D&D Next: Spell Design

It’s long established in this blog that I like to talk about spells and magic, and today is another of those days. D&D Next spell design is currently a bizarre melange of new ideas and old spells that have been with D&D for two or more editions. For example, lance of faith and thunderwave are the only spells in the 12-17-12 playtest packet that aren’t also found in 3.x D&D, and those are both from 4e. On the other hand, the mechanics of each individual spell is often strikingly different from its appearance in earlier editions. Chiefly, the designers have done away with spell scaling by caster level, a move I am inclined to applaud. They have also made major changes to spells that grant bonuses or penalties to attack or AC, in keeping with bounded accuracy principles and reduced usage of fiddly math bonuses.

When the playtest packets first came out, I was surprised at the new spell layouts. At least as far back as 2e, spells break many key stats out into their own separate lines at the beginning of the layout: level (okay, 2e handles that in the page header, a mortal sin against information management), name, components, school or sphere of magic, range, area, duration, casting time, and saving throw. Damage values and other details of the spell effect are contained in the text block. 3e adds spell keywords and whether or not Spell Resistance applies, but otherwise the spell format is just about the same.

4e makes some big changes by converting everything to a standardized power format, though the information is mostly similar. The major change is that attack bonuses, defenses, and damage/effect expressions are broken out into clearer pieces of rules text rather than inhabiting a text block of 1-5 paragraphs. For quick-and-easy use, it’s pretty much the tops; what it doesn’t do as well is anything other than mystically-induced facial murder. Utility spells are a bit text-y compared to attacks, while rituals are structured very nearly like 3.x spells.

So this brings us to D&D Next, where spells have only two to four pieces of information broken out: level, school, casting time (that is, whether or not this spell has a ritual form, and whether or not it is a word of power, which functionally if not literally reduces the spell to a minor action), and any higher-slot versions of the spell that can be cast. All other relevant information is contained in the spell’s description and effect blocks. I’m not too sure what this accomplishes for the designers, except that it strongly encourages them to keep it as short and to the point as they possible.

These are basics, though, if you’re the sort of person to read obscure (yet oddly charming) gaming blogs like this one, I have probably not opened your eyes to the wide world. The more interesting point is that they have introduced new rules concepts to differentiate and balance spells. Words of power I’ve already mentioned. Concentration is another; since a spellcaster cannot (currently) maintain concentration on more than one spell at a time, the designers keep a hand-brake on the most complicated and gameplay-slowing effects. Considering that clerics do not lose concentration due to damage taken in battle while wizards do, it also acts as a very hard-line incentive for wizards to stay well in the back. Mearls has recently commented, though, that some change is coming to this mechanic:

Wizards are the best at area attacks and control effects. I have to admit that I’m really happy with how the concentration rule is reining in spell stacking and buffs. My high-level playtests have limited any balance problems to specific spells, rather than the entire concept of dropping five buffs to create a death machine character or using glitterdust/grease/stinking cloud/wall of fire at once to turn an encounter into a joke. We’re looking at breaking the concentration rule into two separate rules: one rule that covers concentration and the chance that damage ends a spell, and a separate rule (tentatively called focus) that limits you to one focus spell at a time.

It’s a good change, because while it gets rid of the inducement for wizards to stay in the back and avoid taking damage, it means that the DM can splash area damage a bit more without constantly wrecking whatever the wizard is currently maintaining.

Most direct-damage spells also carry an internal scaling mechanic, nominally comparable to loading the benefit of 3.x’s Empower Spell metamagic into each spell. Typically the caster gains one additional damage die of whatever size. While I don’t want to get too deep on the game-balance nerdism, the current schema of spell preparation makes this choice a lot more interesting. So look, I’m going to set it off in a quote block, and if this isn’t for you (or is glaringly obvious and you don’t know why I even mention it), there are no hard feelings on my part if you wanna skip it.

In D&D Next, spells aren’t prepared by spell level the way they are in 3.x. Setting aside cantrips and benefits from wizard traditions, wizards prepare (wizard level + 1) spells – so if you are happy to spend all of your casting slots of a given level on amped-up spells or some other way, you can prepare a disproportionate number of spells of a different level. So that’s neat. But while you can prepare just one low-end attack spell and use slots to crank it up to respectable output, spells inflate less efficiently than preparing a higher-level spell does.

The bad side to storing the power-inflation options within the spell’s description is that it’s not easily expansible without releasing rules updates. By way of comparison, Empower Spell is one among many metamagic options in 3.x, and the designers (including third-party publishers) could introduce new options whenever they wanted, at the cost of a precious feat slot or by requiring the use of a metamagic rod (some of the most flexible and powerful magic items in the game).

Many spells, particularly utility effects, have a ritual form that means the caster doesn’t have to spend a spell slot to cast them, just more time. The caster does still have to spend a spell preparation slot, unless she belongs to the Scholarly tradition – one of their main benefits is that they can ritually cast any spell in their spellbooks. Anyone who knows me will be shocked, shocked I tell you, to discover that I like the Scholarly wizards, and also really like their special abilities. Anyway, I think there will prove to be interesting design space in rituals, though I don’t have my own ideas on that developed enough yet to post about them. I’ll broadly state that I think there will be a concept of “ritualized preparation” that results in many more kinds of spells, especially combat-focused spells, having a ritual usage.

A number of spells, such as destruction and holy aura, have a Requirement. Since most spells don’t specify anything about Verbal, Somatic, and Material components anymore, the Requirement field carries any information on Material components, casting foci, or casting times for a spell that are anything other than a single action. In a few rare cases (feather fall is one) the spell requires an unusual condition. Between the quest-based components or foci and the options for unusual conditions, this field introduces a huge amount of variability in design, opening the door to something as textured and flavorful as the system of Catches in Changeling: the Lost.

I don’t expect anything with that kind of if/then complexity to be part of Basic or even Standard D&D, but I know I’ll be writing a lot of it for my own campaign’s use, and I expect that somewhere around three years into D&D Next’s post-release production cycle, they’ll release something called the Tome of Magic that explores such wild new spell mechanics. This will, generally speaking, be what they mean by “Advanced” D&D Next. (Ironically, this form of Advanced D&D won’t appeal to the grognards.)

There’s one more unusual hook currently found in spell mechanics: the truename. Only six spells in the current playtest packet have a truename mechanic, starting with Planar Ally at sixth level. In all cases, the spell does something without the target’s truename, but has a much improved effect with it; clone is an outlier here in that having the target’s truename allows the target to be someone other than yourself. There are two problems with the truename mechanic.

  1. Story-side problems. What is it that your characters do to get a truename? What is a truename? Is it the name your parents gave you at birth, a mystical name you discovered at the age of majority, or some other thing? This makes a huge difference not in the spell’s mechanics, but in how a character learns a truename. Some characterization as to the appropriate hurdles here is important, especially because of Item 2. Primarily, though: why would you ever write down your truename and risk someone else learning it? (This question has valid answers, but D&D doesn’t currently explore them.)
  2. Mechanics-side problems. The playtest document specifies that a creature’s truename “evolves to something new [after]” anyone uses that name in a spell. So let’s say you find a demon lord’s truename in a musty old tome. Do you have any good reason to believe he doesn’t have his truename changed every couple of years, weeks, or hours? The mechanical effect here is that if one truename spell cast against you fails through some miracle of dice luck, the caster can only use the less-powerful non-truename version for the rest of that encounter, and presumably forever (depending on the answer to Item 1).

I think there are good and bad ways to handle truenames. As a concept, I like their inclusion – after all, we went out of our way to put them into Dust to Dust. I like the theme of secret vulnerabilities that reward knowing a lot about one’s opponents. For that matter, I like just about anything that rewards the players for engaging their curiosity and learning details about antagonists. Truenames are tough, though, because rules that directly interact with player paranoia are always tricky.

Up to this point, I’ve just talked about what’s there, without introducing new ideas of my own. I’ll be talking about this topic a lot in the future, I’m sure, so this is a foundation for those future posts. Nevertheless, I want to trot out the ideas that prodded me into writing this in the first place, what some of my blog-neighbors might call the Joesky tax.

Spells should be intricate in design, yet readily comprehensible in application. The best way I’ve found to handle this is to attach the complex parts through an initial if/then statement that the reader can quickly dismiss if it doesn’t apply to her situation. For example, imagine an alternate version of fireball, that we’ll call Noquinn’s Disperses the Chill. It goes along just like a regular fireball spell – same level, range, area, and so on – but deals 5d6 fire damage (1d6 less than a regular fireball). If the user sacrifices a shard of jade at least 3″ x 2″ x 1″ (MSRP: 20 gp) as an additional component, the spell deals 7d6 damage, and also deals full damage to incorporeal or ethereal creatures. This is the exact opposite of how I’ve seen a lot of spells written over the years, in which the spell sounds great right up until you get to the end and learn that the interesting ideas you were formulating for using this spell are specifically forbidden, or the spell requires a component that is damn-near impossible to acquire. To put that another way, “oh, neat!” is a better way to end a spell than “oh, never mind.”

In another set of ideas: I am interested in creating families of spells that interact with and build on one another in a definite way. For example, the Harbingers of Doom spell family includes a melee damage buff (usually cast on an ally), a DoT/debuff (cast on an enemy), a zone (cast on the battlefield, requires Concentration), and an at-will direct damage spell. Every time you cast one of these spells, you add a Creeping Doom counter to a five-foot square that includes the target. Counters last until you go one minute without casting a Creeping Doom spell within 50 feet of them. As a free action on your turn, you may impose a penalty to all saving throws equal to the number of Creeping Doom counters on one target within 50 feet.

Obviously there’s a lot of writing left to do to get this concept into a fully usable form (such as the spells themselves). I was also imagining similar spell families for healing (something like “when you cast a non-cantrip spell of the Chalice of Life family, heal 2 damage per counter to one creature within 50 feet”). This also has a lot in common with an idea I posted some time back, but probably involving fewer separate pools of counters at any given moment. Given that D&D Next spellcasters have comparatively few non-cantrip spells per day, my feeling is that spell families need to be pretty impressive to make up for the sacrifice of flexibility involved in spending 3-5 spell preparation slots on a particular list of spells. Introducing a “finisher” spell that cleans up multiple counters from the field as a cost to cast the spell might also be interesting.

Somehow I don’t think I’ll get this written before the end of January, though. Kainenchen and I have big plans for the next week and a half. 😉

Leave a comment

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