MOAR! Monsters Know What They're Doing

THE MONSTERS

HUMANOIDS

XVARTS (V)

Xvarts are difficult to devise tactics for, because their ability scores, their features, and their flavor text in Volo’s all seem to be at odds with one another. Their ability scores suggest Dexterity-focused sniping and shock attacks. Their Overbearing Pack feature suggests a reliance on shoving opponents prone, presumably to be followed up with melee attacks (both of which depend on Strength). And the flavor text states that they attack primarily to abduct, which implies either grappling targets or knocking them unconscious.

There is a solution, but it’s tricky.

Xvarts move at the normal humanoid speed of 30 feet per round. Their Strength is low, and their Constitution merely average, so they’re averse to melee slugfests. Xvarts necessarily seek strength in numbers—and allies, specifically giant rats and giant bats. Giant rats make particularly good allies for xvarts, because of their Pack Tactics feature; giant bats, however, are tougher, with a slightly higher challenge rating. A xvart encounter should include, at a minimum, two xvarts per player character, plus an animal ally for every two xvarts.

Xvarts are neither smart nor wise. They have no ability to adapt if their favored strategy doesn’t work, and they may not be particularly quick to notice that it isn’t working. However, unlike the usual low-Wisdom monster, which waits too long to run away, xvarts are cowardly; if anything, they’ll run away prematurely from encounters that favor them. The Low Cunning feature gives them Disengage as a bonus action, but this represents instinctive evasive ability, not discipline.

So how do you create an effective kidnapper out of a low-Strength creature without proficiency in Athletics? By building a strategy around the distinctive Overbearing Pack feature, which gives a xvart advantage on a shoving attack as long as it has an ally—in this case, a giant rat—within 5 feet of the target. The xvart’s Strength modifier is an unimpressive -1, but advantage jacks this up to an effective +3 or +4, allowing it to compete with stronger or nimbler PCs.

Xvarts have darkvision and proficiency in Stealth. By attacking only in twilight or (if the party contains PCs who have darkvision themselves) at night, xvarts can minimize their chance of being seen before they attack. They send their animal allies out first (giant rats have darkvision, and giant bats have blindsight, so visibility isn’t a problem) to distract their enemies, then follow up with attacks of their own.

Let’s say we start with giant rats. The xvarts send in the rats, which lack Stealth proficiency, so they attack in pairs to gain advantage from Pack Tactics. The xvarts wait until their enemies are engaged, then move in and attack to shove their victims prone, intending to follow up by grappling so that they can’t get back up.

Observant readers will have noticed that, based on my encounter building guidelines, there won’t be enough animal allies to engage every enemy. The rats and bats are for use against the front line. The xvarts themselves, which are numerous, can take care of the rest of their opponents by double- and triple-teaming them.

Example: A level 1 party consists of Áine, a wizard; Daria, a marksman ranger; George, a sword-wielding rogue; Lennie, a front-line fighter; and Tolmac, a druid. These five PCs are ambushed at night by a pack of ten xvarts, accompanied by five giant rats. When the attack is launched, three of the rats swarm Lennie, while the other two go after either George or Tolmac—let’s say George. (Sending rats to attack a druid would be a mistake, but then again, it’s exactly the sort of mistake a dimwitted xvart might make. Don’t overthink it. The xvarts pick George because George has a sword, while Tolmac just has a big stick.)

Once the rats have engaged their foes, the xvarts move in: one against Lennie—nah, make it two, he’s pretty big; one against George; two against Áine; three against Daria, who looks a little tougher; and two against Tolmac. (Xvarts judge their opponents mainly on the basis of size, armor, and armament.) Taking advantage of low visibility and their targets’ distraction (Player’s Handbook, chapter 7, Hiding), they slink up to their targets and try to waylay them without being noticed before they strike. If a xvart is spotted as it approaches, it doesn’t attack; instead, it immediately runs for cover and tries to Hide again.

As I mentioned above, the xvarts’ first attack is a shoving attack, to knock their opponents prone. If two or more xvarts are attacking a single target and the first succeeds in knocking the target prone, the second immediately attacks to grapple. If the target isn’t grappled by the end of the turn, they’ll probably get right back up, which will scare the xvarts and cause them to run for cover.

If the target is grappled, the xvarts remain adjacent to one another as they drag the target off, the auxiliary xvart using the Help action to give advantage to the grappling xvart on its Strength (Athletics) checks, while menacing the victim with its shortsword. If there’s a third xvart (as against Daria above), it gives the victim a smack with its shortsword for good measure, aiming to knock them unconscious.

Additionally, as the Volo’s flavor text says, the xvarts throw nets or sacks over their grappled victims, since probability dictates that a strong victim will get lucky and break the grapple after a few rounds even if the xvarts have advantage on their rolls. You’ll have to use DM discretion and combine the restraining effect of the net (Player’s Handbook, chapter 5, Special Weapons) with the xvarts’ advantage on Strength (Athletics) checks to resist escape, or it will be too easy to get out. Even so, being restrained imposes disadvantage only on Dexterity checks, not on Strength checks, so in all likelihood, Lennie will still burst his bonds sooner or later—and that’s okay. Making things fair for the xvarts doesn’t mean making them unfair for the PCs.

Xvarts are team players by necessity, not by nature, and if one is moderately wounded (reduced to 4 hp or fewer), it runs off—using the Disengage bonus action, then the Dash action—and leaves its companions to their fate. Xvarts hauling a grappled or restrained enemy who breaks free also make a run for it, as do their companions, as soon as they notice. So do xvarts abandoned by their partners in the initial attack.

The presence of a xvart warlock changes the parameters of the encounter. For starters, xvart warlocks, if they leave their settlements, are usually leading expeditions in search of magical treasure, not sacrificial victims to abduct. In this scenario, the xvarts apportion and comport themselves differently: A giant rat attacks each target, then two xvarts attack that target normally with their shortswords (statistically superior, in terms of expected damage, to giving up one of those two attacks in order to try to knock the target prone). Then both xvarts Disengage (bonus action) and retreat 15 feet from their target; if the target pursues, it’s subject to an opportunity attack from the rat. Moderately wounded xvarts still run away, Disengaging (bonus action), then Dashing (action).

Meanwhile, the xvart warlock, fortified with mage armor, keeps its eyes peeled for enemies who are seriously wounded, whom it can finish off. The warlock always wants to try to get in the last blow, because Raxivort’s Blessing gives it 4 temporary hit points whenever it reduces an enemy to 0 hp.

The warlock has four offensive spells, two leveled (burning hands and scorching ray) and two cantrips (eldritch blast and poison spray).

Burning hands requires a Dexterity save to resist and deals 4d6 fire damage to each enemy in its area of effect who fails their save and half that to each enemy who succeeds; the warlock wants to make sure there are at least two (Targets in Areas). Thus, we’re looking at about 21 expected damage.

Scorching ray, on the other hand, is a ranged spell attack, and it does only half as much expected damage—but it’s good out to a range of 120 feet, as opposed to burning hands’ 15 feet.

Each of these spells costs a spell slot, of which the warlock has only two—and it wants to save one for either expeditious retreat or invisibility, its escape-hatch spells. Thus, it saves them for moments when it can finish off at least one of two opponents at close range or at least one of three at long range.

It has no limit on eldritch blast and poison spray, but neither of these can be expected to deal more than a few points of damage—in fact, if an enemy comes within 5 feet of the warlock, it’s going to take a swing with its scimitar rather than take a chance on poison spray. Poison spray tends not to work so great on the kinds of enemies who’ll charge a warlock.

Expeditious retreat and invisibility both require concentration, so the xvart warlock can cast one or the other but not both. It casts the former if its path of escape includes an abundance of cover, the latter if it has to run away across open terrain. Like other xvarts, the xvart warlock is a coward, and it takes to its heels when it’s only moderately wounded (reduced to 15 hp or fewer).

DERRO (M)

Derro are small humanoids native to the Underdark. Equal parts fearful and vicious, says Mordenkainen’s, [they] prey on those weaker than themselves, while giving simpering obeisance to any creatures they deem more powerful. You know the type: They believe absurdities, and they applaud atrocities.

With high Dexterity and above-average Constitution but merely average Strength, derro are skirmishers, but they’re not especially mobile ones. Their Intelligence is average, but their Wisdom, for some reason, is in the cellar. This mental ability contour is unusual; the reverse is far more common, especially since Wisdom supports the Perception skill. Not only are they easy to get the drop on, they also have an underdeveloped survival instinct, making them more likely to fight to the death. They are, however, proficient in Stealth, predisposing them toward an ambush strategy.

They have excellent darkvision and Sunlight Sensitivity, so they rarely venture aboveground for any reason, and absolutely never during the day. These traits combine with their innate paranoia to suggest an intense territoriality: Not only do they defend their turf viciously, they hardly ever leave it at all, except to try to conquer an adjoining sliver of new territory.

Derro have two weapon attacks, Hooked Spear and Light Crossbow. One option with the spear is to knock an enemy prone (presumably by hooking and tripping them), which would give any adjacent melee attacker advantage on a follow-up attack. However, a ranged attacker has disadvantage against a prone target, so this doesn’t help the crossbow-wielding derro at all. Even worse: It turns out, if you run the numbers, that even if the first derro in a group successfully hooks and trips an enemy, its allies nearly always deal less expected damage, despite having advantage on their attack rolls, than the group would do if all of them simply attacked to deal damage.

This holds true for any group of two to five derro. It takes six or more derro attacking a single opponent in melee for the advantage from hooking and tripping to produce an increase in overall damage, and at that threshold, it works only against unarmored, lightly armored, or moderately armored opponents.

Reflecting on this fact, I think the hook-and-trip should be considered an advanced derro tactic. Derro have a challenge rating of only 1/4, so you can throw them against even level 1 PCs. Against these PCs, they don’t appear in great enough numbers to do anything but stab. But once your PCs are up around level 5 and higher, they’re going to be fighting hordes of derro, not just patrols and platoons, and in that instance, the first in each group of attackers hooks and trips to try to give the rest advantage. (For the sanity of your players and yourself, use the Handling Mobs rule in chapter 8 of the Dungeon Master’s Guide, and figure that advantage gives +4 to hit.) Assuming they all share the same initiative count, derro wielding crossbows all shoot first, before any of their enemies fall prone, and then the derro with spears attack.

Alternatively, if you’re more interested in flavor than in optimization, always have the first derro in a group of three or more wielding spears attack to hook and trip! The average difference in damage is less than 1 point, and your players probably won’t do the math on the fly and realize that being flat on their prats doesn’t put them in any more danger than they were in standing up.

After all that, it seems almost anticlimactic to point out that the light crossbow deals significantly more damage than the spear—86 percent more, on average. So rather than divide up a derro unit between crossbowmen and spearmen, assume that every derro carries both a crossbow and a spear; that they prefer to use their crossbows over their spears; but that when an enemy rushes them, they switch, and so do their immediate neighbors. Also, if at all possible, they launch their first crossbow volley from hiding, to gain unseen attacker advantage.

Derro paranoia and low Wisdom mean they don’t flee when seriously wounded but rather keep on fighting until they’re down to 1 or 2 hp. At that point, the gravely wounded derro run, baiting out opportunity attacks—and their erstwhile allies seize that opening to retreat out of melee range themselves and go back to attacking with their higher-damage crossbows. If combat drags on beyond three rounds, all derro flee the scene, Dashing away. But this is simply a strategic retreat. They’ll gather some more allies, stalk their opponents, and ambush again as soon as they get the chance.

The Derro Madness sidebar in Mordenkainen’s is mainly a roleplaying guide, with little or no impact on derro tactics. Advantage on saving throws against being charmed and frightened doesn’t change much, since derro already have Magic Resistance; it adds only a small number of edge cases, such as being resistant to a Berserker barbarian’s Intimidating Presence. The effect of Insanity is already encapsulated by the derro’s pitiful Wisdom score, which scrambles its awareness of how much danger it’s in.

Derro savants are derro with sorcerous ability. Aside from having high Charisma and slightly below-average Strength, they have exactly the same ability contour as a regular derro. Because that below-average Strength makes them even less effective in melee, however, they always attack from range, and other derro run interference for them in case an enemy tries to close to melee distance.

Lightning bolt is the big gun in the derro savant’s arsenal, but it has the drawback of affecting only a narrow, straight line. Invisibility, however, gives the derro savant the freedom to position itself where it can cast a lightning bolt that nails four or more enemies, if they’re properly lined up. It’s most likely to get this chance if the battle has a well-defined front line. In a more all-over-the-place battle, there may never be a good opportunity to cast lightning bolt.

Normally, I’d say, the derro savant should use its 3rd-level spell slots for lightning bolt and nothing else. But I’d also say that because of the length of its area of effect, it’s practically wasted if cast against just one or two enemies. So what about, say, boosting chromatic orb with a 3rd-level spell slot? That would make it deal 5d8 damage (22, on average) against a single enemy with a ranged spell attack roll, vs. 8d6 damage (28, on average) against one or two enemies, with the burden on them to make a Dexterity saving throw, and half damage dealt even if they succeed. There’s no comparing the two. Chromatic orb falls far short.

Burning hands? At least that one requires a Dex save, deals half damage on a success, and can affect a second target, but even when boosted to 3rd level, the base damage is only 5d6 (17, on average). This one’s a self-defense measure for when the derro savant gets sacked, nothing more. As for sleep, it just doesn’t scale well. So save those 3rd-level spell slots, even if the opportunity to cast lightning bolt doesn’t seem to present itself. The derro savant holds out hope that the moment will come eventually, and when it does, it will be ready.

As for cantrips, the derro savant has two that deal damage: acid splash and ray of frost. Ray of frost is better, but neither is that great. The derro savant uses up its lightning bolts and chromatic orbs before resorting to cantrips. Spider climb is useful for casting from the safety of a high ceiling, escaping in a high-verticality environment, and not much else—and the derro savant is as unlikely to flee as any other derro. Sleep won’t be much use once the PCs are past level 4 or so, unless a significant number of them are seriously wounded. Then again, derro—even derro savants—are poor judges of the strength of their opposition, so a derro savant may unwittingly waste sleep on a group in which it will affect one target at most.

GRUNGS (V)

I have to hand it to Volo’s for giving us grungs, undisputed winners of the Most Adorable Evil Creature title, formerly held by kobolds.

Clearly based on poison arrow frogs, grungs are arboreal rainforest dwellers, tribal and territorial. In the latter respect, their behavior in groups therefore resembles that of lizardfolk, so I refer readers to my entry on them in The Monsters Know What They’re Doing (34–38). Grungs’ amphibian nature also invites comparison to bullywugs.

Lizardfolk are brutes, but grungs are low-Strength, high-Dexterity, high-Constitution skirmishers. Their low Strength means they’re going to be encountered in large numbers: no fewer than half a dozen at a time, I’d say. If they’re going to initiate an encounter against your PCs, rather than vice versa, they’ll have to outnumber the party at least three to one.

Grungs share the Amphibious and Standing Leap features with bullywugs. This means they’ll often be found in swampy areas, around rivers, and in other sorts of difficult terrain, which they can get around in easily by jumping. They’re quicker than bullywugs, though not as quick as most PCs, and since they can climb as well as jump, they use their proficiency in Stealth to hide in trees and drop on their enemies from above.

Grungs aren’t stupid. They have average Intelligence and Wisdom, so they’re not going to keep using a certain tactic if it’s not working, and they know when they’re beaten. Once they no longer outnumber their foes by at least two to one, or once a majority of them are seriously injured (reduced to 4 hp or fewer) or killed, they Dash off toward the heart of their territory.

Ordinary grungs wield daggers. Using them as a ranged weapon isn’t a completely absurd proposition, but their range is so poor that grungs can easily close with any enemy they could reasonably hit with a dagger and engage them in melee. Ideally, therefore, grungs would like to outnumber their opponents and surround them; they throw their daggers only if this isn’t an option.

Fighting grungs should be like fighting popcorn. Anytime an opponent lands a hit on a grung, the first thing it does on its next turn is jump away, either off into the undergrowth or up into a tree. Either way, it can then use its action to Hide, since its fellow grungs can keep whoever struck it occupied. Once it’s hidden, it maneuvers around until it can attack from hiding again. This assumes, of course, that the damage the grung took wasn’t enough to seriously wound it—in that case, it simply flees.

The Grung Poison variant suggests a variety of quasi-hallucinogenic side effects of failing the saving throw against grung poison. What’s not clear is whether these effects are meant to apply only to the toxin in the grung’s Poisonous Skin or the one on its dagger as well. If it includes the latter, an army of grung warriors will consist of green grungs, whose toxin limits a poisoned creature’s movement to climbing and standing jumps. Once an enemy is poisoned, the grungs stop attacking it and take it captive.

If you’re not using the Grung Poison variant, capture is still the grungs’ ultimate goal. Two or more grungs that have knocked an enemy out drag it back toward the heart of their territory; one alone can only move at a speed of 5 feet per turn if its burden weighs more than 105 pounds, and since it would be moving across difficult terrain (rainforest), that speed would be reduced to 2 feet per turn. (Jumping while carrying the dead weight of an unconscious abductee simply isn’t workable.) However, if enough of them are cooperating to manage the burden, the only applicable penalty is the terrain penalty, meaning they’ll move at 12 feet per turn. I’d also say the half-movement penalty for moving a conscious grappled enemy (which would further reduce their speed to 6 feet per turn) doesn’t apply if two or more of them are cooperating to carry them off.

The Poisonous Skin feature is an interesting one, because it explicitly describes what happens when another creature tries to grapple a grung but not what happens if a grung tries to grapple another creature. Should the same effect occur? I’d say yes, based on the phrase or otherwise comes into direct contact with the grung’s skin. This ruling aids the grungs in their efforts to kidnap trespassers.

In summary, here’s how I see a grung combat encounter playing out:

Green grung warriors hide in trees, scouting for trespassers. They attack when they outnumber the trespassers by three to one or more.

On their first turn, they leap down upon their (hopefully) surprised targets, attacking from hiding (advantage) with their daggers.

On their second turn, if they’ve taken a light or moderate hit (up to 6 damage), they leap away (movement, possibly incurring an opportunity attack), Hide (action), and prepare to attack from hiding again. If they’ve taken a serious hit (7 damage or more), they Dash (action, possibly incurring an opportunity attack) back home. If they haven’t taken any damage, they attack one more time.

On their third turn, assuming they still outnumber their enemies by at least two to one, they all try to grapple their targets. If the first grung of a pair succeeds on its grapple, the second one takes the Help action to grant it advantage on its roll to keep the opponent grappled should they try to escape. A successful grapple requires the target to make a DC 12 Constitution saving throw against being poisoned. The poisoned condition gives the target disadvantage on all attack rolls and ability checks; in addition, if you’re using the Grung Poison variant, the poisoned target can no longer move except to climb or make standing jumps.

On their fourth turn, those grungs that have successfully grappled their opponents haul them off, while those that haven’t keep trying until their opponents are subdued or their numbers are too badly reduced for them to keep fighting.

If a grung fails two grapple attempts in a row against a particular target, it gives up on this tactic and goes back to attacking with its dagger.

Grung wildlings have spellcasting ability, and this ability is nastier than it looks at first, because of how their spells can be combined. Specifically, spike growth plus plant growth is a killer combination. Plant growth instantaneously causes all normal plants within a 100-foot radius to grow so thick that movement is slowed by a factor of four. (This doesn’t inhibit the grungs’ movement, since they can jump over the growth rather than have to slog through it.) Spike growth, meanwhile, deals piercing damage to any creature moving into or through its area of effect. Because plant growth gives not only a grung wildling but also all its allies a comparative advantage over their foes, a wildling casts this spell on its first turn and follows up with spike growth on its second, casting it where it can catch at least four enemies in the area of effect. If the wildling’s enemies are too spread out for it to target four of them with spike growth, it uses its second turn to cast barkskin on itself instead. (It can’t sustain both of these spells at the same time.)

On its third turn and thereafter, the grung wildling supports its team, shooting into the melee with its shortbow. If at any time it needs to jump away, it does so with the help of the jump spell, which allows it to jump a distance of 75 feet (and remains in effect for a full minute, without concentration). It uses cure wounds only to support an elite warrior, if one is present.

Grung wildlings are red grungs, and a creature affected by their toxin is overcome by hunger and uses its action to eat any food it can get at.

A grung elite warrior has the same combat abilities as an ordinary grung, only better, plus the Mesmerizing Chirr action. This feature stuns non-grungs within earshot who fail their saving throws, so that they can’t take actions or reactions and automatically fail Strength and Dexterity saves, and attacks against them have advantage. If a PC spots the grungs before they launch their attack from hiding, the grung elite warrior uses this feature right away, to try to regain the advantage denied by their blown cover. Otherwise, it uses it on its second turn, and again as soon as it recharges.

Aside from this, the grung elite warrior fights as a normal grung, with a couple of exceptions.

First, the elite warrior can take a lot more hits than an ordinary grung can, so it doesn’t leap away when it takes one, unless that hit seriously wounds it (reduces it to 19 hp or fewer).

Second, in that instance, it calls a retreat and takes the Disengage action rather than the Dash action, fighting as a rearguard while its allies disperse; it only leaps away itself once all its allies are safely gone.

Third, grung elite warriors are orange grungs. Their Grung Poison causes the victim to be frightened of their allies. A close reading of the frightened condition is important here: "A frightened creature has disadvantage on ability checks and attack rolls while the source of its fear is within line of sight" (Player’s Handbook, appendix A, emphasis mine). This rule is easily distorted by imperfect memory into having disadvantage on attack rolls against whatever the frightened creature is frightened of; in fact, a PC who’s frightened of their own allies has disadvantage on all attack rolls—and ability checks—as long as any ally is visible. Thus, once a grung elite warrior grapples an enemy, if that enemy fails their save against being poisoned, they have disadvantage on attempts to break the grapple as long as one of their own allies is still in view. And if the grungs are carrying the whole party off together…

Fourth, key locations within the grungs’ territory may be heavily guarded by elite warriors. These grungs don’t mess around with the kidnapping shtick. They stand at a distance and pincushion intruders with poisoned arrows.

KOBOLDS (V)

Volo’s includes stat blocks for three new varieties of kobold: the kobold dragonshield, the kobold scale sorcerer, and the kobold inventor.

The kobold dragonshield is simply an exceptionally strong kobold with a melee Multiattack and selective elemental damage resistance. These don’t affect its tactics, nor does its Heart of the Dragon feature, which allows it to shake off the frightened or paralyzed condition and do the same for allies around it. But its above-average Dexterity and Constitution, as well as its greater number of hit points, reduce its dependence on ranged combat and make it an effective skirmisher. Kobold dragonshields are bold nuisances, charging in to jab with their spears, then Dodging as they retreat. In kobold lairs, they use this maneuver to draw their pursuers into tight situations or traps, Multiattacking until they take a light wound (5 hp or greater), then Dodging and retreating. Elsewhere, they retreat when moderately wounded (reduced to 30 hp or fewer) but don’t go far, hoping to resume their attack later; when seriously wounded (reduced to 17 hp or fewer), they retreat for real. Since the kobold dragonshield’s Armor Class is predicated on its carrying a shield, assume that it wields its spear one-handed.

The kobold scale sorcerer has a more impressive set of spells than you’d expect a kobold spellcaster to have. Its 2nd-level scorching ray and 1st-level chromatic orb are sound and solid damaging spells that it doesn’t need to think twice about casting. Expeditious retreat gives it a way to slip away from a charging melee attacker; fire bolt and poison spray are useful self-defense techniques if it runs out of spell slots.

Its two metamagic options, Heightened Spell and Subtle Spell, are curiously only useful for charm person and mage hand, respectively. Any ranged spell attack would give the kobold sorcerer’s position away, regardless of whether the spell had a verbal or somatic component, and charm person is the only spell the sorcerer possesses that requires the target to make a saving throw. Both spells strike me as particularly useful for setting off traps—either using mage hand to do so remotely or using charm person to get an enemy to walk directly into a trap or to trigger one against their own allies.

The kobold inventor is sheer lunatic brilliance. It’s an ordinary kobold, and it fights like one, except that it uses its Weapon Inventions in lieu of normal weapons. But which of its Weapon Inventions is most effective? Here’s my attempt to rank them:

Green Slime Pot. On a direct hit, deals an average 6 acid damage and distracts the target, who has to spend their next turn figuring out how to get it off, or it will keep dealing damage; on a miss, creates an environmental hazard. Its value depends on its dealing ongoing damage, so it’s best deployed in round 1.

Basket of Centipedes. Marginally more powerful than Wasp Nest in a Bag, in that a victim reduced to 0 hp by the centipedes’ bites is paralyzed for an hour. That probably won’t come to pass, though, since the centipede swarm deals only 8 expected damage, total, over three rounds of combat. This invention and the Wasp Nest in a Bag are strongest when they’re first deployed, so they’re the best choices for the second and third rounds of combat.

Wasp Nest in a Bag. Like the Basket of Centipedes, deals only 8 expected damage over three rounds, but you get to yell, "Beeeeeeeeees!" (Yeah, they’re wasps, not bees. Who cares?)

Rot Grub Pot. Its drawback is that an enemy has to walk into it for it to do any damage—or not walk away from it, if it was deployed to the enemy’s space. But in either of those cases, or if the rot grubs get to act before the enemy does, it deals 9 damage on a hit and requires the victim to burn themself with fire in order to keep the rot grubs from burrowing in and dealing the same amount of damage again, round after round, until the hapless victim is cured or killed. Like the Green Slime Pot, it’s best used in round 1—but since it’s consistently just a bit inferior to the various alternatives, it never quite makes the cut until round 5. I rank it fourth anyway because it’s a viable alternative to the Green Slime Pot in round 1, should you prefer it.

Acid. Deals 7 acid damage on a hit (4 expected damage). Nothing special, except that it can bypass the damage resistance of a raging barbarian. Best in the last round of combat, but still not as good as the Basket of Centipedes or Wasp Nest in a Bag. Break it out if the combat encounter drags into a fourth round.

Skunk in a Cage. No. 1, hilarious, except to the unhappy skunk. No. 2, can potentially take an enemy out of combat entirely. But, No. 3, can end up disabling the inventor or one of its allies ("It rolls initiative and, on its turn, uses its action to spray musk at a random creature within 5 feet of it"—not necessarily at an enemy).

Scorpion on a Stick. Deals only 2 expected damage, with no other benefit, in any given round. However, the kobold inventor can keep using this gizmo round after round, unlike its other inventions. It comes in behind the Basket of Centipedes and the Wasp Nest in a Bag because the default length of a combat encounter in fifth edition D&D is assumed to be three rounds.

Alchemist’s Fire. Weak damage, easy to put out. Can rob the target of one turn’s action, but even if it lasts three rounds, it probably won’t deal more than 2 damage.

Note that Acid, Alchemist’s Fire, and Green Slime Pot are all ranged weapon attacks with a normal range of only 5 feet. If it uses them at that range, the kobold inventor will have disadvantage on its attack roll from attacking point-blank (Player’s Handbook, chapter 9, Ranged Attacks in Close Combat); if it uses them from any farther away, it will have disadvantage from attacking at long range. Having disadvantage either way, the kobold inventor makes these attacks only against targets it has an ally adjacent to, so that Pack Tactics cancels its disadvantage out.

TORTLES (M)

Tortles are easygoing, nomadic, testudinal humanoids who rarely fight except in self-defense, having friendly dispositions and no territorial impulse. Except when they gather to exchange goods and news, they tend to live relatively solitary lives and are unlikely to be encountered in large groups.

The common tortle is tough, with high Strength and above-average Constitution, Wisdom, and Charisma. They lack proficiency in any social skill, so if being straightforwardly nice isn’t enough to get them out of a jam, they’ll either defend themselves in melee combat or, if they can tell they’re outmatched, retract into their shells until the danger passes.

According to their stat block, tortles typically carry quarterstaves—repurposed walking sticks, perhaps—and since these deal more damage than their natural claws, they’re the preferred mode of attack. Indeed, since tortles’ Claws attack deals only 1d4 + 2 damage, they’ll prefer any melee weapon deadlier than a dagger over attacking unarmed, as long as they’re proficient with it. You can therefore choose to arm an individual tortle with any melee weapon you like. As long as they’re wielding a versatile weapon such as the quarterstaff, they wield it two-handed, since they don’t carry shields.

Tortles also carry light crossbows, but they’re not as good with these (or any other ranged weapon) as they are with a staff. They use their crossbows only against enemies who won’t come within their melee reach or who repeatedly slip away by using a feature that denies opportunity attacks.

When melee ensues, tortles slug it out until they’re moderately wounded (reduced to 15 hp or fewer), at which point they use their Shell Defense action to hunker down. They remain hunkered until their assailants go away—or, if those assailants keep attacking, until they’re seriously wounded (reduced to 8 hp or fewer). At that point, they realize that their shell is insufficient to keep them safe, and they pop back out (bonus action) and make a run for it, Dodging as they flee. Tortles are surprisingly spry, and their high Armor Class makes the Dodge action a strong choice: With disadvantage on their attack rolls, most attackers will be at an effective -5 to hit them.

The tortle druid is lawful neutral rather than lawful good. Presumably preferring wilderness treks over urban sightseeing, it’s less friendly than most other tortles; its default attitude is indifference. It may be talked into casting cure wounds as a boon, but it will expect suitable payment for the service.

Since its Wisdom is slightly higher than its Strength, its preferred method of dealing with a belligerent foe is spellcasting, and if they fail their saving throw against hold person, the tortle druid considers the problem solved: It simply walks away from its paralyzed enemy. Only if seriously provoked—by an intractably hostile opponent, or by chaotic characters chaosing all over the place—does it take advantage of that spell’s effect to deliver a beatdown with its staff.

A tortle druid can figure out relatively quickly—say, after just one round of combat—whether it’s a match for its opponents. If they’re level 2 or lower, it’s happy to do whatever it takes to teach them a lesson: It casts hold person to stop a lone assailant, thunderwave when it’s double- or triple-teamed, and produce flame to answer ranged attacks from less than 30 feet away. If none of these conditions applies, it attacks with Quarterstaff (two-handed, just like non-druid tortles), preferring paralyzed targets over nonparalyzed targets. When it reduces a foe to 0 hp with its staff, it merely knocks them out, unless it’s annoyed beyond the limits of its patience.

On the other hand, if its opponents are level 3 or higher, it uses hold person to neutralize threats and thunderwave to get multiple melee opponents off its back before beating a hasty retreat—also using the Dodge action. It uses Shell Defense when moderately wounded (reduced to 23 hp or fewer) and flees when seriously wounded (reduced to 13 hp or fewer).

FIRENEWTS (V)

Firenewts are quasi-humanoids adapted to conditions of extreme heat, and they display the corresponding fiery temperament: aggressive, wrathful and cruel, according to Volo’s. They’re raiders, slavers, and zealots. If you encounter a small band of them, they’re probably looking for captives. If you encounter a horde, they’re on the warpath.

Firenewt warriors have above-average Dexterity and Constitution and merely average Strength. Despite this ability contour, they fight as brutes, because Dexterity is their primary offensive ability (they wield scimitars, a finesse weapon) and because they lack any feature that would adapt them especially well to skirmishing. They also wear medium armor and carry shields, and they have Multiattack.

They’re not bright. With an Intelligence of 7, they show no imagination or adaptability in their tactics, essentially fighting like primitives. Nor do they discriminate among targets. ’Tis always a fight to the death for them, so ’tis always one for ye, says Elminster in Volo’s, but I’d consider this optional, not gospel. It’s true that they’re described as fanatics, so they may well fight to the death out of conviction. But their Wisdom of 11 is high enough that they can be presumed to have a normal survival instinct. I might split the difference and say they’re more likely to fight to the death when they’re on some kind of mission, in the company of other firenewts; if they’re just minding their own business, they’ll Dash away if seriously wounded (reduced to 8 hp or fewer).

Firenewt warriors have two attacks: their Scimitar Multiattack and the Spit Fire action. Spit Fire can be used only once per encounter, and it does more damage on average than the melee Multiattack, dealing partial damage on a successful saving throw as well as full damage on a failed save. So the simple pattern of a firenewt warrior attack is to charge, use the Spit Fire action as soon as they come within 10 feet of their targets, then Multiattack for the remainder of the engagement. If a firenewt warrior is somehow captured, however, it regains Spit Fire after an hour of inactivity, the equivalent of a short rest.

If you’ve got firenewt cavalry, they’re riding giant striders, basically beaked dinosaurs. These mounts move at a speed of 50 feet and have the recharging Fire Burst action, which they use whenever it’s available against any two enemies who are within range and close enough together (20 feet apart or less). Otherwise, they Bite. Since their Dexterity modifier is the same as a firenewt warrior’s, I’d have both mount and rider act on the same initiative count, even though the giant striders are technically independent mounts.

Giant striders have only animal Intelligence, therefore no capacity for zealotry, so they invariably Dash away if seriously wounded (reduced to 8 hp or fewer), taking their riders with them unless the riders are feeling fierce enough to dismount and keep fighting. (Remember that dismounting costs a rider half its movement.)

A platoon of firenewt warriors may be led by a firenewt warlock of Imix, the spellcasting variant of the firenewt. These, interestingly, fit the brute ability profile more closely than firenewt warriors do, and they carry a melee weapon, but their primary offensive ability is Charisma. Since they can cast mage armor at will, without spending a spell slot, they always have it cast upon themselves before initiating battle.

Firenewt warlocks have only two spell slots available to them, and at least one is always saved for hellish rebuke, which is cast as a reaction to taking damage. As for the other slot:

Flaming sphere requires a Dexterity save and deals half damage on a success. It’s best against at least two targets but in practice often ends up being used against just one. Because the spell is continuous, it can be used to chase an enemy around. The fact that it’s controlled as a bonus action enhances the caster’s action economy.

Scorching ray is a ranged spell attack that can be cast at up to three targets; its chance to hit is 5 percentage points greater than the chance of hitting with a melee weapon. On a hit, each ray deals an average 7 damage, for a total of 21.

Burning hands also requires a Dexterity save and deals half damage on a success. On a failed save, a target takes an average of 14 damage. The 15-foot conical area of effect indicates use against at least two targets.

Here’s our heuristic, then: If one enemy stands out as obviously, exceptionally dangerous (the only way for firenewts to judge this is by their size, so to them, obviously, exceptionally dangerous equals the 6-foot-8-inch, bellowing barbarian, not the quiet wizard in the back), the firenewt warlock casts flaming sphere as its first action, using its bonus action on the same turn and on subsequent turns to ram the sphere into that enemy. If no such enemy stands out, but there are at least two front-line fighters or supporters close enough to one another and to the firenewt warlock that it can catch them in a 15-foot cone, it casts burning hands; if it can’t do so this turn but will be able to do so on its second turn, it uses its first turn to position itself accordingly, also making a melee weapon attack if the opportunity presents itself. If the conditions aren’t right for either flaming sphere or burning hands, it casts scorching ray at up to three opponents in the enemy back line.

The only other spell the firenewt warlock casts is the cantrip fire bolt, a ranged alternative to its Morningstar melee attack. Which attack it makes is a matter of battlefield positioning. The firenewt warlock generally favors spellcasting over melee attacks, casting fire bolt at back-line opponents, but it’s not shy about clobbering any foe who comes too close.

Firenewt warlocks share the firenewt warrior ideology and flee—or not—according to the same criteria.

GNOLLS (V)

There are four new gnoll variants in Volo’s: the gnoll hunter, the gnoll flesh gnawer, the gnoll witherling, and the flind.

The gnoll hunter has high enough Dexterity to qualify it as a shock attacker and high enough Wisdom to identify weak targets, and it has proficiency in Stealth. Its longbow attack also has the added virtue of slowing its target’s speed by 10 feet on a hit, giving it the combination of attacking from hiding, then running down its target to finish it off with melee attacks.

The gnoll flesh gnawer has a high Dexterity, but without a ranged attack, it’s clearly a shock attacker, not a sniper. Also, it has the Sudden Rush feature, which lets it charge into melee faster; the exemption from opportunity attacks allows it to charge right past its opponents’ front line to get at weaker targets in back. Its merely average Wisdom suggests that it’s not that choosy about which weak back-liner it goes after; anyone not willing to engage on the front line is fair game.

The gnoll witherling is undead, and it gets an extra attack as a reaction whenever one of its allies is killed. It’s also immune to poison and exhaustion. That’s it.

The flind has a magic flail. The flail has three different powers, but the flind doesn’t get to choose among them; instead, its Multiattack comprises one use of each. The flind’s Aura of Blood Thirst power also grants gnoll allies within 10 feet of it an extra bite attack as a bonus action. That ups their damage, but it doesn’t affect their tactics in any way. It affects the flind only to the extent that it benefits from being surrounded by lots of allies, as opposed to running off on its own, so that’s where it stays.

ORCS (V)

Volo’s gives us fantastic resources to use when designing bases for kobolds and hobgoblins and planning out how they’ll behave in a combat encounter. When it comes to orcs, though, Volo’s ducks these topics, instead giving us an anthropological (orcological?) overview of the highly theocentric structure of orc society. This offers some guidance on encounter building but no new insight on how orcs might behave during a fight, save one detail: war wagons.

We can surmise that a group of orcs escorting a war wagon are less likely to charge Aggressively if doing so means leaving the war wagon unattended. Also, as reluctant as orcs are to retreat to begin with, they’re even more reluctant if doing so means abandoning a war wagon. To allow a well-laden war wagon to fall into the hands of an enemy by fleeing would be unforgivably disgraceful. Orcs are so hung up on pride and valor, they don’t even use the war wagon for cover if they’re seriously wounded. If they have 1 hp left, they place that 1 hp between the enemy and the war wagon.

What Volo’s does offer are five new varieties of orc, two of which are spellcasters, all of which build on the orcish pantheon of the Forgotten Realms setting.

The orc Blade of Ilneval is a front-line battlefield commander. It has three distinctive features—Foe Smiter of Ilneval, Ilneval’s Command, and Multiattack—that together make its combat sequence surprisingly simple.

Foe Smiter grants an extra die of damage when the Blade hits with a Longsword attack. Well, gosh golly, seems like the Blade will be favoring its Longsword melee attack over its Javelin ranged attack, then. The Blade’s Multiattack grants two weapon attacks (either both with Longsword or both with Javelin), plus one use of Ilneval’s Command, if that feature is available. Ilneval’s Command is a recharging ability that grants attack reactions to three of the Blade’s allies within 120 feet, a distance that allows the feature to work at battlefield scale.

Just about all we need to know about the orc Blade of Ilneval is right there: First, it charges up to the front line and fights there, using its longsword. Second, whenever it can, it also confers attack reactions on three of its allies at the end of its turn. The only other meaningful difference between a Blade and an ordinary orc is that it has slightly higher mental abilities. It can tell when a certain tactic isn’t working and make adjustments, it leads the pre-combat parley (which, for orcs, consists mostly of taunting), and it prioritizes the targets that seem most threatening.

The orc Claw of Luthic is a support caster, although it can also do wicked melee damage with its claws, so it’s not going to hang back as spellcasters often do.

The key spells in the Claw’s arsenal are bestow curse, warding bond, bane, cure wounds, and guiding bolt.

Bestow curse is an interesting one. The Claw can sustain only one instance of it at a time but can cast it twice. Suppose it chooses the option to add necrotic damage to its attacks against the target. The Claw’s Multiattack grants it four melee attacks as long as it has half or fewer of its maximum hit points. Every one of those attacks, if it hits, will deal an extra 1d8 necrotic damage. One way for the Claw to use this spell, therefore, might be to charge the desired target using the Aggressive feature when seriously wounded or nearly so, cast bestow curse upon them, and finally, in the following round, make a four-Claw Multiattack, hopefully dealing bonus necrotic damage with each strike. However, this tactic has two problems. First, it’s slow: It takes a full turn to set up, and if the Claw defeats its foe, it takes another turn to set up for the next one as well. Second, the extra 1d8 damage is less than the damage it deals on a normal hit (1d8 + 2), so rather than cast bestow curse, it’s better off using its action on another Multiattack. Consequently, since its Dexterity and Constitution are greater than its Strength, it casts the disadvantage-on-attacks option against its desired target instead. The Claw has fairly high Wisdom, so it’s choosy about its target, favoring enemies who seem likely to have low Wisdom themselves, such as bards and rogues. Alternatively, when it’s not yet wounded that badly, it chooses instead to target a fighter already engaged by one of its allies with the turn-wasting curse option, which has greater impact the more Extra Attacks an opponent has.

Warding bond, cast on a more powerful ally, effectively lets the Claw take half the damage that would normally go to that ally; it also provides nominal boosts to Armor Class and saving throws. Because it doesn’t need concentration to sustain, and because it’s useful right away and remains useful as long as it’s in effect, it’s the kind of spell a caster would normally cast first thing in a combat encounter. The trouble is, once again, it takes a full action! If a Claw uses its first turn to cast warding bond and its second turn to charge and cast bestow curse, the combat encounter is already well on its way to being over. As the DM, you can sleaze this by deciding that the Claw has already cast this spell on an ally before the combat begins, which means it has one fewer 2nd-level spell slot to work with, but if I were to do this, since I’m an honest sort of guy, I’d first want to make sure the Claw had some reason to expect that combat was soon to break out. Maybe the Claw casts this while its commander is parley-taunting. Maybe, in fact, part of the reason for the parley-taunting is to give the Claw cover while it casts warding bond? In any event, if the Claw is casting this spell within view of your PCs, they should be given a Wisdom (Perception or Arcana)I

roll to notice it, as a contest against the Claw’s Dexterity.

Bane requires concentration and therefore is no good if the Claw is sustaining bestow curse. However, since it targets Charisma rather than Wisdom, it may be useful as an alternative to bestow curse if the latter seems unlikely to take hold. If the Claw casts bane as an alternative to bestow curse, it may as well do so at bestow curse’s level (i.e., 3rd), allowing it to target five opponents.

Cure wounds will be cast on any ally that’s seriously wounded, which I define as being reduced to 40 percent or fewer of its maximum hit points. If the target is a powerful ally, the Claw casts cure wounds using a 2nd-level spell slot, since warding bond uses just one of these slots and can only be cast on one target at a time. If the target is a grunt, the Claw uses the normal 1st-level slot.

Guiding bolt is a nice prelude to an Aggressive charge, dealing an expected 7 radiant damage and giving the Claw advantage on its next attack roll. It’s not quite as good as bestow curse, though, so the Claw will probably use this spell only if the combat encounter drags on and it runs out of higher-level spell slots.

The orc Hand of Yurtrus is also a spellcaster, whose role varies more between support and assault. It can cast silence or blindness/deafness to shut down enemy spellcasters (the former is absolute and affects a whole area; the latter allows a Constitution saving throw and affects only a single target, but it can be used against anyone, not just a spellcaster, and it gives them disadvantage on attack rolls and their opponent advantage on same). It can cast inflict wounds, boosted to 2nd level, to deal 4d10 necrotic damage on a hit—much better than the 2d8 necrotic damage of the Hand’s normal melee attack, with a better to-hit modifier to boot. It can cast bane in the same manner as the orc Claw of Luthic, if there’s no spellcaster among its opponents who needs to be stifled.

The most important thing to keep in mind about the orc Hand of Yurtrus, I think, is that despite the Aggressive feature and its lack of any ranged attack, it’s not nearly as well suited to toe-to-toe melee combat as orcs usually are. Of its physical abilities, only its Constitution is high. It doesn’t float like a butterfly or sting like a bee. All its damaging eggs are in the one basket labeled inflict wounds. So while it still charges its enemies, because that’s what orcs do, it’s relying on that single spell to take its opponents down rapidly. It casts it twice using 2nd-level spell slots—three times, if there’s no need for silence, blindness/deafness, or bane—then keeps casting it at 1st level. It simply doesn’t have any other, better way to hurt its opponents, so it uses this one over and over.

The orc Nurtured One of Yurtrus has three distinctive features: Corrupted Carrier, Corrupted Vengeance, and Nurtured One of Yurtrus. The first two make the Nurtured One a walking plague bomb: As an action, it can voluntarily cease to exist and explode, splashing toxic bodily fluids over a 10-foot radius. If anyone in this radius fails a DC 13 Con save, they’re poisoned, giving them disadvantage on attack rolls and ability checks, and they take 4d6 damage. In contrast, the Nurtured One’s Claw melee attack—it gets only one per turn—deals a measly 1d4 + 2 slashing plus 1d4 necrotic damage. The Nurtured One’s sole purpose on the battlefield is to sacrifice itself for the team: It charges into the midst of its enemies and detonates itself where it will contaminate at least two of them, preferably more. In contrast, Nurtured One of Yurtrus is kind of a pointless feature: It simply grants the Nurtured One advantage against being poisoned or infected by someone else’s filth. If the Nurtured One is fulfilling its intended purpose in combat, its opponents will never get a chance to do so.

The orc Red Fang of Shargaas is an assassin. Its Shargaas’s Sight and Veil of Shargaas features allow it to cast darkness at will, without the material components, and to see through the magical darkness unimpeded. Its Slayer feature gives it advantage on attacks against targets that haven’t taken a turn in combat yet, and any hit against a surprised target during this turn is a critical. Since its Scimitar melee attack does more damage than its Dart ranged attack, it uses this weapon if it can. The Red Fang, alone among orcs, doesn’t have the Aggressive feature, but the Dash bonus action of its Cunning Action feature has the exact same effect, so you can assume the Red Fang behaves the same way, Dashing (bonus action) toward its foe before Multiattacking with Slayer. It actually doesn’t help the Red Fang to drop darkness on its target before doing this. First, the advantage gained from attacking a blinded target is no better than the advantage gained from Slayer; second, although darkness might frighten and disorient the target, it removes the element of surprise, so that Slayer no longer turns hits into crits. Instead, the Red Fang uses darkness as a defense when its presence is already given away and it no longer has the element of surprise.

There is one other creature listed alongside the orc variants in Volo’s, the tanarukk. But the tanarukk is a straightforward Rrrrraaaaahhh, stab stab stab brute, whose features do nothing except increase its damage resistance and ferocity. It’s stupid and savage and has no self-preservation instinct, fighting until either its enemies are destroyed or it is. You can run this one on autopilot.

SKULKS (M)

Skulks are former travelers who somehow stumbled into the Shadowfell and got so lost that they eventually lost themselves. Summoned back to the material plane to perform tasks for their summoners, they do their jobs but also take their frustrations out in all sorts of inappropriate ways. They’re invisible except in mirrors, to children, and in the light of a particular, outlandishly extraordinary candle, and even when they can be seen, they no longer have any individual identifying features.

Their stat block establishes them as chaotic neutral. The flavor text’s description of their behavior is unambiguously chaotic evil. Decide from the outset which way you want to go. A chaotic neutral skulk performs its tasks in the most slapdash manner possible and throws everything into disarray wherever it goes, but it’s not bloodthirsty. A chaotic evil skulk is a murder machine—a messy murder machine.

The skulk’s ability contour has only one peak: Dexterity, which is extraordinary. Everything else is average or below. This contour marks the skulk as a hit-and-run attacker that deals as much damage as it can in a turn or two, then retreats. Because it’s (mostly) invisible, it can withdraw from melee when it feels like it without subjecting itself to opportunity attacks.

With 120 feet of darkvision, it has a strong preference for dim light or darkness, but it doesn’t require these conditions. Importantly, unlike the invisibility spell, Fallible Invisibility doesn’t end if the skulk attacks or casts a spell. The skulk is invisible—full stop. This trait is subject to only three workarounds, one of which (the candle) is a horror story all its own and is so rare and specific that finding it is likely to entail an entire side quest. The other two (mirrors and kids) are unusual enough in D&D adventures that as a DM, you’re almost forced to make sure they’re included. Otherwise, you have an adversary that’s completely invisible, all the time.

Not only that, the skulk is Trackless: You can’t identify its location by looking for footprints. You can listen for it, as well as watch for jostled objects in its vicinity, so it’s not entirely impossible to figure out where it is. But without see invisibility, faerie fire, or something similar, you’ll always have disadvantage on attack rolls against it, and it will always have advantage on attacks against its foes.

Most