Friday, December 17, 2021

Send in the animals

Last time I talked about the DnD spell Animate Objects, where you attack your enemies with chairs.  This time I want to look at Conjure Animals to see how that spell stacks up.  Conjure Animals is a *lot* more complicated because it gives a similar choice betwen a few big summons or a ton of small ones, but there is a lot of variety in the summons.  Animated objects always have the same stats, but a wolf is not the same as a giant wasp!

The two spells share the same fundamental flaw - the game is balanced around the idea that attacking once for 15 damage is about the same as attacking twice for 10 damage, which is about the same as attacking 8 times for 6 damage.  This idea is patently absurd, but all the summon spells and indeed the entire Challenge Rating system is based upon it.  Yesterday we saw that animating a ton of small objects is vastly superior to a big one, and with animals it is the same.  Pick the big one, you get a huge, tough critter than hits for 10 damage / round.  Pick 8 small animals though, and each of them hits for 5 / round.

Conjure Animals lets you summon 1 CR 2 animal, 2 CR 1 animals, 4 CR 1/2 animals, or 8 CR 1/4 animals.

Animal              CR     Dmg/animal   Total damage

Giant Elk             2             11                11

Brown Bear         1             9                  18

War Horse           1/2          7                  28

Wolf                    1/4         5.5                 44

You see the problem.  Sure, swarms of animals are at risk of being Fireballed, but they deal preposterous damage.  A fighter at level 5 swinging with a big sword probably does about 20 damage a round, and we don't want summon spells to completely replace fighters.  Given that, I think summoning a Giant Elk that is worse than a fighter but has a ton of health is a reasonable thing.  It lasts a full hour, has as many hitpoints as any characters in the group, and some animals have special abilities like flying, tracking, swimming etc that you can leverage.  It is a spell I would happily cast.

However, summoning swarms of weak animals is a huge problem.  If you max out on stuff like wolves your fighters are going to feel utterly useless.  The wolves do several times as much damage as they do, and if the enemies do decide to start chopping through the wolves that is *great*.  The wolves have 88 HP total!  They will still do a ton of damage and save your group a huge amount of incoming damage too.

The solution here is simply to nerf the number of animals you get when you go for lower CR ones.  Instead of doubling the numbers each step down, I would change it from 1,2,4,8 to 1,2,3,4.  Nice and easy, and it actually works out well with the damage numbers.  The CR 2 is the worst still, but at least the others all deal similar overall damage.

The other issue with this spell is that it is extremely overpowered when cast at higher level.  If you level up Fireball from 3rd to 5th level, for example, the damage goes up 25%.  However, if you level up Conjure Animals from 3rd to 5th level, you get *double* the animals.  This is a problem!  It makes all higher level summoning spells a joke compared to Conjure Animals, and uplevelling spells is supposed to be a way of filling in spots, not invalidating high level spells.  The way I would fix this is to make the spell summon 1 more animal of CR 1 at level 5, or one extra CR 2 animal at level 7.  This is still worth doing, but leaves room for actual high level summon spells to compete.

Even after both of these nerfs Conjure Animals is still better than any other summon spell in the book, I think.  At 5th level it deals more damage than the Bigby's Hand or Animate Objects spells, and it makes Conjure Elemental and Mordenkainen's Sword look like ridiculous jokes on multiple fronts.  Even after my suggested nerfs, I think you have to make the types of animals that are summoned be determined by the GM, and have the GM pick weak animals.  I tried to pick the best animal types for my numbers above, but if you instead pick randomly the spell is a lot more fair.  Many of the animals have worse attack routines than the optimized ones above, and that would reduce its power level reasonably.

I like the idea of summoning animals to attack your enemies.  It is also cool that their abilities are based on the abilities of the base creatures.  As usual, the theme is good, but the math sucks.

Wednesday, December 15, 2021

Send in the chairs

The Animate Objects spell in DnD is thematically great.  You are in the middle of a fight in a workshop, and you cause the bench, a couple of crowbars, and a table to rise up and bash away at your enemies.  You could also use it to help carry people, break stuff, or accomplish other tasks.  Awhile ago MattInTheHat came up with a *fantastic* plan for the spell - he threw some coins at his enemies, telling them to buy some better insults, then animated the coins and bashed them with them.

This was a great dramatic moment.  The coins were marvellously effective.  I eventually picked up the spell myself and paid for a bunch of small metal pieces with my character's initials on them so I could toss them at my enemies and do similar things.  This is all well and good, except that it is hideously unbalanced.

The way the spell works is that you get 10 slots of objects.  The bigger the object, the more slots it takes up, but the more damage it does.  However, tiny objects are a huge outlier in terms of their damage per slot.  Consider the chart below, written against AC 16.

Size    Slots    Damage  Hit      Total Damage  Dam/Slot/Round

Tiny    1          1d4+4     +8              4.225        4.225

Small   1          1d8+2     +6             3.575        3.575

Med     2          2d6+1      +5            4               2

Lrg       4         2d10+2     +6            7.15          1.789

Huge    8         2d12+4      +8           11.05        1.381

The problem is obvious.  The thing that really matters is the damage per slot, and tiny things utterly dominate that metric.  Large things hit hard, but there are so few of them that they do terrible damage.  It feels to me like you ought to want to animate big stuff if you can, but the opposite is true.  You might think that coins are perhaps too small to work, and that is reasonable.  (The spell does not specify.)  However, you could always just have 10 daggers on a bandolier or somesuch, if need be.  

The real thing we ought to figure out is how good *should* the spell be?  I will use the spell Disintegrate to benchmark just how absurd Animate Objects is.  Disintegrate is a single target 6th level spell that does 75 damage with a save to avoid.  Animate Objects, if up levelled from 5th to 6th and used on Tiny objects does 78 damage with attack rolls - quite similar.  However, Disintegrate is a single action, and Animate Objects does that damage every round!  Also Animate Objects uses many attacks that can all smash enemy concentration and also will take attacks of opportunity if they enemy tries to move.

It is clear to me that Animate Objects on tiny objects is completely busted.  It is the best single target attack, and is as powerful as a fighter, but the spellcaster can cast other spells while the objects beat down.

So how do we fix this?  There are two basic approaches - first, to restrict how often you can use it, and second to change the numbers.  The first one would entail insisting that tiny objects be of highly specific sizes.  You could insist that they be the size of a sword, for example, so that having 10 of them on your person is extremely difficult, and thus you could only use Animate Objects under specific circumstances.  Unfortunately there are ways around that - I would probably have a summoned minion carry around a box of swords for me to animate in that case.  This approach also doesn't fix the issue that large objects are terrible.

The other approach is to fix the numbers.  Generally small things have the problem that they have low strength so they do little damage, but these small things have massive dexterity and use dexterity for their damage rolls.  I don't see any reason why we need to shackle ourselves to the listed strength and dexterity numbers.  The tiny object get many attacks, which is especially powerful against casters, so their damage should be the weakest overall.  Large things are harder to come by and are often restricted by their size, so they should hit hard.  Here are my suggested new values:

Size    Slots    Damage  Hit      Total Damage  Dam/Slot/Round

Tiny    1          1d4+1     +5              1.75           1.75

Small   1         1d8          +4             2.025         2.025

Med     2         2d6+2      +6             4.95           2.475

Lrg       4         2d12+3    +7             9.6            2.4

Huge    8         4d12+5    +9            21.7           2.713

These new values look a lot better.  If you have a bunch of small things they are still fantastic at bonking spellcasters many times to disrupt casting.  Their damage is still good, and definitely worth casting.  If you animate big stuff it is worse at disruption, but hits very hard.  (But still not nearly as hard as the base version of the spell on tiny objects.)  Tossing coins at people is a fine thing to do, animating swords does better damage, and finding chairs or tables or anvils to punch people with is better yet.  That feels way more appropriate thematically, and stops this spell from being quite so brutal.  

Having some spells be better than others is fine, but you have to keep the best ones from dominating everything.  Fireball, for example, is so overpowered that it is often the best thing to cast in a single target situation, even though it is obviously designed for AOE.  That is a problem.  Animate Objects, as written, is similar.  It is supposed to be a spell you maintain over many turns, but it is actually the best one shot single target attack.... and then it keeps doing that every turn.

Next I may tackle Conjure Animals, which has many of the same problems that Animate Objects does.  Summoning a horde of wolves or bears delivers absurd damage in the same way, because the game is designed around the idea that 4 attacks for 10 each is about as good as 1 attack for 20.