As I’ve mentioned in previous posts, players controlling steeds, sidekicks, pets or summoned creatures on the D&D battlefield is a bit of pet peeve of mine. I don’t like it when one player has dramatically more ‘screen time’ than all the others (like most players, whether they admit it or not, I get more satisfaction from playing my character, than watching others). Familiars, steeds and conjured creatures reduce one’s own percentage of the D&D action pie, and I really can’t get invested in the character development arc of your pet gerbil.
I came around to Conjure Animals, however, when a player I was DMing got his party out of a tricky situation, by summoning eight wolves to fend off the much larger pack of wolves that were attacking them en route to Candlekeep. He even managed to do it without noticeably eating up the clock. In truth, he got me out of a hole too, as my meticulously planned murder mystery story wouldn’t work so well if the detectives got TPK’ed before they could even discover the death of the Keeper of Tomes, let alone investigate it.
Conjure Animals is an ace up the sleeve kind of spell that is as versatile as it is powerful, and since ‘discovering it’ I’ve been casting it a lot as both a Circle of the Moon druid and a College of Lore bard. When we were jumped by over a dozen devils in a recent session, I used the eight wolves trick to soak up most of their first wave of attacks, making a potential TPK a relatively comfortable encounter; when we needed to interrupt a ritual, but couldn’t break through a metal grill, I simply summoned my wolves on the other side of the grill to rip apart the evil warlock in question; and when a catoblepas death-rayed our barbarian in round 1, leaving just me and a wizard (with no offensive spells and an AC of 9, because ‘roleplay’) standing, the four crocodiles I summoned grappled and chewed the monster to death as we watched on from a safe distance.
Conjure Animals is a get out of jail card, in so many tight situations, and given that D&D is usually more fun when you don’t all die, I’ve found judicious use of it to be a boon at the table.
Problems With Conjure Animals
But while I’ve been enjoying using it a lot, the spell undeniably has a lot of design problems.
- Rules as written, it’s the DM who chooses the beast type (after the player first chooses the number of beasts, as per the CR options). This a) kills player creativity and ingenuity; b) can set up a situation where the DM tries to pervert the player’s intentions; and/or c) adds to the DMs creative and technical workload, slowing down the game.
- The spell is probably too powerful, esp. against foes that don’t have ranged weapons (to break the caster’s concentration). I’ve tried to only cast it in truly desperate moments, but a less discerning player might find themselves using it as their go to spell during even routine encounters, watering down the contribution of the rest of the party.
- Summoning the eight beast option, in particular, is a significant time drain (see my opening paragraph).
- In terms of combat effectiveness, it’s nearly always the best option to summon eight beasts of CR 1/4, a) underscoring the above point, and b) making the other options a bit pointless.
- There are so many beasts to choose from (and they’re not ordered by CR in the Monster Manual), meaning decision paralysis and further time wasting can come into play, as players (if they’re permitted by the DM) frantically search for the perfect tool in their humungous toolbox .
- The upcasting options are just going to exacerbate all these problems to the nth degree!
- Beast stat blocks in 5e are nowhere near strong enough to feel realistic… leading to the head-scratching moment you summon a rhino and it doesn’t give you any advantage in charging down a door (for example).
So what can we do about these problems? Here are some potential solutions…
1. Change number of summoned creatures…
My gut feeling is that the number of beasts you can conjure is simply not weighted correctly. While this might seem drastic, it probably should be more a choice between: 1 x CR 2 creature, 2 x CR 1 creatures, 3 x CR 1/2 creatures and 4 x CR 1/4 creatures.
But let’s test that gut feeling a bit more rigorously, by comparing some ‘classic’ summoning options of the spell as written, and see just how the different CR tiers compare. There’s so many factors to consider that it’s impossible to do a complete comparison, so I’m gonna keep this real simple… I’m going to compare the number of hit points for each CR option, and the average damage done if every creature hits with all its attacks.
1 x Polar Bear (CR 2): HP = 42 / Damage Potential = 21
2 x Brown Bears (CR 1): HP = 68 / Damage Potential = 38
4 x Black Bears (CR 1/2): HP = 76 / Damage Potential = 48
8 x Wolves (CR 1/4): HP = 88 / Damage Potential = 56
Now, as I mentioned, there’s all kinds of factors I’m not taking into account here, like slightly better ACs for the higher CR creatures, potential bonus action or charge attack damage (if you were to select elks, tigers or panthers for example), better / worse to hit modifiers, and abilities like pack tactics, constrict (giant constrictor snake), web (giant spiders), or swallow (giant frogs and toads).
But I’m happy this gives a broadly accurate picture of how the options compare, as, while the higher level CR beasts have better to hit mods and AC, there’s an inherent advantage to having more beasts to control.
Now I’m curious what it would look like if we follow my instincts.
1 x Polar Bear (CR 2): HP = 42 / Damage Potential = 21
2 x Brown Bears (CR 1): HP = 68 / Damage Potential = 38
3 x Black Bears (CR 1/2): HP = 57 / Damage Potential = 36
4 x Wolves (CR 1/4): HP = 44 / Damage Potential = 28
That looks more balanced to me. The first and fourth option are now nearly identical (you might argue that the wolves’ lower to hit bonus, and the fact that their damage potential will go down as individuals die, is evened out by their pack tactics and the greater number of opportunity attacks they will be able to make). While now the stand out options are the middle two options, both of which would see a significantly smaller ‘time tax’ levied on the table when the spell is cast, compared to the classic ‘I conjure eight wolves’ in RAW.
(You could balance this revision further by saying each beast gains temporary hit points equal to your spellcaster modifier when they are summoned, so that the four wolves and three black bears gain a little on the premium option of two brown bears).
Of course, you could compromise between RAW and my revision: maybe allowing 1 x CR 2, 2 x CR 1, 3 x CR 1/2 or 5 x CR 1/4 creatures to be summoned… or if you’re not too bothered by the spell’s time tax: 1, 2, 4, 6, is still fairer, and means you could possibly introduce 8 x CR 1/8 as an option. This latter option probably pushes people towards conjuring 4 creatures, which I think is a nice balance of the fantasy of summoning creatures to your aid, while not bogging down the table too much as a result.
For upcasting, I’d allow the 5th level spell slot to pick one CR 3 beast, or summon one additional beast on top of any of their existing options. For 7th level, I’d allow one CR 4 beast, two CR 3 beasts or two more of the existing options etc.
2. Players get to choose which beasts… but from a limited pool
It really doesn’t make sense for DMs to choose the animals, and, speaking as a player, having the DM take that decision off you is far more of a fun ruiner than any power downgrade. If you summon eight giant frogs to swallow the kobolds for comedy effect, and the DM gives you eight riding horses instead, just to be contrary, the fun is lost and you feel cheated.
On the other hand, I myself have been guilty of slowing the down the game as I try to work out whether a wolf’s pack tactics is going to work out better than a panther’s situational bonus action attack, or if maybe I’d be better off conjuring the constrictor snakes or the giant frogs and get grappling. I’ve so far overlooked the elk, which does potentially the most damage and has the benefit of being a large creature (great if you’re looking for some meat shields to conjure between you and the enemy). And that’s even assuming I’ve settled on the eight creature option…. maybe I should just summon two giant eagles for my buddies, wild shape into one myself, and we all fly off into the sunset.
I think the game would really benefit if players had a limited pool of creature to choose from when casting conjure animals, and I think that number should be equal to the spellcaster’s level. That’s five at level 5, when you pick up the spell, so still plenty of room for versatility.
The effect of this change is that players will be forced to do their little ‘optimal damage dealing’ calculations and tactical planning BEFORE the session, rather than slowing down the game the minute its supposed to be at its most dramatic. DMs might also appreciate there being a limit on the surprises that can be sprung on them, while everyone will appreciate that the caster is going to know their summoned creatures stat blocks better and the game won’t have to keep pausing to find out the AC, movement, attack modifier etc. of a giant boar.
It just makes sense.
Optionally, the DM could limit the caster’s choices to beasts they might have seen or studied.
Conjure Animals: Best Beasts by CR
While attacking is likely to be your most common tactic, you can use your conjured critters as meat shields, for battlefield control, for mounts (in and out of combat), for intimidation and no doubt a host of other reasons. The one thing you need to remember is that, being a concentration spell, they could disappear any minute (so flying mounts during combat is risky… especially if you’re applying some realistic falling damage in your D&D).
Let’s break down some of the best beast options and what they’re particularly good at.
CR 2 Beasts
As we’ve seen, conjuring a single CR 2 beast is your weakest option in most circumstances, but there are some fun ones to consider when the time is right.
Polar Bear. With two attacks (with +7 to hit) come what may, this is your best damage dealing option. If its dark, switch to a Cave Bear.
Sabre-Toothed Tiger. Ten more HP than a polar bear, the smilodon needs to succeed in knocking its target prone to get two attacks. Others can also benefit from a prone enemy, however, so it can work out better than the polar bear, esp. if you plan on ganging up on one opponent.
Giant Constrictor Snake. If you want to entangle a single target, this serpent’s constrict action is a +6 attack that does 13 damage and automatically applies the grappled and restrained conditions (which is considerably more powerful than knocking prone). Comes with a sack of hit points.
Rhino. Statistically not your best bet, although the temptation to conjure it 60 feet above your enemies’ heads may be too hard to resist.
Other options include: allosaurus (similar statblock to sabre-toothed tiger), aurochs, giant boar, giant elk, hunter shark, plesiosaurus (swimming) and quetzalcoatlus (flying).
CR 1 Beasts
If you’re using the Hipster Remix of the spell, then two CR 1 beasts is likely your best option. Luckily there are plenty to choose from here:
Brown Bear. 34 HP, two attacks, plus a climb speed make the grizzly a solid choice.
Tiger. Unlike the bear it has to move 20 feet, land its weaker claw attack, and the enemy has to fail its saving throw, for the tiger to get a second attack. Does have 37 HP and dark vision though.
Dire Wolf. Similar stats to the tiger, without the situational bonus attack. However everything else is better: 50 ft speed (vs. 40 ft), AC 14, stronger bite attack, plus pack tactics.
Giant Toad. 39 HP, decent bite damage, plus auto grapple/restrain on a hit. Can even swallow a grappled Medium-sized target on a second hit! Just 20 feet movement on land.
Giant Spider. Weak HP (26), but AC 14, decent bite (with poison), plus blindsight, darkvision and spider climb could all come in handy. Its ranged web attack could be used to good effect.
Giant Eagle. Crazy fast (80 feet) with two attacks, the giant eagle is a great choice for taking out enemy casters or generals. Why not sacrifice one (or both) attack(s) for a grapple attempt and hoist a victim or two into the sky? Cue epic falling damage.
Other options include: deinonychus (three, potentially four attacks, but all puny… good option against goblin hordes?), giant hyena, giant octopus (10 feet movement probably means they are a bad choice on land, but they are a go-to water option), giant vulture (v. similar to giant eagle, but slower and significantly worse AC. Does have pack tactics though!) and lion (unless you want to leap across a chasm, dire wolf is better version of a v. similar stat block).
CR 1/2 Beasts
Lots of cool options here, although if you’re playing RAW you can skip to the CR 1/4 creatures.
Ape. A poor choice in most circumstances, the ape is the only beast with a ranged attack (‘rock’) that deals damage. It’s rarely gonna be better than having a flying beast doing a flyby however…
Black Bear. Two attacks and a climb speed. Note, it’s only Medium sized (other bears are Large).
Crocodile. The auto-grapple and restrain on a hit can prove handy indeed, while being able to hold its breath is particularly handy against catoblepases, in my experience. Still, you might be better off with constrictor snakes or giant frogs (see CR 1/4 beasts).
Giant Wasp. Could do some nasty damage to those that fail their Con saves. Flies, obviously.
Warhorse. Has one strong attack (2d6+4), with a decent chance of getting a second go at it per round. Large and v. fast.
Other options include: giant goat, giant sea horse, reef shark.
CR 1/4 Beasts
Creatures in this tier tend to only have 10-11 hit points, but given how many you’re getting you’ll want to choose from the selection below more often than not.
Axe Beak. Large, fast and above average hit points (19). Poor damage.
Constrictor Snake. Auto-grapple and restrain on a hit. Comparable to giant frog below.
Elk. Large, fast, and potential for decent damage on a charge. Plus the knock prone feature means other elks can pile in with advantage.
Giant Frog. Poor to hit and damage, but good HP and another auto-grapple and restrain. While the grapple is not as good as the constrictor snake, the giant frog can swallow Small creatures.
Giant Poisonous Snake. Good damage if victims fail Con saves against poison.
Giant Owl. It flies and does decent damage. If the enemy doesn’t have range attacks they’re going to struggle to take these guys down, as they have the flyby ability.
Panther. Its ‘pounce’ is inferior to the elk’s ‘charge’, but has Stealth, a climb speed and better AC. Plus is way cooler.
Stench Kow. A variant of cow, tucked away in Volo’s Guide to Monsters, its similar to the elk only without the knock prone ability, but with a nasty stench effect instead that could slow down the enemy (or your allies if you’re not careful!).
Velociraptor. Two attacks and pack tactics make this a very respectable choice, with reliable damage and a decent AC too. As they’re tiny you can completely encircle even a Medium-sized foe.
Wolf. Decent bite damage, knock prone ability, and pack tactics places this, more or less, alongside elks and velociraptors. One of my DMs plays stacking advantages, so prone + pack tactics becomes a very reliable way of registering hits.
Other options: boar, cow, draft horse, giant badger (good if you need to burrow for some reason!), giant bat (could be useful in combination with magical darkness… otherwise giant owl better), giant centipede (same as giant poisonous snake but much less HP), giant lizard (if you need multiple mounts that can climb, here you go), giant wolf spider, riding horse.
Your Experiences with Conjure Animals?
I have a feeling one or two of my regular readers will already have a house rule or two pertaining to this spell, so do please share guys. New readers, the comments section is below…
Oooh it’s a tricky one isn’t it. As a DM I would probably prepare a shortlist of beasts that match the environment the players will be in, and ask the caster what they want from the spell when they cast it. Then I will try to match what they want from the beasts available so they both get what they want, and get to be surprised by what turns up.
Like, “I want something big to hold that thing down so we can escape”, or “I need lots of things to engage those archers on the hill until we can get close”, or “I need something fast, something that can swim, something that can carry poor dead Eric etc…”
That or make each conjure beasts variation a different “spell” that the player has to negotiate/win from a spirit. Get on good terms with the wolf totem? Your conjure spell gives you 8 spirit wolves. You only get 8 wolves until you build a relationship with another totem.
I’d probably go with whichever option the player prefers. Some would love the RP option, others would like to be surprised.
Thanks for the comment Daniel.
I feel like, by specifying that the spell conjures fey spirits and not actual animals, the designers successfully got around the whole ‘woah, what’s a polar bear doing in a jungle?’ and ‘how did it get here so fast…?’ (hence conjure, not summon). So I don’t think the animals have to fit the environment…
I still don’t see any benefits to the DM choosing the beasts, or getting involved here at all. I am casting a spell… I don’t want to be surprised. Just like I don’t want the DM to decide what illusion I cast when I cast Major Image. Or what battlemaster maneuver I select when I decide to use one…
There’s no role for the DM here. This is a player taking their action.
My guess is the PH only specifies the DM makes the decision because the designers figured new players would be overwhelmed, while know-it-all DMs could cope more easily. However, I don’t think that’s the case and so that part of the spell should be revisited.
I think the 1/2/3/4 is a no brainer, to be honest. 8 x CR 1/4 is BOTH the most powerful choice in almost all circumstances AND the most time-hogging. Lots of people seem to have adopted the 1/2/3/4 house rule over the years, and reported it works a lot better and is far more balanced.
We also play that you can only give one, simple instruction to the beasts, given time constraints and their extremely low intelligence. No more advanced tactical manoeuvring.
I think the question of “DM decides or rolls randomly, based on regional ecology” vs. player choice needs to be thought through between players and DMs, taking into account player and DM game knowledge (and thus speed of response), the level of fun in surprise vs. the fun of tactical flexibility in player choice, whether you’re playing online or offline (online makes it much faster), the level of trust that the DM wants the player to have maximum fun vs. risk of antagonism, etc.
Ok, I honestly didn’t know this was a common house rule!
I thought people would be against it, since it is a major downgrade… and I know how that normally goes whenever I propose one! 🤣
While the whole ‘fey spirits’ thing could imbue them with more than animal intelligence, I do agree, that having the beasts carry out basic instructions only somehow feels more credible and more balanced. Maybe the level of a well-trained dog would be a good top limit?
I still don’t see what this has got to do with the DM.
As I mentioned in a previous comment, I think this was some kind of prediction on how this would work in gameplay that has just muddied the waters.
If it was designed to be a power check, then the designers should and could have just make the spell less powerful (which in any case we’ve just done, so even less reason for the DM to get involved).
I’ve read about a lot of problems with this spell over the years, but no one in my campaign or the one I play have the spell. And my players generally don’t like adding house rules for things they haven’t seen to be a problem yet. So we don’t have a house rule in place yet. (One player just multiclassed to druid, but is only Druid 2 at this point.)
My players tend to travel quite a bit, so while I like the “DM prepares a list of available animals” as an idea, that pretty much means I need to through together a list of animals for every biome, just in case. Last session, for example, they started in a forest, traveled through hills and plains, hunted a beast through a riverside forest into a cave, then crossed the river and travelled through hills again into town.
I like the roleplaying “what animal spirits have you made friends with” idea a *lot*, given the nature of my campaign setting. Unfortunately, though, that will result in “8 wolves” all the time (for the aforementioned fledgling druid)!
My thought, though was to remove the spell, and replace it with Tasha’s Summon Beast. This would work well with your “what spirits have you befriended” idea, too. So you don’t summon 8 wolves – you summon a manifestation of the Wolf Spirit, one that is as strong as you can manage (i.e. spell slot level); in this case, the summoned beast has Pack Tactics chosen by default. And if you befriend the Eagle Spirit, you can summon an Eagle – again, *one* eagle – with flying and flyby attack. If you befriend the Bear spirit, maybe I give you +10hp and +2 damage, but the bear doesn’t get Pack Tactics. And so on.
So you only have access to special abilities that match the spirits you’ve befriended. And you only get one Beast that answers the call. [Maybe I might allow multiple beasts, but I would just take split the HP among them, and lower the attack and damage bonuses by 1 per additional critter. You can have you spirit Wolf with 35hp (30 + 5 for 3rd level to match Conjure Animals), +6 to hit and +7 damage, or two lesser wolves with 17 hp, +5 and +6; or three minor wolves with 11hp, +4 and +6. If you wanted a swarm of something, then maybe I’d just give you *that*: a swarm of rats (yay, back to one creature)!
Even with your (fully justified) nerfing, I still like the idea that the player chooses a creature type to attempt to summon and has to roll an Arcana check (otherwise the DM chooses), with the DC determined by how naturally the chosen creature would turn up in the current location. I’ll admit though that the DC gets very low if I don’t have a ready alternative in mind.
I like the idea that the player has to come up with a restricted list to choose from in advance.
Given that we’re talking about fey spirits and not actual animals I feel like that is a false argument.
Do we have to make an Arcana check to cast fireball in an arctic environment?
Or to cast Entangle in a desert or urban environment?
Certainly I don’t think you can have a restricted list AND an Arcana check.
I think this spell benefits from a lack of area of effect attacks from monsters in 5e.
It certainly makes taking out eight wolves harder…
While a lack of effective long range attacks on most stat blocks makes it easy for the caster to maintain concentration.
As a “back in my day” reminiscence, there was in 3.5e the amazing Psionic power: Create Astral Construct. I loved this power. Super versatile, but mostly balanced, and didn’t require knowing a lot of animals. Here’s the base set of stats; pick one power from Menu A. 3rd level? pick from Menu B, or two from Menu A.
Do you need to glide over a gap? create a construct that can fly. Need a tank? Upcast the spell, and give it better AC and more HP. More damage? Add on Extra Attack and maybe Energy Touch. and so on.
Translated to Conjure Animals, avoid the encyclopedia flipping. Conjure a generic animal with the specific trait(s) you’re looking for, call it whatever you want to call it, and get on with the game!
Hi Rick, interesting… by 5e standards that seems like a v. complex spell, but I can see how, if pared down, it would offer a neat one page alternative to conjure animals – effectively offering a much faster pick’n’mix Swiss army knife than CA.
The spell also reminds me a bit of how Tasha’s handles the Beast Master companions, with three generic beast options that power up as you level up.
On the other hand, I must confess that bland astral blobs or even Tasha’s catch-all ‘Beast of the Land’ etc. don’t inspire me much!
Meanwhile Dylan Sprouse summoning an ‘army of beavers’ who walk and talk like 1920s Chicago gangsters is pretty epic…
I have never run into this being a problem in my games as player or DM because no one has ever chosen it as a spell to cast.
I think I would prefer the 1-2-4-8 as I would be willing to bet that if someone were to game it out that the to hit bonuses and such would probably even out the end results to being almost the same with the choice devolving to do I need something big or do I need more bodies to buy us time or to tear apart the enemy on the other side of the portcullis.
As for the DM choosing what type of creature – I agree that the DM should have zero participation in this.
Rather than being able to choose any creature, I think I would rule that the caster must have an ‘intimate’ knowledge of the creature being summoned. Just because someone has seen a wolf doesn’t mean that they truly understand wolves – specifically things like pack tactics, movement, etc.
I understand these are fey spirits taking the form of the creatures but they are being commanded and the creatures being created/molded according to the knowledge of the caster.
I would probably rule that the caster, in order to gain the necessary knowledge, would need to study each type of creature for a time in order to gain the knowledge to form them out of nothingness basically. My feeling on this would be a scaled study time based on the CR of the summoned creatures.
CR 1/4 2-3 days or more
CR 1/2 1 week
CR 1 2 weeks
CR 2 3 weeks
CR3 4 weeks
and so on.
I think I would also put in a rule that for any creature, that it would require studying them in their natural environment(this would include hunting, sleeping, attacking, defending, interactions with others in its group or possibly even its mate and so on) and for any creature CR 1 or higher possibly dissecting to view internal organs and such in order to gain a ‘greater’ understanding.
Such as the Catoblepas, would need dissection in order to understand the poisonous gas production.
For Polar Bear it might mean living in the artic area and studying them AND not getting eaten.
I know it is a fantasy game and magic is just that, but I try to have some semblance of realism in a game such that if a player says I have heard of a large bear that lives in artic regions so I will make one of those – without the study all they are going to get is a bear that they do know that has white hair.
G Van Booven
I’m late to the party. In a game I played in, we had a player spamming giant poisonous snakes during every encounter, and those things are nasty. When I started DMing and he did something similar, I talked to the player and pointed out that RAW, it was the DM’s choice of creature, but went on to say that it didn’t feel right for me (the DM) to make that call. So we generated random tables for each of the CRs and when the player wanted to cast the spell, he just rolled on the appropriate table. Based on the roll, he had a 1 in 8 chance of getting his giant poisonous snakes. It worked for us.
I’m strict: Are they really a Druid???
No Dinosaur types, no polar bears unless you can show how you’ve seen one beforehand. The younger you are the less choices you have, ie. horse, wolf, bear, eagle, hawk, leopard, jaguar, alligator?,
– if you dislike dinosaur in your timeline
– easier to choose summoned animals if you even need to (list is 70% shorter)
– DM decides what animals are introduced based on events, if DM wants to move up to the next level, increase opponents CR level i.e. Cave Bear or Giant Spider and now character can Wildshape or Summon to the same, allowing level up to meet the challenge. (if they don’t notice, and somehow live, let them know it is now available)
– Number of summoned very small, one to two, no killing off entire ecosystems of animal life over the span of a three week journey.
– Choices allow character to choose what is summoned as all will fit within vague bounds of the adventure
If a druid is summoning in “cannon fodder” while they sit back drinking from their sippy cup, without it being a total wipe situation, they are to be reported to the nearest actual druid to have their Shillelagh kicked.