Making Magic More Interesting

by Jim Vassilakos

Making Magic More Interesting (and Dangerous)A Brief Mana System with Spell Failure for AD&D

After several years of playing AD&D back in the 80s, I grew somewhat dissatisfied with the game’s magic system. Looking back, I think what was lacking for me was that ineffable sense of wonder that initially attracted me to roleplaying games in general. By that point, of course, I had become reasonably familiar with what most of the spells and magic items did, and so magic in the game unavoidably acquired a certain degree of mundanity, the very antithesis of the sort of feeling that magic is supposed to arouse. Of course, I’d heard about other GMs modifying the rules to shake things up and had even seen some examples of this, so that’s exactly what I began doing with the magic system.

All I knew, initially, is that I wanted to recapture that sense of wonder, but I didn’t exactly know what that would entail. Thinking about it now, however, I realize that I did have various specific goals:

1) Make magic more chaotic and therefore less predictable.
2) Make it more expensive to be a spellcaster (giving the PCs something to do with their treasure).
3) Make magic potentially even more powerful than usual (but only at the cost of even greater risk and treasure).
4) Incentivize and reward specialization.
5) Modify certain spell descriptions to make them less confusing and/or less abusable by imaginative spellcasters.1

Rather than going through the whole magic system in great detail, I’m just going to present a bare-bones version that will hopefully give you some idea of how I achieved some of these goals. Hopefully, this will give you a feel for the system as a whole, and with any luck it’ll give some ideas that you can port into your own game, should you feel so inclined. However, if you want to check out the whole thing, it’s available in the campaign archive at

Having said that, I should warn you that magic systems are a can of worms, and once you start fiddling with one aspect of the system, it may soon draw you into tinkering with other aspects. Ultimately, there’s really no end to it, so decide for yourself how much work you want to put in or you may find yourself drifting down the slippery slope of designing your own RPG.3

The Thirteen Colleges of Magic

I started off with this notion that magic would be separated into various colleges. Hence, in addition to a general education in Common Magic, each initiate would select a number of colleges in which to apprentice. The colleges were as follows:

1. Conjuration: The conjurist makes things through magic.
2. Divination: The diviner, oracle, or astrologer foretells to future, sees into the past, and generally acquires knowledge.
3. Divine Magic: The cleric or priest manipulates the essence through a god or deity.
4. Elementalism: The elementalist calls upon and manipulates substances from the elemental planes. These are your stereotypical combat mages.
5. Enchantment: The enchanter charms and manipulates.
6. High Magic: The wordist or runemaster uses magical runes and power words.
7. Illusion: The illusionist confuses the senses.
8. Low Magic: The spiritualist enters the spirit world and calls upon spirits from that world.
9. Mysticism: The mystic or psi manipulates the essence though the inner power of the mind.
10. Naturalism: The naturalist or druid manipulates living things and the forces of nature.
11. Necromancy: The necromancer or summoner calls upon devils, demons, undead, and the like and binds them into servitude.
12. Thaumaturgy: The thaumaturge or alchemist creates potions and magical substances.
13. Transmutation: The transmuter or changer changes attributes of things.

So, basically, what I did here was to divide the spells that I wanted to allow into the campaign between these various colleges (sometimes I’d allow a spell into more than one college if it seemed to apply to both). Likewise, studying within each college came with its own set of special abilities. Now, I realize that this is a lot of work that you may not want to do, but the reason I did this in my campaign had to do with goal #4, incentivizing and rewarding specialization. Read the section below, and hopefully it’ll become clear.

Spell Level Restrictions and Spell Failure

Intelligence Score
Max Spell Level
Wisdom Score
Base Failure Probability

Note that intelligence controls the maximum spell level a mage can theoretically cast, whereas wisdom controls the base probability of spell failure. In short, intelligence gets you into trouble, and wisdom gets you out. In my game, I tend to use wisdom as a proxy for willpower, so in this sense, wisdom is what a spellcaster uses to control the spell he or she is casting.

The base failure probability is increased by 5% per college within which the caster studies in excess of one (hence, one way to keep spell failure down is to study within only one college, and, indeed, there is social pressure from peers as well as superiors to do exactly this). I should note at this point that in my campaigns, spellcasters cannot multi-class as fighter-mages or thief-mages, as they are allowed to do in most other campaigns. If you still want to allow this sort of multi-classing, I’d suggest building in a similar rule for this, such as +10% per non-spellcasting class. Hence, a fighter-mage-thief would be casting at +20% base spell failure. Or you can lower this to +5% per non-spellcasting class, if you think +10% is too extreme. It all depends on whether or not goal #4 is one that you share, and if so, to what extent.

The base spell failure is then reduced by sl/cl (keeping the fraction), where sl is the spell level and cl is the caster’s level. Hence, a 6th level caster with a 16 wisdom who is studying within only one college would cast a 3rd level spell at 6.5% spell failure (13*(3/6) = 6.5). This is termed the natural probability of failure. If they’re studying in two colleges, however, then the natural probability of spell failure rises to 9% ((13+5)*(3/6) = 9). This percentage can be further modified by various environmental factors, but for now, it is enough to give you a basic idea for how the natural probability of spell failure is calculated. This section serves to satisfy goals #1 and #4, but like I said, if goal #4 isn’t important to you, you can just dispense with the parts that incentivize specialization.

I think the main reason I decided that I wanted to incentivize specialization was that I realized that introducing a mana system (see below) would basically make the mages more powerful. They’d be able to cast the same spell over and over, and if they had the right mana, they could get superior effects. Hence, in order to bring them back down in power, at least a little bit, I figured that by forcing them to specialize, I’d be reducing their overall attractiveness to potential players. But like I said, your mileage may vary.


Spellcaster’s expend mana in their use of spells at a rate of one point per level of spell cast (a half-point for cantrips). These mana points may be absorbed at a rate depending on the
source of mana being utilized. There is no upper limit on the amount of mana a caster may have in his or her system. However, materials for meditation are typically expensive, the prime cost being the crystal or incense which casters must burn in order to acquire mana. The most common varieties of this crystal are described below:

Poor quality: poor effects, a little iffy
Standard quality: normal effects, non-addictive
High quality: good effects, somewhat dangerous
Highest quality: great effects, very dangerous

Here’s a more explicit chart showing how different types of mana affect spell failure as well as the caster’s effective level for purposes of determining a spell’s effects:


Caster’s Effective Level
Spell Failure Modification
Absorption Rate
1 per 4 min

1 per 3 min
1 per 2 min
1 per 1 min

All the forms of incense are burned in an open flame which first melt the tiny sphericals and then cause their slow transformation into an ethereal essence. Other casters within the vicinity of the burn may absorb the mana (even unwillingly), thus stealing the effect of the crystal from its owner. Mana is expended on a FIFO (first-in first-out) methodology. If two forms of mana are used within the same spell, the practice is known as “mixing”, and bizarre effects may occur. Incense is typically stored in small vials, each holding as many as a hundred grains of the substance.

Mana types, other than Hail, affect the caster’s probability of spell failure. Ice increases it by 50%, Tempest by 100%, and Maelstrom by 200%. This is termed the modified failure probability. Hence, our 6th level caster, with a natural failure of 6.5%, would, when using Tempest, have a modified failure probability of 13%.

This section of the rules is basically meant to satisfy goals #2 and #3. #2 was especially important to me, as the PCs in my games usually had more treasure than they knew what to do with. For a few years I tried being the stingy GM, basically giving out silver when the rules said gold and gold when the rules said platinum, but the players were soon on to what I was doing. Telling them that I thought the game handed out too much dough was hardly satisfying (at least from their point of view). They, of course, wanted to buy magic items and castles, and these, of course, are things that PCs typically do once they amass enough wealth, but I wanted some means to bleed them dry long before they ever started looking around for a magic shop. This system didn’t achieve that entirely, but it went a long way towards that goal.

Mana Tolerance/Addiction

The caster must make a tolerance check using the following table every time spell failure occurs.

Ice 1%
Tempest 3%
Maelstrom 5%

If the check is successful, the caster is said to be tolerant to the mana in question. For every point taken in by the caster (including the points currently in the caster’s system), only half a point will be retained. What this means, essentially, is that the caster must now obtain twice as much mana as before. The caster may break tolerance by moving to a new form of mana, staying with that new mana for d20 spell castings before the tolerance breaks. However, in order to muster the will to do so, the caster must make a wisdom check (which may be made weekly until successful). Failing the check results in the caster being unwilling to change mana, and in this instance, the caster is said to be addicted.

Mana tolerance is usually accompanied by some physical manifestation, such as a change in eye color, a doubling in the size of the ears, a sudden loss of hair, or the like. Casters who refuse to break tolerance (and hence are addicted) eventually become ill or crazed or both if there is too little mana in their system at any given point in time. Likewise, strange magical effects are said to follow them, and it has not been unknown for them to entirely disappear for some time, wandering as ethereal beings or even drifting into astral space.

I threw this rule in basically because it seemed apropos. If the spellcasters were taking drugs for their mana, there should be some consequences. To be honest with you, however, I ended up not enforcing this very much, because most of the time it just seemed to get in the way of the game. You may just want to toss it out entirely, or if you do use it, you may want to come up with some magical ways of breaking addiction, so it doesn’t end up slowing down the game.

Casting from Scrolls

For casters casting from scrolls, normal mana is not expended. The power for the spell comes from the scroll itself, so the effects are assumed to be the same as if the caster were using Hail as his mana source. As with normal spellcasting, the sl/cl rule is used to determine natural spell failure. However, the caster may attempt to cast spells from scrolls that he wouldn’t normally be able to cast (for example, the spell’s level may be too high). Hence, the sl/cl rule applies. This modified failure value then doubles if the caster is reading a scroll from outside his or her college (if you’re not breaking the magic system up into various colleges, then you can ignore this rule).

Other Situational Factors & Spell Failure Results

Regardless of whether the spell is being cast from memory or from a scroll, the modified failure value is doubled if the caster has been using metallic weapons and/or metallic/leather armor during the past few (d6) days. For this reason, most casters use only bone daggers and wear only cloth garments. Even their sandals are made from woven straw, a sure sign someone is a mage, for even peasants will wear leather boots or sandals if they can afford them.4

Other situational elements also play a role in determining the final spell failure probability. If the caster is below half hitpoints, the modified failure increases by 50% of its value.5 Likewise, if the caster is attempting to do something (like running) while casting, modified failure goes up by another 50% of its value (running while heavily wounded would essentially double the spell failure probability). So, just as an example, our hypothetical 6th level caster’s 13% would rise to 19.5%, if he were casting while heavily wounded. If he were running at the same time, it would go from 13% to 26%.

Interruption of casting (usually by being hit) will cause automatic spell failure. Whether or not this occurs is determined by the GM, but in general, the segments×10% rule may be used. To illustrate, all actions are declared at the beginning of the combat round. Fighters do combat, mages do spellcraft. If the mage is hit during the round, the number of segments required to cast the spell is multiplied by 10. That is the percentage chance that the casting was interrupted by the hit, resulting in automatic failure.6

When a spell fails, the mana is expended as if the spell were cast, however the effects may vary from unnoticeable to catastrophic. There is an x/cl% probability (where x=# of mana points reserved prior to the spell being cast, and cl is the level of the caster) during spell failure that the caster will “barf up his cookies” resulting in the accidental dumping of all available mana into the failed spell.7 This may boost even a low-level spell to extreme power, and since failure is already occurring, nearly anything can happen.

Spell Failure Table (d10)

Reduced Effects (×½)
Random Target (if another target is in range, otherwise fizzle)
Twisted Effects (spell works but with some minor modification)
Backlash on Caster
Mana Detonation (1d6 dmg per mana in spell to all with 30’, save for ½)
Doubled Effect
Something Unusual (take a luck roll)8
Multiplied Effect (×d4+1 = ×2 to ×5)
Roll Twice, ignoring this result

(Die modifiers based on mana: Ice: -1, Tempest: +1, Maelstrom: +2)

Needless to say, this section of the rules was written to satisfy goal #1, and for me it’s really the section that does the most to recapture that sense of wonder I initially mentioned at the beginning of this article. Most other spell failure systems I’ve seen don’t go into much detail about what it means if a spell fails. With this one, I didn’t just want spells to fizzle. I wanted the possibility of them exploding with unforeseen consequences.

Note that not all spell failures need result in disaster for the mage or his party. Indeed, both multiplied effects and unusual occurrences can be of value in a tight situation. However, they can also spell doom for the caster, particularly if conjoined with a backlash. Casters, under this system, are known best for dying by their own hand in the heat of battle, particularly the low level ones who haven’t had the time and experience to learn how to control their magic. Likewise, many a party member has been unintentionally toasted by his own mage. For this reason, casters are feared and despised nearly as much as they are sought, and many henchmen, having heard horror stories in various taverns, will refuse to join any party containing an inexperienced caster. Fortunately, there are certain magics which are docile in nature, such as simple cures and detection spells, and these will only fizzle or act upon a random target, perhaps yielding the wrong information in the case of detections. In all cases, consult the GM before assuming that a given spell is docile.

In addition to spell failure, I also like to make various high level spells inherently unstable. In other words, things may go wrong but not necessarily because the mage miscast the spell. Sometimes you do everything right and things still go wrong. That’s just life. Hence, I’m going to present two special cases, just to show you how this works.

Special Case #1: Resurrection

Like many GMs, I’ve seen how resurrection-like magic can be abused to create rather bizarre situations. This is particularly true when you monkey with the rules as much as I do, and, more to the point, when you allow your players to do the same. One AD&D campaign witnessed a particular PC visiting a local slave city in order to purchase (or in some cases capture) live bodies into which she could place the souls of her recently departed companions. Giving a PC that sort of power (never mind that she’s a 20th level Necromancer) places her in the position of being able to recruit some very loyal henchmen.

“If you have any questions about how dearly I value my employees, why don’t you go talk to Lyngar. I brought him back to life just last week…and that was after visiting the demon realms to find his soul.”

Yes, I made things more difficult and dangerous than the simple casting of some heinous spells, but all fun aside, what would be the social ramifications of this sort of magic?

The most obvious one, I think, is effective immortality for those who can afford it. And this is a big deal. It means that rather than succumb to old age, King Phoobar can have the college of royal wizards transfer his soul into the body of that dashing young bandit, Robin whatshisname, who’s about to be sentenced to death. Sure, you can make an adventure of the merry men coming to save Robin, but the end result is that if you want to kill the King, you’d better kidnap him, take him somewhere remote, incinerate him with something hot and deadly, and then scatter his ashes to the four winds. It takes all the fun out of poison, if you ask me.

So what are we to do about this sordid state of affairs? Well, before jumping into the problem, let’s look at what some other people have written on this topic. Here are three articles on the topic of resurrection in AD&D:

Ashes to Ashes (Graeme Drysdale), White Dwarf #78 (1986): Drysdale divides the traditional races according to whether they have a soul (dwarves, half-elves, halflings, humans, and gnomes) or a spirit (elves, half-orcs, and monsters).9 He says that only creatures with souls can be resurrected or raised (as per the “raise dead” spell). In his view, those with spirits would only be able to be reincarnated.10 Drysdale also suggests changing the resurrection spell so that it only recalls the character’s soul but doesn’t restore their body. He provides rules for spell failure, which may result in the wrong soul being recalled. Even if the right soul is recalled, he says that the shock of inhabiting a new body may result in insanity.
Dead or Alive? (Michael Satran), Dragon #210 (1994): Satran discusses some of the possible social and legal ramifications of resurrection and presents a number of scenario ideas. As for limiting its use, he suggests only allowing resurrections to take place on certain holy days of the year as well as erecting various legal and political barriers, and he brings up the notion of having resurrected characters suffer from post-resurrection stress disorder as well as other personality changes.
New Life for Resurrections (Rod Meek), Dragon #263 (1999): Meek suggests a number a variant outcomes for resurrection magic, such as (1) the recipient being required to make a regular blood sacrifice to the resurrecting deity, (2) the recipient’s resurrection being only temporary, (3) the recipient being required to find someone willing to give up their life in exchange for his or hers, (4) the recipient only being resurrected for a certain period of every day, the remainder of which he or she exists either as a corpse or an animated undead, (5) the recipient’s soul having been replaced by that of an infernal spirit (and, hence, the resurrection itself being more or less phony), (6) the recipient being bound to remain within the general proximity of the resurrecting priest or the temple where the resurrection took place, (7) the recipient’s soul being unwilling or unable to return to his or her body, (8) the recipient requiring a new body because the old one is so badly mangled, (9) the recipient’s soul being already bound to another person’s body, and (10) the recipient coming back as an undead creature or as a living one who carries the secret scent of death, thus attracting the ire of the jealous undead due to their ability to sense that the recipient has been given a second chance at life, a gift which they themselves were denied.

It is interesting that Drysdale and Meek both came up with the idea of a resurrected body being inhabited by the wrong soul. When I initially began thinking about what could go wrong, this was also my first thought. However, before putting forward some generic rules on limiting the reliability of raise dead, reincarnation, resurrection, and similar soul-manipulatory magics, I need to first illustrate the concept of the luck roll, because that’s what we’re going to be using.

In my game, a great many results are decided by the roll of a d6 (or “luck die”). Whenever the players have a question about some aspect of the scenario which I did not previously devise, or say on a spell effect as to which I’d rather not make a snap-ruling, I’ll say, “Take a luck roll,” keeping in my mind some number which they have to meet or beat in order to arrive to a favorable circumstance. On the pro-side of things, it helps circumvent a lot of potential arguments and keeps the players from feeling like I’m nursing a grudge. This, in turn, keeps the game moving quickly. On the con-side, the luck roll can be overused and should often be rolled in secret by the gamemaster for determining the answers to questions before they get raised.

In any case, here are the new rules I promised:

New Rule #1: Law of Skulls. Generally speaking, if you want to cast a resurrection magic, you need the skull of your dearly departed. Otherwise, your odds of summoning the right spirit are next to zero. The only time this can be avoided, again speaking generally, is when you already have the soul inside a soul-gem (in which case, forget about question #1 below).

New Rule #2: Law of Chaos (a.k.a. Murphy’s Law). Even if the spell is cast correctly and is successful (i.e. spell failure has been avoided), there is still a degree of chance involved due to the inherent complexity of calling a soul from the realms of the dead and joining it to a living host. A number of misadventures could potentially occur purely by chance. Consult the following questions:

Question #1: Do we have the right soul? Roll d6:
1: Nope, you’ve got yourself a demon. He’s a cunning SOB and will likely fake amnesia unless he’s already done a thorough interrogation of the dead soul in question.
2: Nope, you’ve got yourself the wrong guy. He’s no demon, but he’s not your friend either. Last thing he remembers, he was falling down some pit yelling, “My Precious!”
3-5: No problem. You’ve got your man. And boy is he happy to be home…except for one thing. He wants his experience points and his share of the loot!
6: Not only do you have your guy, but he’s brought a friend! There’s now a second consciousness inside his brain which is able to feed him all sorts of information as well as the ability to use skills he didn’t even know he had. Possible downside: the other consciousness can take control of the body while the initial one is sleeping. Can you say split-personality?

Question #2: What is the state of her memory? Roll d6:
1: Amnesia. You’ll need some powerful magic to restore the bits in her brain. Until then, she’s back at zero-level without a clue in the world.
2: Temporary amnesia. At first she doesn’t remember much, but over d20 days, everything will come back, albeit slowly: “Hey, wait-a-minute…if we’re really married, where’s my ring?!”
3-5: No problem. She remembers who you are. That’s why she’s crying: “Wah…not you guys again…I thought I’d finally gotten away from you idiots!”
6: Not only does she remember who she is, but she has some memories from the other side, memories of conversations with the dead or with powers beyond those of the mortal realms: “And then I saw Zeus, and he said, ‘Frag off, ya lousy sod!’”

Question #3: What is the state of his connection between body and soul? Roll d6:
1: Not so good, unless you like having a vampire in the party.
2: Everything seems to be fine until he starts seeing ghosts, the souls of past enemies who like to come along every now and then and beat the living tar out of him (see the movie Flatliners for an idea of how this could work).
3-5: No problem. He’s alive and well. A bit groggy, of course, but that’s to be expected, having been dearly departed and all that good stuff.
6: All appears well, except during the night when he sleeps. Then his soul has a tendency to go wandering out from his body, and when he wakes in the morning, he often finds that his dreams are more real than imagined: “Was yakkin’ with Zeus again last night. He says we’re screwed.”

These rules may not prevent kings & queens from being resurrected, but they’ll sure make it that much more difficult. In any event, if you plan on including soul-transference in your campaign (along with elixirs of youth and so-forth), then you’ve really got to plan to have some long-lived rulers, leaders who have been in power for centuries, at the very least. These leaders will likely be some tough mothers, at the very top of their respective classes.

As for politics, instead of court intrigue revolving around the passing of the crown, it will revolve around the passing of favor, with all the resultant brown-nosing and status-building one might expect. Most likely, kings will be watchful of their subordinates, and many will likely seek to dispose of those who become too powerful, either personally or politically. In part, that’s because their enemies can’t just wait for them to die. On the flipside, if their enemies have access to the same magics, then the showdown will come eventually. Best to force it before their enemies grow too powerful. The particularly far-sighted ruler may decide to take care of a potential threat before his enemy even becomes his enemy: “I’m genuinely sorry to have to kill you. You were a good friend…are a good friend. But you must understand… in time you would have become weary of serving another. It is natural… it has happened before…and it is the reason you now must die.”

Special Case #2: Teleportation

Nearly everyone who has played AD&D is aware of the teleportation spell and may have even used it on occasion. Like the transporter in Star Trek, teleportation magic is a great plot device, however, I’ve found it useful to limit it in various ways.

1. Gorgon’s Blood: Kings have been known to pay a high price for the blood of a gorgon, as when mixed into cement mortar, it conveys a permanent seal against teleportation. Those who have access to such blood will always use it when erecting major fortresses.
2. Lead: Dwarves often dig their lairs among veins of this common metal, as just like Gorgon’s Blood, lead also protects against unwanted magical intrusions. Unfortunately, Dragons also know this, and have been often known for enslaving dwarves to hollow out their lairs, lining their great caves with molten lead.
3. Spells: I’ve also postulated spells of magical sealing to temporarily prevent teleportation to or from a given location.

Even after adding these limitations, however, teleportation is still too easy for my taste, and were I to edit the rules some more, I would make teleportation a great deal more hazardous and hopefully more interesting as well. Of course, when devising such rules, it is worth considering what teleportation actually constitutes and what might go wrong as a result. For my own game, I would presuppose that teleportation involves some interface with the astral plane (or in the case of inner-planar teleportation, an interface with the ethereal). Hence, something as simple and seemingly unpredictable as an astral storm could completely ruin a spellcaster’s day. Using this as a basis, the following questions should be asked:

Question #1: Will the spellcaster appear where he wants to appear? Roll d6:
1: No. Astral winds blowing across the surface of the plane pick up the caster (along with anyone accompanying him) and deposit him elsewhere.
2: Yes, but he appears facing the wrong direction.
3-6: Yes. (I dispensed with the old altitude control problems. I didn’t think it fair, or very interesting, to have a caster’s feet materialize inside a block of stone.)

Question #1a: In the event of being sucked off-course, where might the caster end up? Roll d6:
1: On another plane. Determine randomly.
2: Thousands of miles from his chosen destination. Determine randomly.
3: Hundreds of miles away.
4: Tens of miles away.
5: 1-10 miles away.
6: Within a mile.

Question #1b: In the event of a party being teleported during a mishap, do they at least remain together with the caster? Roll d6:
1: No, they all go their separate ways. Determine each person separately.
2: They are split into two groups (randomly determine who goes with which group).
3-6: They remain together.

Question 2: Will the caster’s entrance (or re-entrance) into the plane be noticed by high-level mages and other local powers who make their home in this area? Roll d6:
1: You bet. Not only has someone sensed the vibrations of his entrance, but they also know his direction.
2: Yes, somebody sensed his entrance, but it’s difficult for them to figure out the direction where he materialized without the use of additional magics.
3-6: No. Unless someone was actively sensing for magical vibrations, they didn’t sense his entrance.

Question 3: Will there be an unusual mishap or other event? Roll d10:
1: Yes. See below.
2-10: No.

Question 3a: What is the Unusual Mishap/Event? Roll d10:
1: Caster and companions are all naked. They teleported in without their clothes or any other belongings.
2: Caster loses his mana in transit. If you’re using a system where he must memorize spells, assume that he forgot them all and must take time to study.
3: The teleportation bubble broke in transit, and the party finds itself on the astral plane.
4: Same as #3, but they face an immediate threat.
5: Everything goes fine, except that when they appear at their location, the caster and companions find themselves para-ethereal: like ghosts, in essence. This condition will last d10 minutes, at which point they will precipitate fully into the prime material plane. Ignore for the outer planes.
6: Same as #5, except that the condition is permanent until dispelled.
7: The passage of the teleportation bubble through the astral plane caught the attention of a major power, perhaps a demon lord, a devil, or even a God. GM’s discretion.
8: The passage of the caster creates a temporary rip in the fabric of the multiverse either at the point of origin, the point of destination, or both (determine via d6: 1-2: origin,
3-4: destination, 5-6: both). For 3d20 minutes, a hole shall exist which will either suck people into astral space (or deep into the ethereal plane if the spellcaster was trying to get to or from one of the inner planes), or, in the case where both sides are torn, will allow free passage from one side to the other, as were a magical gate created.
9: Same as #8, except the rip will last for d20 hours.
10: Same as #8, except that the rip will remain permanently until magically repaired. A dispel is not sufficient, however, a wish is. If the gate is open-ended to the astral, it will eventually be noticed by higher powers, and they may wish to learn who created it.

Conclusion & Society’s View of Magic

Obviously, these sorts of rules serve goal #1, making magic more chaotic. In so doing, they also tend to reduce a spellcaster’s overall effectiveness, which is greatly enhanced by the use of the mana system. Hence, in terms of game balance, spell failure and the law of chaos tend to counterbalance the inclusion of mana (magic becomes more potent but less wieldy), but the critical point for me is that they do so without not necessarily diminishing the individual spellcaster’s aura of power. One could even argue that they enhance this aura by making magic all the more magical. This style of magic-system, therefore, lends itself to that sense of wonder that I described at the beginning of this article.

Having said this, one must naturally consider the social ramifications of magic being so chaotic and unwieldy. As I view it, there would generally be two sorts of societies with respect to attitudes toward the spellcaster.

The first and perhaps most commonly held perspective, would be that spellcasters represent a threat to public law and order in much the same way that a thief's talents might represent the same, although the former case is all the more serious, as a mage has more sheer destructive potential than any thief. In these sorts of societies, spellcasters would need to find niches in which to hide their talents, and so the propagation of secret organizations would be the likely result.

The second sort of society would temper its fear of magic with a greed for its potential, and thus the position of being a known spellcaster would likely become smothered in bureaucracy, which may include restrictions to travel and magic-use within the community. To compound the problem, spellcasters would likely form subcultures around their various colleges, and some of these might even fall into conflict, such as the traditional animosity between priests and necromancers.

Hence, although there is great power to be won, there are also various potential social consequences to being a spellcaster that every GM should consider prior to starting a new campaign. Likewise, as long as lower-level spellcasters are dependent on higher-level ones for progressing in power and learning new spells, these spellcasting subcultures, whether legally sanctioned or otherwise, can become a real force in the campaign. Instead of simply starting a character out as a first level spellcaster, it might be appropriate to assign each spellcaster to a master, this being an NPC to whom they must report on a semi-regular basis. In this way, the GM can more easily convey to the players the social rules of being a spellcaster in her or her campaign.


1. In the early days of Dragon Magazine, there was a column called “Sage Advice” (see, where players would write in asking questions about the game, and occasionally these questions had to do with how to interpret specific spells. A particularly enterprising OSR-GM might want to go through these archives to incorporate various answers into the spell descriptions. Also, there are certain spells that are clearly overpowered for their spell level, and so players always choose those spells at the first opportunity they get. Rewriting the spell descriptions would naturally allow a GM to tweak the levels of various spells.

2. I should warn you in advance that the version in the campaign archive is, like the version I’m presenting here, incomplete. In short, if you want to use it, there are some holes you’ll likely need to fill in as your campaign progresses (in my case, I allowed the players themselves to propose various rules and spells for inclusion into the system). In any case, if you do end up using the system or some subset thereof, please drop me an email to let me know. Eventually I may get around to putting out an updated version, so whatever questions or comments you have may come in handy.

3. Not that there’s necessarily anything wrong with that, but it can get time-consuming.

4. I was trying to provide a rationale for why mages don’t wear armor.

5. Looking at this rule in retrospect, it may be a little too harsh. It may also present a bit of a book-keeping challenge, as mages in combat often get seriously wounded by virtue of the fact that they’re potentially the most dangerous weapon that the party has, and so any intelligent opponent will focus on taking them out first.

6. I did this as a way to speed up combat. In most games I’d played in, the GM counts segment by segment, figuring out who attacks when based on their initiative roll for that combat. When I GMed, I decided that making everyone roll for initiative every round unnecessarily slowed things down. Likewise, the segment count slowed things down, forcing everyone to roll their attack only when it was their turn to do so. In my game, initiative is rolled once per combat, and all the players roll their attacks simultaneously. But in order to achieve this, I had to make up some rule to handle the situation of a spellcaster getting hit at the least opportune moment (when they’re in the middle of casting a spell). Hence, this is the rule I came up with.

7. This rule used to be just x% rather than x/cl%, but I decided to change it after watching too many incidents where a high level caster inadvertently ruined his own day most spectacularly. In one incident that stands out in my mind, the party had, after a series of adventures, finally confronted the big bad guy, a high level mage. With his minions protecting him, he was in a good position to toast the party from a distance, but unfortunately for him, I rolled spell failure, and then I rolled that he dumped all his mana into the spell (because, being a high level mage, he had a lot of mana in his system), and then I rolled a backlash on the spell failure table, so the battle was basically over even before it got really got started. As I recall, he’d splattered himself over a wide area, his minions all ran, and the entire party just stared wide-eyed in disbelief. Hence, a word of advice: When you are tweaking your magic system, try to remember how your modifications will affect all the various power levels that will come into play.

8. In other words, roll a d6. 1-3 is considered unlucky and 4-6 is considered lucky. The GM consults the die and then simply makes something up depending on what would be good or bad for the spellcaster. For example: “The fireball appears, toasting the orcs where they stand, but after the initial blast, it doesn’t dissipate. It’s just hovering there, and as you look at it, it now appears to you too big to be a fireball. It starts coming toward you. What do you do?” The party will eventually figure out that the mage accidentally gated in a fire elemental, and needless to say, he’s not real happy.

9. This rule also exists in the AD&D (1st edition) Deities & Demigods (1980) in the section on mortality and immortality, where it is also stated that spirits are typically reincarnated by their respective deities, whereas souls are not. This idea probably came from J.R.R. Tolkien (see

10. Drysdale does not speculate on what would happen to an elf’s spirit if he or she were reincarnated as a human.