline and judgement to change something with a reasonable expectation of balance that it could be to simply say: the feature is so bad that this is the only sotuation it’s used. Sometimes those rules cause encounters to be far more interesting anyhow!

    2. Analysing the monsters as they are written is a MAMMOTH task. Seriously. It’s enough, without asking him to speculate about house-rules (he’s addressed this a few times). Your request is much much bigger than I believe you intended.

      In cases where KA thinks there’s an error, he tells us so. Sometimes he creates wholly new monsters for us. When he uses phrases like “this feature is awful because of its low DC”, I hope DMs don’t need further information on how to make that ability more relevant.

      There’s only so much one person can do. KA set out to analyse the tactics of published monsters.

      (Furthermore I think the encounter-building guidelines in the DMG and other book address your objectives. There’s even info somewhere in published material about how to reassess a monster’s CR if you change its stats).

  4. Tremendous work on the duergar. I really appreciate recognizing how deadly some of the combinations can be. Gives me, as the DM, the flexibility to take the PCs to the limit if things are going too easily.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.