Monthly Archives: March 2012

Castle Greyhawk Elevation

Both Grognardia and Stefan Poag recently mentioned WGR1 Greyhawk Ruins. Apparently this is one of those megadungeon publications that I had never heard of, and does not seem to be very highly thought of by Greyhawk fans (or some Grognardia commenters: exhibit 1, exhibit 2). The Amazon reviews, however, are uniformly positive (even the two star review criticizes it as “a location not an adventure,” which sounds like a compliment to me).

What was the point of this post? Oh yeah, while I was looking for images of the product (because I was not sure if it was a boxed set or something else; the answer is a paperback), I found this evocative elevation picture on the Wizards site:

Click to make larger (source)

The image can be found in this free gallery (cached here since it is freely available), and seems to be from some Third Edition take on Castle Greyhawk. I kind of just want to take that elevation and draw my own maps for it.

Shared Languages

Catacomb Librarian is unhappy that the OSR focuses on D&D. I am somewhat sympathetic to this point of view as, despite the fact that traditional D&D is my game, I do not think that it is THE fantasy game. I wish that more people would talk about other games so that I can learn about them. I learned about Traveller from Grognardia. I decided I might be interested in Warhammer because of Chris Hogan’s SBVD. I learned about Iron Heroes from Monsters & Manuals.

The first two of those led to purchases, and I will likely buy a copy of Iron Heroes at some point too. Even though practically speaking I’m not likely to play Warhammer or Traveller (though you never know), I still think I can learn from them. Despite the fact that I am interested in other games, D&D is the most common game, and this makes the language of D&D the shared language of tabletop RPG players. It is a common baseline that you can assume, much as you used to be able to assume that an educated person would have read Plato. (That’s also why it’s dangerous to create giant infodump settings, because doing so raises barriers to entry that are not present if you just say “Moldvay Basic” or “Third Edition Core,” and it’s why the development through play OSR dogma is so valuable.)

I’ve been running a Fourth Edition hack game for the past 8 months or so (2-4 times per month), and I’m just now getting to the point where I feel like I understand the system and its implications. And, despite all its changes, this is a system that still shares some DNA with the traditional D&D I am more familiar with. A game with fewer similarities (like, say, GURPS) would presumably take even more time. Not to mention the fact that you need to find other players that are similarly interested if you want to actually play. G+ and ConstantCon games make this easier, but still not easy (compare uptake of FLAILSNAILS and non-FLAILSNAILS games). In any case, playing via videoconference is still a poor substitute for playing at a real table.

This same dynamic is part of the reason why many people who are used to traditional D&D are turned off by Third and Fourth edition: they make the language less shared. They move or change the landmarks. They fracture the community not just in the sense that some like save or die and some don’t, but in the sense that they change meanings. It’s almost as if half your friends start speaking Japanese. You now have to learn Japanese too if you want to participate, and whether or not you think Japanese is an elegant or beautiful language, it is still work to learn.

There are only so many hours in the day. In addition to gaming, I’m heavily into weight training (a hobby that rules your routines like few others). I also like to read literature and history and spend time with my loved ones. I collaborate with some academics on social science research. Oh, and there is that pesky job thing that occupies 40+ hours of my life every week. When I do get to game, I want the rules to fade into the background.

To quote Noisms from Monsters & Manuals on new rulesets:

Nowadays, the prospect of picking up a 200 (or more likely 400) page-long tome of new rules fills me with dread and boredom rather than excitement, and I find that my attention span is only a fraction of what it used to be.

It’s not that I hate other systems or think that D&D is perfect (though I do think the pre-AD&D versions are pretty elegant), it’s that D&D is the common tongue. Like English, we are stuck with it for now.

Abulafia Accounts

Abulafia (also known as is a nifty wiki that allows one to create automated random tables. Zak over at Playing D&D With Porn Stars has mentioned it a few times (like here). However, I can’t seem to create an account. Does anyone know, by any chance, what the procedure is? Clicking on the log in / create account link only seems to provide login fields.

Am I missing something obvious here? Or is there someone I need to ask?

The Shadow People

This is a strange little book. It’s a mix of creepy urban fairy tale and paranoid 60s counterculture novel, terrified of the encroaching police state. It’s set in Oakland, on Telegraph Avenue. Everyone at Cal is studying physical education and criminology due to the wave of urban crime. The main character Dick Aldridge (who works for a “hip newspaper”) sets out looking for his missing girlfriend Susan.

As a novel, I can’t say that The Shadow People is very good. Important characters show up by coincidence with no explanation. Setting details are dropped with extraordinary specificity by characters that don’t seem like they should know such things. I never did much warm to Dick or Susan; they don’t feel like fully-formed characters and I never found myself really caring what happened to them.

That being said, there’s lots of interesting inspiration to be taken from the middle section on the underearth; for example: “There is always a water barrier between our world and Underearth” (page 33). I also like the take on faeries. Elves are described as small and boneless, and are malicious, somewhat stupid creatures, with countless taboos. They hate metal that comes to a point. They also don’t like nudity, because “it’s too much like the light” (page 124).

The Secret Commonwealth by Robert Kirk is mentioned several times. I have a copy of the wonderful recent NYRB edition which I highly recommend. From the NYRB description:

Late in the seventeenth century, Robert Kirk, an Episcopalian minister in the Scottish Highlands, set out to collect his parishioners’ many striking stories about elves, fairies, fauns, doppelgangers, wraiths, and other beings of, in Kirk’s words, “a middle nature betwixt man and angel.” For Kirk these stories constituted strong evidence for the reality of a supernatural world, existing parallel to ours, which, he passionately believed demanded exploration as much as the New World across the seas. Kirk defended these views in The Secret Commonwealth, an essay that was left in manuscript when he died in 1692.

This is clearly the tradition that St. Clair is drawing from. The entire text of Kirk’s book is available for free on the web, but I find that version rather unpleasant to read and much prefer the hard copy. I had read bits and pieces of The Secret Commonwealth before, but finishing The Shadow People prompted me to go back and read the whole thing (I also mentioned The Secret Commonwealth recently, regarding elf height).

There’s a bit of the Orpheus legend woven into The Shadow People also (Dick going down to the underearth to rescue his girlfriend). The gray dwarf’s plan is just so twisted that I want to base an entire dungeon area around it. Overall, The Shadow People is probably only of interest to Appendix N chasers, but I’m glad I read it.

Counterspells Draft

Rather than take a standard action, a magic-user may ready a counterspell. When deciding to ready a counterspell, the magic-user must select a target caster and must be aware of the target (this awareness can either be via direct physical perception, or can use an intermediate medium such as a scrying spell). From now on, the target will be referred to as the caster. As when casting a spell, no movement is permitted. A magic-user may not use a counterspell when surprised. Only magic-user spells (that is, not cleric spells) may be countered. Countering spells may result in gridlock while both magic-users wait for their counterparty to take the first action. This is intended.

There are two kinds of counterspells, temporary and permanent. A temporary counterspell prevents the casting of a spell, but does not wipe it from the target’s mind (or consume the scroll if they were casting from a scroll). Any magic user may always attempt a temporary counterspell. A permanent counterspell wipes the spell from the caster’s mind as if it had been cast or consumes the scroll. A permanent counterspell costs a spell slot of equivalent level from the countering magic-user. For example, if the target is casting levitate (a second level spell) and the countering magic-user wishes to attempt a permanent counterspell, the countering magic-user must also have a second level spell prepared, though it need not be the same spell. That spell is expended during the countering attempt (whether the counter is successful or not). If the countering magic-user does not have such a spell prepared, treat the countering attempt as temporary.

If the target casts a spell while the magic-user is in countering mode, a counter attempt may be made. If the spell being cast is one that the countering magic-user has in a spell book, the countering magic-user will be aware of which spell is being cast before they must decide whether or not to attempt a counter (though they will still be unaware of details such as the spell target). The caster must make a saving throw versus spells, with penalty equal to the highest level spell the countering magic-user can prepare. If the save is successful, the spell goes off as normal (i.e., the countering attempt has failed). If the caster fails the counterspell save, the spell is countered.

If a 1 is rolled on the counterspell saving throw, the caster must roll on the counterspell catastrophe table (see below). If a 20 is rolled on the counterspell saving throw, the countering magic-user must roll on the counterspell catastrophe table. Engaging in sorcerous combat is always dangerous, and opens a countering magic-user to a magical counterattack. Rather than rolling on the catastrophe table, referees may also make something up that is suitably nasty.

New Magic Item: Counterspell Scroll

A counterspell scroll is a “bottled” version of a permanent counterspell. A counterspell scroll has a level, determined by the scribe. Costs are as per Holmes scroll creation rules (page 13): 100 GP and 1 week of time per scroll level. A higher level magic-user may produces a lower level counterspell scroll at correspondingly lower cost. Thus, a seventh level magic-user may produce up to fourth level counterspell scrolls. Such a fourth level counterspell scroll would cost 400 GP and take four weeks to produce. Counterspell scrolls are consumed when used even if the target caster makes the counterspell saving throw.

Counterspell Catastrophe Table

  1. 1d6 psychic damage
  2. Lose another prepared spell (determined randomly)
  3. Thrown 2d6 feet in a randomly determined direction
  4. Struck blind for 1d4 turns
  5. Screams due to severe pain (wandering monster check)
  6. Age 1 year (may include hair and fingernail growth)
  7. Knocked unconscious (as per sleep spell)
  8. Mind violation (as per ESP spell); duration 12 turns
  9. Develop spell allergy: future casting of countered spell always requires save
  10. Polluted luck: penalty of 1 or 2 (depending on die) to all rolls for 12 turns
  11. Weakness: movement and all physical ability scores halved for 12 turns
  12. Warped reality: all missiles within 50′ arc to target the subject for next turn
  13. Feeblemind (as per spell) for remainder of encounter
  14. Hostile 1 hit die elemental summoned (determine element randomly)
  15. Strange gravity: as if in 0 G environment for 1 turn
  16. Riposte: other magic user gets free spell or attack against subject
  17. Entrancement (as charm person, another save applies)
  18. Moon curse: exposure to moonlight causes 1d4 damage (remove curse ends)
  19. Sun curse: exposure to sunlight causes 1d4 damage (remove curse ends)
  20. Hostile 1 hit die demon summoned

Two Hanging Scrolls

Michael over at The Grumpy Troll posted this great John Blanche picture a while back that is the inspiration for an entire new campaign. The intricate line work in that piece reminded me of some eleventh century Chinese hanging scroll paintings. Check out the scale, especially in the first piece (Early Spring). Do you see the temples? What about the person? This says something about the relationship of the individual (and perhaps the adventurer, in a fantasy gaming context) to the cosmos.

Early Spring by Guo Xi (11th Century)

Buddhist Temple in the Mountains by Li Cheng (10th Century)

Wacky Multiclassing

I’m not generally a fan of multiclassing or overly flexible skill systems. I like archetypes, and I don’t like systems that reward too much synergistic optimization. I like my game choice to occur during play, not before play. That being said, here’s an interesting collection of links:

Taken together, these gave me some ideas that I want to jot down. So here they are. This system below is meant to replace the classes (that is, characters would just accumulate perks from the list).

When you gain a level, choose any two of the following benefits:

  1. +1 hit die. This is a d6. Max hit dice is 10.
  2. +1 attack bonus. Max attack bonus is +10.
  3. +1 spell slot of level N. To qualify for a spell of level N, you must already have N + 1 spells of level N – 1. For example, to take a third level spell slot, you must already have two second level spell slots. The number of slots for spells of level N must always be less than the slots for level N + 1. The first time you take this benefit, you also gain the read magic and scribe scrolls abilities. A color of magic system would work well here too. Maybe call spell levels ranks to distinguish them from experience levels.
  4. Advance a rank in turn undead & demons (whatever that means); the first time you take this, you gain the ability to scribe protection scrolls based on some subsystem.
  5. +1 bonus to AC. Max +5. AC can never go below 0 no matter what (or above 20 if using an ascending system).
  6. Thief ability: pick or roll.
  7. +1 backstab multiple. Default is 1. Max 5.
  8. Psionic ability: pick or roll.
  9. Long list of feats or whatever (easily ignored by those of us that like simpler systems).

0 level characters have one d6 hit die. When making a first level character, pick two benefits exactly as if you were gaining a level.

Max level is 25. I first thought 20 would be nice and iconic, but 25 is required so that a dedicated magic-user can gain the ability to prepare sixth level spells (well, 22, but the magic-user might want an extra hit die or something along the way, and 25 is a nice round number). Use your favorite XP progression table. Actually, you could set the max level to reflect whatever power level you want the campaign to top out at.

Anyone can use any weapon and any kind of armor, but there are drawbacks (chance of spell failure, penalty to thief skills). If you don’t put ranks in the combat competencies, that two-handed sword is not going to do you much good.

Layer on races at the beginning if that’s your kind of thing. Humans would probably get one or two extra benefits at first level to compensate (depending on how interesting the racial bonuses are).

To get a truly schizophrenic interesting character, you could even roll on the benefits table during level-up rather than choosing.

I actually wouldn’t be surprised if Fifth Edition does something somewhat similar, as it would allow the game to include 4E style powers as a replacement for something like a standard attack bump. This would make a traditional fighter compare favorably in terms of power levels to a fighter built with lots of powers and maneuvers. If the 5E designers go in this direction, I’m guessing it probably will not be quite as open (as in, there will probably be a suite of options per class, like in Zak’s post).

Tower of the Stargazer

[Caveat lector: some partial module spoilers contained below.]

As mentioned previously, I ran The Tower of the Stargazer this past Monday using Moldvay Basic D&D for my standard group of players. Character generation was 3d6 in order. No adjustments were allowed. The PCs included two clerics, two magic-users, two thieves, and a halfling. That’s right, no fighters. Players seemed to choose class based on their highest ability score, making class selection effectively random as well.

This is another way that basic D&D is humanocentric that I had never noticed before: none of the demi-human classes have obvious prime requisites in the way that the big four human classes do. When players rolled intelligence highest, they immediately gravitated towards magic-user (despite my saying several times that playing against type was completely viable due to the decreased emphasis on ability scores). When a few replacement characters were rolled up following some PC deaths (more on that below), one player did choose to make an elf, but only because he rolled a 16 strength and a 17 intelligence.

I said before that I was trying to avoid provisos and exceptions, but I did introduce three house rules. The first was LotFP alignment restrictions (elves and magic-users had to be chaotic and clerics had to be lawful). I wanted to emphasize that alignment was about cosmic forces, not character behavior or morals. The second was a save versus death ray when reduced to 0 HP by damage. Make the save and the character is unconscious rather than dead. This rule was used several times, and it saved one retainer (but none of the PCs). The third was my first Vancian magic variant, a save versus spells to retain a memorized spell after casting. None of the magic-users cast their spells though, so it didn’t come up.

One cleric rolled exactly enough GP for Plate, so he spent all his money on armor and armed himself with a heavy branch as a club (you can’t get much more murderhobo than that). In general, buying equipment was much smoother and quicker than I expected. I told them to not worry about rations since this was a one-shot, but that they would definitely need light sources. I suspect this process was so smooth because all available equipment is displayed on one page (including weapons and armor) and everything is priced in terms of whole GP (no fiddly fractions or multiple coin types). Game designers take note.

I also included two retainers generated using Meatshields! so that there would be some backup characters for players to control if primary PCs were killed (I wasn’t really worried about combat strength). I was originally going to include four retainers, but seven players showed up (six of my normal players and one curious coworker who was entirely new to tabletop RPGs). 11 adventurers seemed like a crowd for the tower.

The session went very well. Several players told me how much more they enjoyed the simple rules and quick character generation. One player said that PC death was the fun part and that being unhappy with character death was like being unhappy that eating an apple leaves a core. This same player wondered if he would be happy with a B/X character for an extended campaign though.

In hindsight, there are a few things that I would have done differently if I was running this adventure again. The first is that I would have included more retainers. The murderhobo cleric was killed by a poisonous spider and then one of the hirelings was knocked unconscious, leaving only one retainer for the cleric’s player and no backups for other players.

When more PCs died because of an exploding door trap, there were no more retainers to take over. I wanted to get the players back in the game as soon as possible, so I had them roll up new characters and begin at the entrance of the tower. This was a mistake, both practically (splitting the party) and atmospherically. Prior to the introduction of the replacement characters, all the PCs had been keeping pretty close to each other, and I felt like this emphasized the strangeness and danger of the tower. After the introduction, everything felt a bit more scattered (though it was still fun). Lacking extra retainers (obviously, you can’t have unlimited retainers, even if I had included more), it would have been better for the new PCs to be discovered trapped in stasis by the wizard 60+ years ago (I wish I had thought of that during play).

I would also create some diagrams for myself of the lever possibilities for the treasure room force field puzzle. Before the session, I reread that area description several times and thought that I had the details down, but during play I still found myself referring to the description continuously and confusing myself. They didn’t solve the puzzle.

There is a set of generative tables in this module for creating labels for crates in a storeroom. The tables creates results like “the rib … of a sailor … who collected … happiness.” I knew that I should pregenerate a set of results, but for whatever reason I didn’t. Rolling on these tables during play felt artificial. After two results, one of my players asked if there were any labels that didn’t follow that formula. Totally my fault here; I knew it would be an issue. (Being a teaching module, I’m surprised the text did not suggest pregeneration though.)

You need to have a backup plan ready in case the players decide to free the wizard and take the 100 GP reward, because that ends the adventure. My players didn’t choose to do that, but if they had I don’t think the adventure would have been nearly as enjoyable. I had some basic ideas for several more encounters in the wilderness, but they were not very well planned out. Since this was a one-shot, there was not a populated sandbox ready for them. I’m glad I didn’t have to improvise.

The ghost encounter was not very impressive the way I ran it. We didn’t have time for a full game of chess, so we just diced for it. I’m not a big game player other than D&D, so there was not an immediately obvious alternative. This encounter needs a carefully planned game or a redesign. Even Raggi admits in the text that during his play testing, a number of his players were bored while one or two concentrated on the challenge. I still think this was totally my fault though, as I foresaw the problem, I just wasn’t able to find (and learn how to play!) another game prior to the session. An empty room plus a generic secret door probably would have worked better. It wasn’t horrible or anything, it just wasn’t very creepy. Live and learn.

I hope we get a chance to continue the adventure, as they were just about to explore the real laboratory and the telescope when we had to stop. Also, the clerics were brainstorming ways to kill the trapped wizard, and it would have been interesting to see if they could pull that off without a TPK.


My regular game group agreed to play a one shot game with a traditional rule set. We normally play a 4E hack. My favorite version of D&D is B/X as edited by Moldvay, Cook, and Marsh. So, to give them a taste of the old ways, I wanted to find something close to that. Most of my current players are not aware of any differentiation between anything prior to and including First Edition.

Why not just use B/X directly? Well, that is actually what I ended up doing. Since I only have one copy of Moldvay Basic though, I had hoped to find a free legal PDF of a clone that was “close enough” so the players could check it out before play. As it turned out, that was not necessary. I just chaperoned character creation along with a few copies of key pages. It turns out that pages B5 through B13 are all you need for a B/X Player’s Handbook (with pages B16 and B17 for magic-user and elf spells). Cleric spells would also need to be added for a game that proceeded past first level, obviously.

Making all the characters from scratch was relatively painless and took less than 20 minutes. In total, there were seven players in addition to me. The group included an experienced 4E player, two players who had only ever played D&D in my normal 4E game, several who had played a few different editions, and one who had never played a tabletop RPG ever. They rolled 3d6 in order, picked a class, rolled for money, and bought equipment.

I was going to run either Keep on the Borderlands or Stonehell, but Aplus helpfully suggested Tower of the Stargazer in response to a G+ query. This is one of my favorite modules, new or old, and I have never had a chance to run it. It is also relatively small and self contained, and I felt confident that they would get to some of the interesting areas within a three hour session. It went well; I’ll write more about the session in another post.

Though I decided to use Moldvay, this is what I was looking for in a ruleset, with reasons in parentheses:

  1. Highest level spells as 0E: magic-user max 6, cleric max 5 (power level)
  2. Race as class (quicker chargen, distinctive demi-humans)
  3. No cleric spells at first level (against the perception of cleric as medic)
  4. Five saving throw categories (atmosphere)
  5. Variable hit dice (I think OD&D’s all d6s would be confusing)
  6. No class ability score requirements (quicker chargen)
  7. No ability score modification (quicker chargen, no optimization)
  8. Freely downloadable (convenience)
Perhaps surprisingly, I don’t care one way or the other about ascending versus descending armor class.

Here is the disposition of various games regarding my criteria:

  • B/X D&D is perfect but not freely downloadable
  • Labyrinth Lord gives first level clerics a spell
  • Labyrinth Lord also has spells higher that the 0E levels
  • Equipment and armor are more complicated in Labyrinth Lord
  • Swords & Wizardry uses a single saving throw
  • S&W Core has a rule for the 5 saves, but separates race and class
  • S&W Core also has spells higher that the 0E levels
  • Lamentations of the Flame Princess has strange attack bonuses
  • It also has spells higher than the 0E levels
  • All the 0E clones (e.g., S&W:WB) are out because of the hit dice
  • Labyrinth Lord Original Edition Characters is not freely downloadable (yes it is, see here)
  • Original Edition Characters is also out because of the hit dice
  • The Microlite games seem to be very much their own thing
  • Basic Fantasy has separate race and class (and odd cleric spells)
  • Dark Dungeons goes much too high in level (Rules Cyclopedia clone)
  • Dark Dungeons also has skills
The highest level of spell available is probably the least important of the criteria, as they are not likely to even notice. The rules directly exposed to first level characters are clearly the most important.

I know I could house rule anything, but I wanted their first traditional D&D experience to be absent provisos and exceptions as much as possible (though I believe I am constitutionally incapable of avoiding house rules entirely; more on that later in the actual play report). I just wanted the rules to fade into the background.

Practically speaking, these seemed to have been my options within the clones (had I not gone with Moldvay):

  1. Swords & Wizardry Core and live with the single save and separate races 
  2. Labyrinth Lord and live with the 1st level cleric spell and SRDisms
  3. Basic Fantasy and live with separate race and class
It seems like Basic Fantasy comes closest to traditional B/X, the only major deviation being the separation of race and class (sixth level cleric spells are odd, but not a big problem; at least magic-user spells only go up to level 6). This is surprising, because Basic Fantasy is a clone that I have almost no experience with (I think I only glanced at the website briefly a few months ago and downloaded the PDFs).

I’m sure any of the clones or neoclassical games would have worked, and this is not to be taken as a dig against them. But it is interesting too see how the various options compare to my idiosyncratic preferences, and how many of the clones diverge in various ways.

2012 04 24 edit: note about free text-only version of Original Edition Characters.

Two Vancian Magic Variants

Prompted by this discussion over at Jeff’s place, I had some ideas for Vancian magic house rules. Here they are again, cleaned up somewhat.

Variant 1: Spell Retention

Prepare spells as normal, standard charts. However, when a spell is cast it is not automatically wiped from the magic-user’s mind. The magic-user gets a save vs. spells (penalized by spell level) to retain it. At low levels this means that about 30% of the time spells will not be expended the first time they are cast. For powerful magic-users, it makes low level spells closer to at-will powers (but not actually unlimited). Thus, the resource management aspect of the class is not destroyed.

What’s the downside? If the magic-user rolls a 1 on the saving throw, that is considered a spell fumble and they lose the spell and must roll on a magical mishaps table (or suffer some other campaign-specific penalty). Fumbles can only occur when casting spells of the highest two or three levels that can be prepared. That is, by the time a magic-user can cast fourth level spells they have completely mastered first level spells and can no longer fumble their spell retention save. The highest level spells, however, always carry some level of risk.

Variant 2: Improv Casting

Prepare spells reliably as per the traditional rules. Any unprepared spell that the magic-user is familiar with (i.e., has in their spell book) may be cast but requires an action and a successful save vs. spells (penalty equal to spell level as above). Failure and the action is wasted, fumble and bad stuff happens. Further, a fumble occurs on any die roll equal to or less than the spell level. Thus, casting an unprepared third level spell would fumble on a roll of 1, 2, or 3. Improv fumbles should always at least prevent the magic-user from trying the same spell again until they can return to their spell books and puzzle out what went wrong.

“I saw this one thing this one time and it kinda went like this…”

Any spell that the magic-user has witnessed may also be attempted (for example, if the magic-user has seen another magic-user cast fireball). This uses the same rules as casting unprepared spells, but the save penalty and fumble threat range are doubled.

Both of these variants increase the magic-user’s power or versatility, but also expose them to fumbles. Many people prefer magic to include an element of danger, so that may be a feature rather than a bug. I like this mechanic because it is thematically coherent, is simple (no bookkeeping), and uses saving throws (which are like my favorite thing ever).

This does probably make magic-users more powerful, so if that is a problem you should control spell acquisition carefully (for example, no free new spells on level up). If one was so inclined, one could use both variants together, as they cover different aspects of casting, but that would result in a larger divergence from the traditional game.

I gather that there is a mechanic in the Dungeon Crawl Classics RPG which has a separate fumble table for each spell, so that might provide some inspiration, though a fumble table per spell seems like overkill to me.