Friday, 18 May 2012

Fatigue based spellcasting in old-school D&D

As I recently mentioned, I've been running James Raggi's the Grinding Gear for a small group. I've been running the module as a one-shot, which means that I've not had to insert it into an existing campaign -- giving me free reign on the setting. Obviously, it being a one-shot, the setting isn't that important, and hasn't played a huge role, but the vague background I've imagined is a weird fantasy-esque, early modern setting, with no demi-human PCs, no divine magic, and much more limited arcane magic than one typically finds in D&D.

This slightly atypical setting, combined with the fact that it's a one-shot (well, a two-shot now) means that I've been able to play around with a few ideas for new rules. I've already talked about the introduction of mixed-class characters and firearms, so the remaining (and probably most experimental) house rule is my system for fatigue-based spellcasting.

Regular readers will know by now that my favourite D&D-related pastime (well, joint favourite with actually playing the game!) is messing around with new spells and rules for magic-users! Something of a hobby in itself ;)

I've been mulling over ideas for some time on how one could simply (i.e. without having to rewrite everything!) adapt the D&D Vancian spellcasting system to give it more of a low-magic, dangerous edge. From what I've seen of the magic system in Dungeon Crawl Classics, that's kind of the vibe he's going for, but of course that's a very involved rewrite of the magic rules and all the spells, so not what I had in mind.

At first I was contemplating some kind of spellcasting roll, with the possibility of nothing -- or varying degrees of bad / weird side-effect -- occurring if the roll fails. In the end, however, I came up with something far simpler, with a nod to Fighting Fantasy. It works as follows.

Fatigue Based Spellcasting

In the standard system of magic, as presented in traditional D&D, a magic-user's spells must be memorized in advance and vanish from his mind when cast. Thus a spell-caster's puissance is measured by the number of spells which he can fit into his memory at once.

Using this alternative system of spellcasting, a magic-user does not need to memorize spells. It is assumed that once a magic-user has mastered the formula of a spell, he is able to cast it at will, simply by performing the required gestures and incantations. However, as a counterbalance to this freedom, each spell which is cast drains the caster's reserves, leading to a creeping fatigue and listlessness.

This fatigue is simulated by a hit point cost for each spell cast, depending on its level. Thus, as a magic-user casts more spells, his ability to react in dangerous situations, and to avoid serious damage or death, is reduced.

Spell level        1    2    3    4    5    6    7    8    9
Hit point cost   1     2     4    8      12   16   20   24   28

Hit points lost due to spell-casting are recovered in a slightly different way to normal. A magic-user's player should keep track of both his current hit point total, and the sum of points which were lost due to magic use. After a full night's rest, all hit points lost as a side-effect of spells cast are restored. On the other hand, hit points lost due to spell-casting cannot be restored by healing spells or magic items.

This system enables low level casters (especially those at 1st level) to cast more spells than in standard D&D, while higher level spells, due to their large drain in hit points, will be less frequently used.

  1. The only problem I can imagine with this system, and a point which I'm rather unsure of, is that it makes a magic-user's maximum hit point total extremely important. A 1st level MU with one hit point would be unable to cast spells, for example, and a player who made consistently lucky rolls for hit points as his character advanced would be at a great advantage.
  2. I've been using the standard spell progression chart to determine the number of spells a magic-user knows, rather than how many spells he can memorize. I've doubled the numbers on the chart, so a 5th level MU would know four 1st, four 2nd and two 3rd level spells. This nicely simplifies the question of how MUs learn new spells.
  3. Likewise, I've allowed high INT to grant extra known spells, using the cleric chart for high WIS.
  4. Using this system, a few spells might need re-levelling. For example, given that a 1st level magic-user could be able to cast three spells a day, sleep is probably too powerful for a 1st level spell. I'd move it to 2nd level.


  1. so, with this system you weaken an already weak (hp-wise) class. bad idea. isn't it enough if a MU gets hit twice or so and goes down? i see a very large problem especially at low levels here. you would have to increase the MU's hp but this would treat on fighter's ground. I think this system is not feasible. the intend of having a spellcaster's energy be taxed is familiar but i would not base it on hp consumption. if you want a fatigue-based casting system, what about making a save vs. spells with a penalty equal to the spell cast. if failed, caster becomes tired. if failed a second time, he becomes exhausted instead and if failed a third time, he will fall unconscious, or something like that?

  2. so, with this system you weaken an already weak (hp-wise) class. bad idea. isn't it enough if a MU gets hit twice or so and goes down? i see a very large problem especially at low levels here. you would have to increase the MU's hp but this would treat on fighter's ground. I think this system is not feasible. the intend of having a spellcaster's energy be taxed is familiar but i would not base it on hp consumption. if you want a fatigue-based casting system, what about making a save vs. spells with a penalty equal to the spell cast. if failed, caster becomes tired. if failed a second time, he becomes exhausted instead and if failed a third time, he will fall unconscious, or something like that?

  3. Fighter/Mus become better spell casters compared to single classed MUs under such a scheme at lower levels (if you hav F/MU ).

  4. I like the idea of fatigue-based spellcasting. One thing that I've always liked about Shadowrun is the elegance of the magic system, including the way spells drain the caster of vitality; it's almost enough to draw me back to the game, until I remember all the other fiddly bits I don't like!

    I wonder if you could import a similar system into D&D? Instead of hit point damage, perhaps casting a spell forces the caster to make a saving throw -- perhaps against Paralysation or, well, Spells -- with failure indicating unconsciousness.

    You would have to modify the the saving throw for more powerful spells, perhaps -1 to the roll for each level above first, but it could work.

    1. This is the second mention I've read recently of the Shadowrun magic system. The one time I played that game I hated it (and it might have just been a bad referee). I was also playing a more fighterly character, so I was not really exposed to the magic system. Maybe I should take a look at it again.

      I did, in parallel, come up with a "save to retain spell but fumble on 1" system a while back that I still think is pretty elegant, and sounds similar to a few of the ideas discussed here:

  5. Great idea. Thanks for sharing!

  6. Beastman is right, of course. In order to implement this system, you might have to give every class the same hit point progression, ie. everybody uses the same die for hit points. The fighter uses up his hit points in combat; the magic user exhausts his in spellcasting

  7. One final observation. It might be tempting to use the fibonacci sequence for spell-cost progression (1,2,3,5,8,13,21,34,55). It lowers to cost of spell casting at low levels, but results in 8th & 9th level spells being extremely expensive to cast.

  8. I use an almost identical system, but to mitigate the loss of hit points for spell casting, and to align hit points more closely with their intended concept (not representing damage, but fatigue, skill, etc.), I've given all characters wound points equal to their one-half their Con score, in addition to their normal hit points. Wound point totals do not increase with level. Characters who take hit point damage are becoming fatigued, and when they run out of hit points they start losing wound points (they are too fatigued to defend themselves and are now taking injuries). Once wound points are exhausted the character dies.

    Since hit points do not reflect real injury they are recovered very quickly (1 hp/hour of rest), whereas wound points heal quite slowly (1 wp/2 days of rest).

    While this system works very well, and reconciles the hit point recovery rate with the Gygaxian paradigm, the one problem that I've found with fatigue-based spell casting is that magic users will almost always spam magic missile and never cast anything else. This became such a problem in my campaign that I've returned to the standard Vancian memorization system and imposed penalties for having one spell occupy more than one spell slot.

    Other than that, the fatigue system worked pretty well and had a lot of sword & sorcery flavour; I had the magic user cut his hand with a sacrificial dagger as part of the casting ritual as an offering to the demons of the dark beyond.

    Good luck with your system, I'll be interested in hearing how it works out for you.

    1. Sean- How do you handle Cure Light Wounds spells/potions do they cure Real Hit Points or Wound Points, or both or what? This is important to note what exactly magic is curing- is it "fatigue" or "real damage". If magic just cures Hit points- then it makes Wound Points very important and realistic.

  9. M20 and derivatives have the same system (eg. my M20 Hard Core).

    From M20: Magi can cast any arcane spell, and Clerics any divine spell, with a spell level equal or below 1/2 their class level, rounded up. They have access to all arcane spells in the SRD spell list. Casting a spell of any kind costs Hit Points. The cost is: 1 + double the level of the spell being cast.

    Like people above have said, all classes gain the same amount of hit-points: Hit Points = STR Stat + 1d6/Level.

    Thus, a mage might start with 10 hp and be able to cast 10 zero level spells (cost: 1 hp) or three first level spells (cost: 3 hp).

    I thought it was super elegant and loved it. When I did a one-shot to try it, my players hated it. They felt weaker. I disagreed but that didn't make them happier. We abandoned M20 and returned to D&D 3.5 (and eventually I found new players and started using Labyrinth Lord).

  10. I always liked the Vancian approach to magic purely for nostalgia but I also like the idea of known spells. So what I did was to combine the two. I use the spell progression chart in 3rd edition D&D for Sorcerors ( 2 known spells at first level- but able to cast 3 spells per day with no bonus for high Int) Then what I do is also allow the magic user to carry a spell book that contains their spells. They can choose to memorize a spell they dont KNOW at a cost of two spell slots for one memorized spell ( ex- if a magic user Knows Sleep and Magic Missile and can cast them in any combination up to 3 times a day- but decides they want to memorize Detect Magic, then it costs two spell slots- so the magic user can only cast Detect Magic and one of his Known spells that day). This allows much more versitility for the spell caster without completely throwing out the Vancian rule which is core to D&D spells.

    For Clerics, I do the same thing- allowing them to Know two spells, but instead they can pray "on the fly" for a spell they dont Know ( but is on their spell list) at a cost of two spell slots for the spell desired ( so if a Cleric Knows Cure light Wounds and Bless, but finds he is in a situaion where he needs Purify Food and Drink, he can trade out two of his spell slots for that day- assuming he didnt use them already- and summon the spell on the fly).

    I never liked Spell points for D&D. I dont mind them for other game systems but they just dont seem to fit for D&D. I think the idea of basing spell points on HP is bad since the magic user is already limited in hp as it is. Others use Constitution as a means to guage spell points- but Id dont like this either because then every spell caster is walking around high Constitutions and also gaining extra Hit points, high poison saves, etc. It just doesnt make sense to me. Magic users should be weak due to the constant draw that magic has on their bodies ( at least thats how I rationalize them only having 1-4 hp instead of 1-6 common to average humans- its the constant draw of magic that makes them physically "weaker" )


  11. @Tom: Because of new players that didn't know what spells to memorize, I instituted the following house rule: magic-users and clerics can leave slots empty and study or pray 10 minutes at any time to fill those slots. Thus, you only need to memorize the spells you want to use spontaneously in combat. There's no penalty other than the time delay, however. (And the requirement that magic-users carry their spell books with them, but I think this is standard behaviour for almost all of them these days? I wonder how people lost their spell books in the old days, if at all.)

    1. Yea I like that idea too. What about Clerics?

  12. Hi everyone, wow thanks for all your in-depth comments & feedback! :)

    I'll reply to people's thoughts in blocks, as there are several common themes running...

    Firstly, as a preface, I love the standard D&D Vancian spellcasting system too, so I'm not trying to come up with something "better". I just felt like trying something a bit different for this setting.

    Beastman & Kelvin mentioned the idea of a system based on saving throws with some kind of consequence for failure. As far as I've encountered in my relatively limited exposure to RPG magic systems, that's kind of "the other" method of spellcasting fatigue. (I experienced that kind of system in True20, for example, which works exactly as you described Beastman.) That kind of system is then based on chance and "pushing your luck" in a sense, rather than resource management. It's of course a perfectly workable system, but I decided that somehow it wasn't the feeling I was going for in this setting.

    A third system which is quite common is a pure spell points system of some kind. I find this somehow completely flavourless, however, and ruled that out early on.

    Actually, as I mentioned, the main inspiration for my system was Fighting Fantasy, where each spell cast simply drains STAMINA (= hp). I like the simplicity of it.

    Another theme in several / most people's comments is that this system would make MUs too weak, as they have so few hit points already. Yeah... it's a point I'm not sure about. In the single playtest we've had so far, with 3rd level characters, I didn't feel it was a problem -- MUs just have to be very careful with their magic (which was the intention of the system, to make MUs think about each spell they cast -- is it really worth it?). I was thinking along the lines of a 1st level MU being able to (traditionally) cast a single spell per day, which would equate to a single hit point spent in a day in the new system. Doesn't sound too bad (unless some unlucky player rolls a 1 for hp!, in which case I'd probably let them reroll ;).

    So, yes, upping the MU's hit dice to d6 might be an option. I guess it just depends what kind of balance one desires -- how powerful MUs should be. Certainly the system I've written up above makes MUs significantly weaker than in standard D&D, while giving them a bit more flexibility, and making magic feel a lot more dangerous than in the Vancian system. This was, of course, my intention.

    I've actually been thinking of exactly what Sean proposed though -- a "wound points" pool based on CON / 2 (inspired by reading about Crypts & Things, which I believe has an approach something like that). I think I wouldn't allow this pool to be drained for spellcasting -- is that how you ran it Sean? (Interesting that we have independantly arrived at an almost identical system! :)

    I'll be interested to see how it plays out in our next session, and if the MU's player has any feedback on the system.

    1. Hi Gavin,

      No, I don't allow wound points to be used for spell casting; only hit points may be used to cast spells.

      I played with this system for quite a while and it worked very well; with the addition of wound points, magic users weren't unduly penalized for losing hit points to cast spells, and a first level magic user can still cast a spell even if he rolls only 1 hit point.

      It really is funny how identical 'new' ideas seem to spring up simultaneously, and independently of one another. I had no idea that Crypts & Things was also using a similar system.

    2. Awesome, that sounds perfect!

      We used the "alive until -10" hit points system (like in AD&D) in part 1 of the adventure, so I think I'll just replace that with the CON / 2 wound points system next time and call it a day.

      ps. I only just noticed that one can reply to individual comments now on blogger, ha!

  13. JDJarvis mentioned the fact that a multi-classed F/MU would be a better spell caster than a pure MU. Yes this is something I'd considered. In this one-shot where I'm using this spellcasting system, multi-classed characters aren't allowed. Mixed-class characters are however allowed, and I'd decided that spells cost double the normal amount of hit points for mixed-class MUs.

    Tom mentioned the side-effect of MUs lucky enough to have high CON being the best wizards. Yes, this is a feature I'm not really that keen on either, but I've not managed to think of anything better! (And as simple.) Though, I guess if I were to use the CON/2 wound points (and didn't allow them to be used for spells), then the CON modifier to hit points wouldn't be used, and the problem would be void. Hm, sounds promising actually.

    The other big issue, which I'd only partly considered, was what Sean mentioned about MUs just spamming magic missiles and such. I guess I'd simply make a few select spells higher level. As I mentioned, I'd definitely make sleep 2nd level, probably likewise magic missile and charm person.

    I guess all in all the system I've proposed is tailored fairly specifically to the setting which I'm running, and the atmosphere I'm trying to conjure -- i.e. definitely not designed to be balanced with or equivalent to the standard D&D magic-user.

    1. By "spam magic missile" I assume what is meant is that the MU will spend spell points on low level spells and disregard higher level spells (whereas Vancian "forces" you to memorize some 1st, some 3rd and so on).

      Although this is a problem for all spell point systems (it also allows one to reverse this process and spend all points on a few high level spells when that is more advantageous) I think a partial solution is to not let spells scale up for free by level.

      I.e., higher level wizards' magic missiles fire multiple missiles, and higher level fireballs do more damage etc. If you lock in a magic missile to a single bolt for a 1 hp cost and a fireball to 3d6 damage for 12 hp cost etc. some of this imbalance is ameliorated


    2. Aha, the plot thickens! ;)

      You're right, if spells like magic missile don't scale, then they don't seem so appealing at higher levels.

      The other problem, though, which I think is what Sean was getting at, is that magic missile, as an "auto-damage" spell, is so useful in combat that it overrides the usefulness of any other spell.

      This is always the trouble with D&D house rules isn't it, especially ones affecting the magic system -- there are so many interconnected components that it's very difficult to modify it without unravelling loads of problems... :)

    3. EDIT: should be "a fireball to 5d6 damage for 12 hp cost"

    4. If it were me, I might just eliminate the problematic spells. This is more or less the approach taken in LotFP, where the spell list was very carefully curated to remove the spells with overly crude effects (healing, fireballs, etc).

    5. Yeah, I mean in my own setting I'd probably do exactly that -- make some spells higher in level, adjust some spells, and entirely remove some spells. (And of course add some new spells for fun :)

      That reminds me though that I should check out the LOTFP spell lists again...

    6. Healing is in the LOTFP cleric list, by the way, but raising the dead is out (thank goodness!).

  14. @Alex, ha! that's interesting that M20 uses basically the same system as well! Not sure I'd ever noticed that...

  15. you might like this alternative magic system from the past

  16. Hey Catacomb librarian, thanks for that! An interesting piece of history :)

    I love the quote: "Do you ever get bored with standard Gygax magic?", haha :)

  17. I agree that direct spell point systems are a bit bland, but making different kinds of damage unhealable by magic also seems like it might be awkward to track in play. One might be able to get a similar feel and mathematical result by tracking something like fatigue points next to damage, and have damage + fatigue points >= max HP cause unconsciousness. Or maybe that's how you actually do this in practice?

    1. In the one session we've had with the system so far it didn't seem to be a problem, but what you suggest would be a clever way of handling it, thanks!


Note: only a member of this blog may post a comment.