AD&D Magic System Part 3: Spell Research

Spell design starts with an idea: "It would be cool if..." but should then proceed to introspection:

What if all PC casters in my game memorize this spell every day as many times as possible? Would I be ok with that? Etc. That's because new spells have far greater implications than magical devices in terms of their ability to modify the landscape of the campaign. As such, the number of new spells that this author has seen fit to add to his campaign is small indeed. If you want numbers it's more than 50 and less than 100. Way less than 100.

Pages from the Mages was a Forgotten Realms-y supplement that collated a pile of Dragon Magazine articles by Ed Greenwood and proffered them in book format 'round about 1995.

I picked up a used copy from ebay a few years ago and the thing is serviceable. It does remind me, though, how hard it is to make a good new spell. I think there might be between six and one dozen spells from that tome that I'd say are really worth their material components. Re-skinned damage dealing spells are the worst. But Spider Eyes is cool. I love that it's a low-level scryer the caster cannot fully control and encourages the DM to role play. And it's useful almost ANYWHERE. Sky Hook is also fantastic for the creativity it encourages.

AD&D spells are sometimes explicit, but many times open to debate. Sometimes this is good and sometimes it's not. When it comes to spell text, what you want is a spell like Tenser's Disk or Unseen Servant and NOT like Polymorph Other.

Tenser's Disk and Unseen Servant give us clear parameters including the area of the spell's influence, weight limits and so on. But then leave it up to you to determine HOW you use those forces.

Polymorph Other, on the other hand is too open ended, plus it seems to contradict itself. It also fails to frame the limits of the spell's power in an easy-to-digest way.

This is not a video game. We don't want video game style spells. But hopefully these examples showcase why the vision of the spell and it's mechanical limitations are important to design well.

It is my opinion that new spells are very hard to design, well above magic items because of the DMs inability to gracefully rescind the power should an error be made. Whereas an ill-designed monster might pollute only 20 minutes of the entire campaign and a poorly conceived magic item might wreck a session or two before you send a rust monster after it, a poorly designed (or ruled) spell opens the door to backtracking, iterating and arguments at the table when you'd rather be having a great time.

Spells are, by definition, methods of bending or breaking rules. Bending or breaking rules on a regular basis can quickly eschew reason and balance. When I design a new spell I typically go through the following thought process:

Does this spell replicate another spell in an almost identical way?

  • If yes, do not make this spell.

Does this spell cross boundaries it shouldn't?

  • If yes, do not make this spell until you have considered VERY carefully. (see below)

Is this spell simply a variation of a damage dealer?

  • If yes, the variant you envision must be truly outstanding. (see below)

Does this spell lend itself to abuse?

  • If yes, you must design drawbacks and limitations. (see below)

Is this spell so complicated that the text is longer than the AVERAGE PHB spell?

  • If yes, you probably need to simplify and/or re-scope. (see below)

Once you've iterated to the point that you can answer NO to all of the above questions, you can now ask yourself this question:

Is the utility of this spell easy to grasp AND does it also lend itself to deeper thinking and creativity?

  • If yes, you SHOULD probably make this spell.


Does this spell replicate another spell in an almost identical way?

Self explanatory. Moving on.

Does this spell cross boundaries that it shouldn't?

Examples of this are, IMO, best taken from Magic the Gathering. Counter Spells are the domain of blue. Direct Damage is the domain of red and (at a cost) black. Big Cheap Creatures should be green and so on. In AD&D terms, healing is the domain of clerics NOT magic-users. Nature magic is usually for druids. Undead are not affected by mind-affecting spells: ever.

While wizard-healing and charm-undead spells HAVE both been published, I would not allow them to enter my game without serious drawbacks.

Bending rules at HIGH cost is one thing. Breaking them is another.

There is a single Magic-User healing spell I currently allow in my game: taken from p94 of The Storyteller's Arcana by Scott Taylor. This brilliant spell allows a magic-user to drain 1d6 dmg for every 2 LEVELS of the caster from himself and apply that as healing to the target. Think about this carefully and then remember that magic-user HD is a d4 and caps at 11th level, with only 1 hp / level thereafter. The more powerful the caster, the more likely he / she is to suicide! Perfect.

Is this spell simply a variation of a damage dealer?

Is summoning a giant cat tongue to unavoidably lick for 1d4 dmg per round for as many rounds as the caster has levels different enough from other damage dealing spells that you want to add it to your campaign? I cannot answer this question for you. I can only tell you what I would do...but I will only do THAT via snail mail and only if you enclose $20 with your inquiry.

Does this spell lend itself to abuse?

Another way of thinking about AD&D spells is that they almost never give you something fantastic for free. Unless it's VERY high magic. And even then...even then. Look at Teleport. You always run a risk of death--no matter how small. Look at Permanency. Look at Wish. Look at the casting cost for Augury (100 GP every time). Look at Wall of Force (10,000 GP every time). Unseen Servant BEGS to be abused but can only carry 20 lbs and can only stray 30' from the caster...which limits it correctly. You must make the PCs pay for degenerative mechanics OR place severe limitations upon them...or you will rue the day.

Is this spell so complicated that the text is longer than the AVERAGE PHB spell?

I've read many homebrew spells (even some authored by illustrious graybeards that were influential at TSR in the early days) that attempt to cover every contingency in the text while interacting with things on many levels. Restrict your scope and focus the spell. Do this until you can clearly explain the spell in three hundred words AT THE MOST. Read Polymorph Other again and know that even Gary got it wrong sometimes!

It's also important to compare the new spell to the power of other spells in the game. Quick litmus tests for power are:

  • Does my spell allow a saving throw?

  • If it does, is the spell negated or halved?

  • What is the range? Reduced range is meaningful!

  • Does the caster have to hit/touch the target?

  • How narrow or broad is the spell's utility?

  • Is this a spell you'd ALWAYS want to memorize vs a spell with a specific use case?

  • How expensive are the material components? Increase to mitigate abuse!

Here are two examples of spells I created. I use both in my AD&D campaign and have had both good response and good results during play:

Behold (Illusion/Phantasm)

Level: 3 Components: V,S

Range: 6" Casting Time: 3 segments

Duration: 1r/level Saving Throw: Neg.

Area of Effect: Special

Explanation/Description: This spell affects a 2"x2"x2" area and any number of creatures whose total HD are equal to or less than twice the illusionist's level. All affected creatures that do not save have a magical eye opened, usually upon their head, or anywhere that the illusionist desires.

This eye cannot be closed, sees clearly even in total darkness and cannot effectively be blindfolded by any means. Although the spell will not affect creatures immune to mind-affecting spells (such as undead) it can cause normally blind creatures to see and is in fact a direct channel to the illusionist's magic. For the duration, affected targets will certainly see any and all illusions the illusionist casts, even if they close their eyes, turn away, cast darkness, hide behind shields, or bury the eye in the sand. All saving throws against illusions/phantasms during this time are penalized by -1.

Note that the spell is temporary and does not cure blindness. Sightless creatures will be sightless again when the spell expires. (word count 167)

Consecrated/Desecrated Ground (Abjuration)

Level: 3 Components: V, S, M

Range: Touch Casting Time: 1 turn

Duration: Permanent Saving Throw: None

Area of Effect: 5' x 5' per level

Explanation/Description: This spell protects an area from incursion by those of alignments that differ from that of the caster. Any such being attempting to enter is allowed only ONE roll. Failure means they are barred from entry until the consecration/desecration is neutralized.

NEUTRALS who set foot on either consecrated or desecrated ground have a 55% chance of being forced back due to deep cosmic unease. NEUTRALS who enter suffer –2 on ALL ROLLS (or 10% on percentile rolls) for as long as they remain.

GOODS entering desecrated ground have a base 0% chance of being forced back, which INCREASES by 10% for each level below 11th, so that a 1 HD good creature has a 100% chance of being barred. GOODS who enter suffer –1 on ALL ROLLS (or 5% on percentile rolls) for as long as they remain.

EVILS entering consecrated ground have a base 100% chance of being forced back, which DECREASES by 5% for each level below 11th, so that a 1 HD evil creature has a 50% chance of being barred. EVILS who enter suffer -3 on ALL ROLLS (or 15% on percentile rolls) for as long as they remain.

Consecrated/Desecrated ground can be dispelled by a Remove Curse with a probability of success equal to 10% per caster level. Material component is 1 vial of holy/unholy water per 5’x5’ section.

Great Anthony, but how do you go about allowing the PCs to actually MAKE the spell?

I handle this the same way I handle spell research for existing spells: with my handy-dandy spreadsheet calculator. I have to admit that I had to fix the calculator for this blog post as I discovered it was not working as intended. Therefore if you didn't download the newest version of my DM's familiar TODAY, you should probably re-download to make sure your version is correct.

Glancing over the steps listed in the left column, you can see that this is very straightforward. Enter the value of the library or temple. If no temple or library is owned (see the columns with RED headers: Existing Spell No Library & New Spell No Library) the researcher must pay exorbitant fees to consult books/scrolls in various libraries, plus travel, plus the cost of securing temporary study locations, plus the cost of lost notes, returning to the various libraries, paying again, and again, and again, and again, after each iteration and failure, etc.

In the case of Pray-for-it types without a temple, the sacrifice and prayers must be so outlandish as to get the deity's attention in spite of the cleric's low-standing [and/or] lack of a congregation.

The base cost per week runs 400 GP x the Spell Level (Example: 800 for 2nd level and 3,600 for 9th). For NEW spells the rate is doubled. If the PC has no temple or library, the rate is 2,200 GP x Spell Level and 4,200 GP x Spell Level for NEW spells.

The minimum number of weeks (at the above rate) = the Spell Level +1

For each additional week above the minimum, the odds of success increase by 1%

If the researcher wants to throw money at the problem they can select a success modifier (from +10% up to a maximum of +40%). Adding +10% to the chance of success costs 2,000 GP x the Spell Level. +20% costs 4,000 GP x the Spell Level. +30% costs 6,000 GP x the Spell Level. And +40% costs 8,000 GP x the Spell Level. Note that this money is not recoverable and does not add value to an existing library or temple! It shows up in the table as "Added Cost" and will be piled onto the the Total Cost for step #6.

To research spells of various levels, the library or temple must meet minimum requirements in terms of value (which equates to prestige and or esoteric information). Thus, in order to research a 1st level spell, the researcher's shrine or library must be worth at least 2,000 GP. If the researcher is trying to make or pray for a NEW spell, the shrine or library must be worth at least 6,000 GP.

The progression of minimum values (from 1st thru 9th level) is shown in this screen grab from my XLS sheet.

If the value of the library or temple does not meet this requirement, the researcher is treated as NOT having a temple or library for all cost and success calculations.

Note: the BASE cost of the research is ADDED to the library/temple value at end of research, so even if the attempt fails, future attempts have a greater chance of success.

Also note that if the library/temple value EXCEEDS the minimum value requirement, the odds of success are greatly improved. The difference between the minimum value and the starting value of the library/temple is divided by 3,000. This result is then multiplied by 0.01. Therefore if the value of the library/temple exceeds the minimum requirement by 6,000 GP, the chance of success is improved by 2%. All of this nonsense is handled by the spreadsheet and I don't have to give it anymore thought.

You can see where the sheet gives you the new value of the library/temple AFTER the research attempt is complete (see step #8).

Below this, notice that the researcher's INT (Study-for-it types) or WIS (Pray-for-it types) is entered and provides a 1% chance of success for each ability score point. Researcher level also gives a one to one boost to the odds. Although it isn't called out on the sheet, I impose the logical constraint that a researcher must be of sufficient level to CAST the spell they are researching.

Spell level is multiplied by 2 and this number PENALIZES the odds of success.

The final odds of success are listed at the bottom and rounded to the nearest whole number. The odds of success for those without a library/temple are halved. And in the cases of NEW spells, the odds are reduced to a third of normal.

You will find that spell research is insanely costly and grants poor odds for casters trying to research basic spells at low level. This encourages a go-find-it adventuring approach. But for high level researchers with valuable libraries/temples trying to fill gaps in their low-level rosters, the cost will be low and success guaranteed...although a % to Know Spell roll will STILL be required to LEARN a spell that is successfully researched.

On the up side, the value of a library can be much lower for spell research than for magical item fabrication.

Just remember that all of this was inspired by other writers. I simply took those ideas and built my own tables to form a system. Therefore, none of the above is BTB. It is an interpretation of BTB designed specifically for my campaign.

Lastly, I wanted to talk about spell languages.

My game runs on mostly unique languages (which you can read about here). As part of this system of languages, I've abolished the concept of Read Magic / Write Magic. When you bring a cleric you assume he can turn undead. When you bring a magic-user you assume he can read magic. Reading and (especially) writing magic "usually" happen more in downtime threads than they do in dungeons. Nevertheless, the dungeon is hard enough that I don't expect magic-users to sacrifice a spell slot just to read a language they've arguably been studying every day for 16 hours a day for decades.

So much for BTB, eh? I am an advocate for mechanics. And I could possibly write a persuasive essay in favor of using Read/Write Magic if any good debate student. But my supports would be flimsy. I encourage you to use Read/Write spells if you like them. It's just that I've gone a different way.

I've created a language called the Unknown Tongue for Magic-Users. The Secret Tongue is for Illusionists. Each sect of clerics has their own (un)holy writ as well (similar to latin perhaps in the age were no one could read). Witches, Death Masters and Incantices each have their own cribbed form of language that gathers symbols from other tongues. Thus, in my game, the language is part of the class BUT STILL OCCUPIES A LANGUAGE SLOT according to INT. Therefore my magical languages are akin to Thieve's Cant insofar as they are unavailable except to members of the appropriate class.

This works great for me, maybe not so much for you.

That pretty much wraps my three-post series on AD&D's magic system. I hope it was helpful or at least gave you ideas on how to make it work in your own game. Thanks so much for reading and have a Happy All Hallow's.

peace and happy gaming.

101 views6 comments