Tuesday, 11 June 2013

Monitors: making magic 5

In this seemingly-endless set of posts about magic (currently including how other people do it, what magic means in game terms, approaches to what magic does, and ways of allocating magical power), there's still at least one thing I'd like to consider, which is different ways of casting spells.

A lot of this will involve theorising about the flavour particular approaches will produce, which I haven't actually tested. So take it with a pinch of salt; I cheerfully acknowledge it's just musing.

Bespoke and off-the-peg magic

In terms of actual spells, there are two main approaches.


Off-the-peg magic is the one I'm most familiar with, as used in Dungeons and Dragons, Call of Cthulhu, Deathwatch, Dying Earth and many many others. Here, all spells are predefined by the designers, and casters simply select (or generate randomly, or inherit through background) from the lists. They may be able to choose spells at the time of casting, or may have to pre-select them to various levels of strictness (once, daily, etc.).

Advantages of this system:

  • The spells that exist are known during design, and balance or gameplay issues can be worked out.
  • Specific spells can be explained clearly and coherently. Their capabilities, limits and situational uses can be laid out in one place.
  • Limits can be placed on particular uses of specific spells that might create gameplay issues.
  • Players can see what effects are possible, and choose from a list that has been balanced and themed. While applications of spells may vary, analysis paralysis should be limited.
  • Individual spells can be themed for the game, helping to produce the desired tone and genre.
  • Spells can be made highly individual and esoteric, giving a sense of very arcane magic.
  • Spells can be heavily tailored to particular classes, backgrounds, monsters or individuals, reinforcing their nature.

Disadvantages of this system:

  • Individual spells often take up large chunks of rulebook.
  • Every spell needs careful balancing and testing, to avoid must-haves and nevers.
  • Casters need to pick specific spells and plan characters around their options. This is particularly true of (D&D) wizard-type systems that restrict known spells, vs. cleric-type systems where casters know all spells they're eligible for.
  • Games can be clogged with variations on a theme.
  • Games can be clogged with level-varied versions of the same spell - sometimes even though each one also scales with level. D&D features this with whatever summoning XXI-type spells.
  • You may end up with several spells that do specific but similar things, rather than one more general spell. This might be to block an exploitative option that exists between the intended ones.
  • In real play, there will always be unexpected synergies, situational uses and 'exploits' not spotted by designers, so the system cannot totally prevent this.
  • A small group of designers will never be as inventive as the sum of all players, nor predict the questions and situations that will crop up. By limiting spell use to a set of preconceived options, they may restrict in-game creativity (though there's always creative use of existing spells).
  • Combinations of spells need analysis to make sure broken combinations don't appear by accident.

Bespoke magic

In contrast, this system is based on a small number of flexible parts, which can be controlled or combined to produce specific effects. Ars Magica combines verbs and nouns of various disciplines to allow spell creation on the fly, for example. In a more school-based version, a caster may have the ability to Summon Elementals, or Telekinesis, or Electricity, and perform various feats within that discipline depending on personal ability and intention.

Advantages of this system:

  • Allows wide scope for creativity.
  • No long lists of specific spells clogging up the rulebook.
  • Variation around a theme and level-scaling can be handled automatically, without the need for repetition of similar spells.
  • Gives a feel of creative, imaginative magic and mastery of magical forces.
  • Casters have a great deal of flexibility, with no need to plan spells beforehand. This avoids the problem of little-used niche spells in off-the-peg systems.
  • System has potential to be mechanically very simple.
  • Magic is intuitive, fluid and subject to the caster's imagination, giving a sense of understood or perhaps innate power.
  • Disciplines and their combinations can be themed for the game, helping to produce the desired tone and genre.

Disadvantages of this system:

  • New spells can always be created, so testing and predicting possibilities is difficult.
  • It may be difficult to coherently explain the limitations and possibilities of magic when it has no set spells. This can venture into confusing territory with hypotheticals, exceptions and conditions all over the place. Information can't easily be bundled into relevant packages, but is likely to be spread out.
  • For rulebooks etc., spells must be invented whenever examples are needed.
  • It will be tempting to fall back on the same old spells (or the 'best builds' dreamed up by players), reducing the advantage of creativity.
  • The capabilities of the characters are hard to predict, since much relies on the players' creativity. They may not think of possibilities envisioned by the scenario writer or GM.
  • Some uses of a discipline may end up under-powered, and others over-powered, because they are not individually tweaked.
  • The range of options available may cause choice paralysis, particularly if powers can be combined freely.
  • With spell design in the hands of players, 'spells' as used may not fit the intended feel.
  • Expanding the range of disciplines calls for reanalysis of everything, in case broken synergies are introduced.

Mana, mana, tekel upharsim

The next question is how spells and caster are powered. There are a number of common systems for fuelling spells. These relate to background, the genre and assumptions of the game, and balance.

Mana pool

In this system, characters have a pool of resource points specifically for casting spells (though it might be shared with other supernatural abilities, such as ki powers). This includes Call of Cthulhu Magic Points.

This system gives characters flexibility in spellcasting, and makes magical effort like physical effort: you can do a few major things or many small things before being exhausted. It places a cap on the amount of spells the character can cast in a given interval, useful if you want limited magic. Mana can also allow flexibility in the power of individual spells, if casters can spend more mana to enhance spells.

Mana systems can be combined with a casting roll; you might pay a fixed cost and roll, or bet a number of mana points that determines your likelihood of success. This is good if you want uncertainty, and the feeling that spells are costly and unreliable. The second option introduces tension and choice, and allows for both desperate long-shots and absolute determination to succeed. A disadvantage here is that a mana cost, plus a casting roll, plus any rolls for effect, may feel like just too many hoops to jump through.

Mana systems may also be used to avoid explicit level-limitations on spells. If the mana cost is too high, it's impossible for a low-level caster to use Spell of Utter Destruction. Conversely, in a betting system, even an apprentice wizard can have a shot at Spell of Utter Destruction, but their odds increase with experience.

Vancian magic

In this system, individual spells are prepared in advance and expended when used. It's generally a case of memorising and forgetting spells, but it could also be skinned as actually casting a spell beforehand, and then using a brief incantation to trigger the final stage - that's how I tend to rationalise it myself.

This system emphasises the importance of individual spells, and makes predicting upcoming challenges a key part of playing a caster. Both obtaining suitable spells to memorise, and choosing appropriately, are crucial. It automatically limits what casters can accomplish in any day, and does so more strictly than any other method - it's easier to grant extra mana to a caster than explain why they suddenly remembered a spell in defiance of normal magical law, and why that spell in particular.

An obvious limitation is, this is mostly suited to off-the-peg magic. I suppose in theory you could devise and prepare spells in advance, but the 'study and memorise' aspect of Vancian magic really clashes with the freeform creativity of bespoke magic. It also tends to sideline niche spells, since it's rarely worth preparing one on the remote chance it's incredibly useful, when you could prepare a common spell that'll almost certainly be quite useful.

As a pseudo-Vancian model, you could have casters choose disciplines in advance; this works for me to some extent with clerics, who might pray for particular kinds of blessings (protection, strength of purpose, strength of body) and then determine their manifestations as necessary. In a setting with varying sources of magic to tap, such as elements, a caster might have to actively replenish their power by tapping one or more, and their choices would determine what kinds of disciplines they could use that day. Of course, this partially offsets the benefits of flexible casting, but it might make for interesting play.

Spell slots

In this system, casters have 'slots' of varying level that can be used to cast spells, but can choose which spell to cast as and when. This is the sorcerer model from D&D.

The system grants greater flexibility than Vancian casting, since spells don't have to be prepared beforehand. This makes being a caster less of an odds-juggling role, and removes the possibility that a suboptimal spell choice will cripple you until the next rest. It makes it more likely that niche spells will be cast at all. At the same time, it keeps a mixture of weak and strong spells going, and provides a way to neatly scale caster power with level - just provide new spell slots.

If you want to encourage (well, force) casters to cast a lot of minor spells, a few substantial ones and just the occasional game-changer, this is the model. It prevents them from burning up their arcane power on a couple of ultra-spells and then having nothing to do. On the downside, it removes the option to do exactly that. Back on the upside, given players' prediliction for demanding to rest as soon as the wizard's out of spells, that may not be such a bad thing. But you could introduce other limits (some form of exhaustion, say) to prevent over-spend in other systems.

Some systems let you spend a slot of a different level to produce more or less powerful versions of a spell.

This model can work with either bespoke or OTP magic, as long as there's a way to determine the slot a spell should fit.

One significant disadvantage is that while the model's simple, it's also very artificial. Spell slots don't correspond to anything intuitive (though it's possible to construct settings where they work), and you can end up with the situation where a wizard is only able to cast very powerful spells, having exhausted low-level slots already.

Skill roll

In this system, casting a spell is basically like any other action: you roll against a target number. There is no other limit on casting, but there may be side-effects of both failure and success. This is the system used in Warhammer 40,000 RPGs.

A skill roll system may have a strict threshold, but often the roll determines the effectiveness of the spell, with very good rolls producing a particularly effective casting. Skill rolls dependent on level, or on stats that vary with level, are good for soft scaling, since higher-level characters will find it easier to cast spells and will do so more effectively.

The inherent unpredictability of skill-roll systems gives an unreliable feel to magic, and this is often supplemented in practice with mishaps and brilliant successes, which allow for occasional flashes of brilliance and for horrible side-effects. It's ideal for settings where magic is inherently dangerous or chaotic (cf. Warhammer 40,000). However, that doesn't have to be the case; it can also represent magic as simply difficult to do. Depending on the distribution of casting rolls, all spells might be very difficult to use, or some may be effortless and others challenging. It's a very flexible system; on the downside, the maths needs careful consideration to produce the desired effect. You need to decide what target numbers are needed, how they'll be affected by levelling or stat variance, and whether the ease of casting across the whole range of spells still matches your intention at various different levels.

Another decision to bear in mind is how many rolls are called for. If there's a casting roll for spells, then designers need to decide whether you also need to roll to hit targets, to overcome wills and so on.

Despite the unpredictability, this system gives a lot of power to casters, since the issue is not usually whether they are capable of casting a spell, but whether their odds of success are good enough. Given enough time, they can generally succeed - assuming that any consequences of failure aren't too catastrophic. This means mishap charts and the like are a very strong influence on the feel of the game: without them magic is merely challenging, but a mishap chart's tone can flavour magic anywhere from 'hilarious' to 'frustrating' to 'psychotically dangerous'. There's also the question of whether mishaps affect only the caster (which makes it their call), or spill over to influence other allies and enemies (which makes them a collective issue, and can make casters a liability).

Unlimited casting

Does exactly what it says on the tin. I can't immediately think of anything that allows this fully, but some recent D&D-type games allow unlimited use of certain minor spells.

Logically speaking, there's no particular reason why spellcasting has to be limited when (say) punching people in the head or swimming across rivers isn't. You could use the same system to handle both, whether that's a cheerful "no limitations", or some kind of exhaustion system. If necessary, powerful spells could be limited in the same way that other powerful abilities often are, through resource requirements: missile launchers need missiles, fireballs need bat guano. This would introduce a dichotomy between self-powered spells (the equivalent of physical and mental feats) and magical techniques that require special equipment or supplies (the equivalent of technical feats).

Of course, unlimited casting isn't necessarily cost free; you could perfectly well combine it with a random effects table, so that the decision becomes a cost-benefit analysis. Is it worth risking conjuring a flock of noisy crows as I magic a hole in the castle wall? Or putting everyone to sleep by casting a lightning bolt?

Fundamentally this is going to give a very accessible feel to magic, whether that's the controlled feel of mishapless magic, a chaotic wellspring ridden with unpredictability, or a dangerous but tempting force that's all too easy to tap into.


This system draws on the caster's physical resources to power spells. Effectively, casting spells does damage to the caster.

I'm sure this is used somewhere (I can't imagine I'm the first person to invent it), but I don't have any actual examples.

This seems like it would work best in HP-type injury models, where depletion can be predicted and managed. A system with injury charts would make spellcasting very dramatic, but I feel like it would tend to be too dangerous to actually use magic (or at least, feel that way) if there's a good chance of blinding yourself or rupturing your spleen. Of course, in a system designed for painful choices and self-sacrifice, that might be fine, and the extent of injuries could be limited.

Rather than depleting an HP pool, the other option is depletion of stats. Depletion could affect a 'casting stat' (as the brain's magic centres get tired), or drain physical strength. It could even stack up, applying general penalties to various actions, as a form of exhaustion.

This approach seems like it would make magic seem exhausting and personally costly to use, but relatively predictable. Even a random injury roll is more controlled than magical mishap charts - you might break a leg, but you're not going to create a rain of frogs or reverse gravity.


So you've got a way to define spells, and a way to power them. But how reliable are they?

Randomness is going to be a significant influence in terms of tone. A non-random magic system will make magic predictable, reliable and relatively straightforward - it's a tool that can be called on, providing you can meet the requirements and pay the price. In contrast, a highly random system makes magic unreliable, unpredictable and potentially risky (randomness may or may not include outright danger).

I've covered a lot of this above, but the main types of randomness are casting reliability (will I get this spell off?) and side effects (what else might happen?).

Casting reliability

Casting reliability partly determines the position of casters. Are they masters of arcane energy, or just dabblers? Can they summon up the power when it's needed, or is it always a gamble? The less reliable casting is, the more amateurish casters are likely to seem. It's hard to maintain the 'wise, powerful wizard' schtick if you're failing multiple spell attempts in quick succession. Unreliable casting is good for things like Cthulhu where magic is a mysterious force and PCs are bumbling amateurs dabbling in What Man Was Not Meant To Wot, but bad for Dungeons and Dragons.

Side effects

Side effects, in contrast, are all about the flavour of magic. The choice of whether they're automatic, rare, or a consequence of failure will make a lot of difference: the first makes magic seem chaotic, the second gives a complex and mysterious vibe, and the third makes it technically tricky and risky.

The type of side effects is likely to be a big factor in establishing genre, though I reckon what you don't include is maybe more significant than what you do. Warhammer 40,000 features casters floating into the air and ignominiously falling back down, and accidentally summoning demons, both of which could equally well feature in a slapstick game. However, it doesn't have rains of custard, people transforming into animals, silly voices, and other silly elements; the overall balance of effects maintains the sinister and untrustworthy nature of magic.

Of course, Warhammer 40,000 mixes things up by letting casters decide how forcefully to unleash their powers, and adjusting the chance of side effects accordingly. This helps the tone quite nicely: you can focus your efforts on keeping the spell under strict control at the cost of reduced effort, or you can summon up all the power at your command, exposing you to the dangers of the Warp.

Arguably, 'critical casting' (in some incarnations) can be a form of side effect, with the spell far more effective than expected (potentially creating its own problems). The disparity emphasises lack of control over the power you unleash.

The other facet of side effects is how serious they are. Minor side effects will tend to give an esoteric feel, without substantially affecting gameplay. Major side effects can seriously disrupt plans, threaten characters and generally chance the balance of the game. In the middle you have effects that cause inconvenience (or unexpectedly help out), being about as effective as a player action or two, but aren't likely to ruin things. The more serious the side effects, the more of a chancy proposition magic becomes: the stakes are higher, and it's often going to be wiser not to take the risk. So if you want magic to be heavily used, and you're not running a game where with hilarious consequences is a common theme, I suspect side effects should generally be smaller.

Okay, we're reaching the point where I can't tell if I'm making sense any more, so let's call it there.


  1. This system draws on the caster's physical resources to power spells. Effectively, casting spells does damage to the caster.

    I'm sure this is used somewhere (I can't imagine I'm the first person to invent it), but I don't have any actual examples.

    My first ever RPG - Advanced Fighting Fantasy - used exactly this system. Casting spells drained a character's Stamina.

    There was also a spell that restored Stamina. This was not a particularly well thought-through system...

    1. Cheers! Yes, I think I've seen that kind of combination before - I think it's always really tempting (and sometimes crucial) to supply stat-restoring techniques, but working out the consequences is fiddlier than you'd think. Or else it's not considered a serious problem.