TL;DR version: I propose a system with spell "stacks" composed of spell slots. Casting a spell of a given level expends both that slot and all slots of lower level in that stack. Each extended rest recovers a small number of slots, and when applied to a stack they are always added to the highest level slot available. Since the player will usually choose to recover slots in the most powerful stacks first, and must start with the most powerful slots in a stack, they can be very effective even if they were previously very low on resources, but it remains to their long-term advantage to use spells somewhat conservatively.
---------
I was thinking about different ways to present spellcasters, and particularly if there were a way to create a little more flexibility in how casting resources are expended, saved, and recovered in order to support a wider variety of playstyles within a single subsystem.
We already know that, at least to some degree, 5e spells are going to have improved versions that simply increase their level. Just how pervasive that idea will be taken is unclear, but if most spells have a version for most levels above their base levels we can try some new ideas.
Suppose that an ith level spell "stack" allowed one to memorize a spell and cast it at any level k between 1 and i (instead of just level i as in traditional D&D-style Vancian casting), where these i different potential castings are "slots", since that seems closest to the established usage. The main thing is that casting the level k version would expend all of slots of k and below in that stack.
For example, suppose one memorized magic missile in a 5th level spell stack. (Presumably magic missile has a version of itself for all of levels 1-5.) If the caster cast magic missile at 3rd level she would retain the 4th and 5th level spell slots, but would lose the levels 1, 2, and 3 spell slots (despite not having actually cast it at levels 1 or 2.) This introduces a strong incentive on the character to use the least powerful version of any spell in order to preserve spell-casting resources, without completely removing the ability to go nova. A caster using this system is also unique in that placing different spells in different spell slots could actually have a large impact on daily flexibility. For example, if a high-level caster used a 9th level spell slot for Dimension Door, she might get as many as 9 castings of a Dimension Door-like spell or just a single powerful one, but usually somewhere in-between. If that same caster instead put some attack spell in that slot and used a 4th level slot for Dimension Door we observe the player prioritizing flexibility in attack over movement, but in a way distinct from traditional Vancian systems.
How would spell recovery look? I think that one would recover a fixed number of spell slots per day, but generally fewer than the total number of slots in all stacks. These slots would always be added from the top. So if a 7th level spell stack only had its 7th level spot remaining, recharging a single slot would make the 6th level slot usable again. (My initial preference for number of recharged slots is 1 slot per stack, applied among all stacks however desired, so that if a character cast all their spells at the highest level, thus being totally depleted of spell slots, they could recover at least one use of all their stacks the following day.) As the highest-level stacks are the most powerful usually a player would choose to fill them first, so a character that had gone nova might gain a significant portion of its power back, but would require a more extended period to recover absolutely everything. In times after significant depletion of spell slots it also makes the player balance the benefit of having several potential uses of more powerful spells vs. having greater breadth by making sure all its stacks still have some slots.
My inspiration for this style of recovering spells is actually the charging of capacitors (and similar physical systems), which will charge to a significant portion of their total capacity over a relatively short period of time, but rather slowly approach their maximum by comparison. Obviously one could do the same thing in a normal Vancian system, but in combination with spell stacks which incentivize the player to not "waste" spell slots, I think there is room for some interesting emergent behavior.
Before I do any further analysis I want to suggest possible flavor for such a bizarre system. After all, my usual intuition of how magic "should work" would run toward a spell point system (e.g. casting the 3rd level magic missile in a 5th level stack might expend 3 points, leaving 2 points left for a later casting) or else toward getting rid of the higher level versions of the spell because casting, and therefore "holding on to", the lower level versions should be easier. Instead, I envision preparing a spell in this system as creating a reservoir and dam which holds back and temporarily tames magical potential energy. This energy absolutely wants to escape, and can be represented by objects of different sizes, where "small magic" like low-level spells might be small spheres, while "strong magic" like 9th-level spells are very large spheres. To cast a spell one must create an opening in the dam large enough to release spheres of the corresponding size. In doing so, however, all the smaller spheres simply leak out along with it. In this analogy, therefore, increasing experience enables the caster to create reservoirs (which is hard) large enough for the large spheres, but these largest spheres are actually the easiest to control because they are only lost if one completely demolishes the mental barriers created by spell preparation which are holding them within.
All right, here is a sample spell progression that I could see working in the game.
The columns to the right have the following meaning:
Some interesting things about this progression (which would likely hold for other similar-looking ones) is that, for better or worse, low-level play is essentially as it always has been in D&D. Maybe a few more spell slots, but when you have 6 level 1 spell slots and get 6 slots to fill after a rest the resource management is unchanged from classic Vancian design. One can see that as level increases, the absolute maximum fraction of points one can recover after a rest decreases, so that managing spell slots has a different feel at high levels. While I'm not in love with this attribute, it certainly does mean a character would have a chance to ease into the resource management.
One thing that changes with level is the fraction of power that can be recovered in a single night. At 5th level a completely-expended caster could recover about 75% of its total power (assuming points as defined earlier are an adequate measure) after a rest. At 11th level that fraction is closer to 50%, and at 20th 33%.
Clearly the total time to recover all spell slots after completely expending them increases with level. It requires 2 days starting at 3rd level, 3 days starting at 7th, 4 days starting at 11th, 5 days starting at 15th, 5 days at 13th, and 6 days at 16th. Essentially any spell progression using this system will have such a character.
Another thing that changes is how much power is wasted by not casting efficiently. It's an extremely rough measure, but we can see that by 8th level a caster who casts all his spells in the least efficient manner possible will have wasted about 1/2 of his power compared to someone who does it in the absolute most efficient manner. By 20th level the amount wasted would increase to about 3/4. Now, casting in an efficient manner becomes wildly impractical later on (and by how much I couldn't yet guess), but even so it is clear that always casting a spell at maximum power could have major consequences.
The number of distinct spells that can be prepared is fairly moderate, in the same ballpark as 4e. Compared to 3.5 (where a 20th level wizard might easily have 50-60 slots) and BECMI/RC (where a 20th level wizard would be in the neighborhood of 35) this is modest. This caster, of course, will have much greater flexibility in how to use those spells, perhaps even cast more overall during a day, because they can be cast multiple times by showing a little restraint. And although the maximum number of slots (~100!) a character could cast would suggest some wicked book-keeping issues, the fact of the matter is that only two pieces of information are needed for each spell stack: the level of the stack, and the lowest level slot still available.
So what virtues might this system have? I think it gives a lot of freedom to the player to be reckless or to be very careful without needing to grind the game to a halt. The careful player will conserve spells, use the least level of spell necessary, and generally be able to stretch their resources for quite a while, especially over a many-day adventure. A more naturally reckless player might have previously been tempted to blow his spells more quickly since they're all coming back the next day anyway, but because they don't like "wasting" perfectly good spell slots show more restraint than in a standard Vancian system. Where they might have previously blow all their highest level spells, now they might choose to use the versions 2 or 3 levels below first, which significantly dampens the negative effects of going nova within a single encounter. Finally, the character that can't help but go for the biggest spells available will still be able to contribute significantly the next day despite the more limited recovery of spells. I hope that these incentives could diminish the 15-minute work day while leaving in really interesting long term resource management.
I have a fear that at some tables it would simply lead to the 15-minute work week because the player feels compelled to nova and then insists upon resting up to full, despite never using most of the low-level spell slots anyway. I suppose this is simply the risk of using incentives rather than force. On the other hand, is the 15-minute work week really any worse than the 15-minute work day? Narratively it is, of course, but at the table when you handwave time you can handwave as much as you want. And maybe some DMs that merely tolerated an extra overnight rest would put their foot down on this issue if it became necessary. Either way, at low levels this system doesn't really do anything for solving the 15-minute day compared to a normal Vancian system, and it is in this regime where the 15-minute day might be the largest problem. In that respect this system (or at least the specific progression I gave above) may simply have failed to meet my goals.
An aspect I like is that it feels like a character can construct a lot of thematic emphasis and signature spells despite the relatively few spell stacks. A 9th level spell stack is much more important to a character than a single 9th level spell slot would be in the usual Vancian system because it provides so much potential depth to that ability. If one wants to be a fire mage memorize different fire spells in your 3 tallest stacks and you will have a lot of flexibility despite committing relatively few stacks to them. One can also do other cool things with them. For example, if the wizard used this system one might choose to implement school specialization by specifying that one of the highest-level stacks must always be assigned to a spell from the appropriate school, but then using traditional Vancian casting on its slots rather than the proposed system. It wouldn't increase the actual maximum potential with the class, but for that spell stack one would never have to worry about wasting slots.
There are a couple big caveats for using this system in 5e that wouldn't necessarily be present in a different game designed with it in mind.
1) I've basically assumed spells have a version from level 1 to 9, otherwise all those lower-level slots are wasted. As much as I might wish it, I highly doubt 5e will ever introduce a 1st level spell that scales smoothly to meteor swarm. Personally, I think that would be the perfect way to give this caster passive and/or at-will effects. For example, if fireball is still a 3rd level minimum spell perhaps it could grant a +1 damage bonus to all fire spells or a control fire cantrip, or the like. This sort of thing is tricky, though, because it could lead to all sorts of stacking issues, etc. Maybe just a bonus to its own stack?
If all spells were accessed with this system one could simply make sure the base version is good enough to account for the "missing" slots, but that strikes me as singularly awkward, especially for balance purposes.
2) I've also tacitly assumed that there is a version of every spell at every level above its base, which is by no means guaranteed in 5e. That's less an issue than number 1, in my opinion -- if occasionally one gets a 6th level effect for a 7th level slot I wouldn't cry. If lots of spells only have huge jumps (like from 1 to 4 to 9) then it might be an important issue.
3) Players with excellent resource management skills could completely overshadow other players in the same class, heck even a player with an identical character. A savvy 20th level caster might manage to cast 17 7th, 8th, and 9th level spells. An unsavvy one might only be able to cast 8, but at the end both might have the same level of depletion in the corresponding stacks. Granted, the savvy player was taking a risk that he would need that higher level spell later instead of immediately, but when that gambit pays off it really pays off. This is probably my least favorite aspect of the system, and I usually have a pretty high tolerance for systems with a lot of room for system mastery. For those who hate system mastery... well it's probably a deadly flaw. I'm not sure it can be easily mitigated without changing the entire character of the system.
I hope those of you who read this wall of text enjoyed it! I welcome comments on the mechanical aspects and the fluff justification. I don't see 5e straying this far from standard Vancian casting, but I sure had an interesting time thinking about it, so thanks for reading.
---------
I was thinking about different ways to present spellcasters, and particularly if there were a way to create a little more flexibility in how casting resources are expended, saved, and recovered in order to support a wider variety of playstyles within a single subsystem.
We already know that, at least to some degree, 5e spells are going to have improved versions that simply increase their level. Just how pervasive that idea will be taken is unclear, but if most spells have a version for most levels above their base levels we can try some new ideas.
Suppose that an ith level spell "stack" allowed one to memorize a spell and cast it at any level k between 1 and i (instead of just level i as in traditional D&D-style Vancian casting), where these i different potential castings are "slots", since that seems closest to the established usage. The main thing is that casting the level k version would expend all of slots of k and below in that stack.
For example, suppose one memorized magic missile in a 5th level spell stack. (Presumably magic missile has a version of itself for all of levels 1-5.) If the caster cast magic missile at 3rd level she would retain the 4th and 5th level spell slots, but would lose the levels 1, 2, and 3 spell slots (despite not having actually cast it at levels 1 or 2.) This introduces a strong incentive on the character to use the least powerful version of any spell in order to preserve spell-casting resources, without completely removing the ability to go nova. A caster using this system is also unique in that placing different spells in different spell slots could actually have a large impact on daily flexibility. For example, if a high-level caster used a 9th level spell slot for Dimension Door, she might get as many as 9 castings of a Dimension Door-like spell or just a single powerful one, but usually somewhere in-between. If that same caster instead put some attack spell in that slot and used a 4th level slot for Dimension Door we observe the player prioritizing flexibility in attack over movement, but in a way distinct from traditional Vancian systems.
How would spell recovery look? I think that one would recover a fixed number of spell slots per day, but generally fewer than the total number of slots in all stacks. These slots would always be added from the top. So if a 7th level spell stack only had its 7th level spot remaining, recharging a single slot would make the 6th level slot usable again. (My initial preference for number of recharged slots is 1 slot per stack, applied among all stacks however desired, so that if a character cast all their spells at the highest level, thus being totally depleted of spell slots, they could recover at least one use of all their stacks the following day.) As the highest-level stacks are the most powerful usually a player would choose to fill them first, so a character that had gone nova might gain a significant portion of its power back, but would require a more extended period to recover absolutely everything. In times after significant depletion of spell slots it also makes the player balance the benefit of having several potential uses of more powerful spells vs. having greater breadth by making sure all its stacks still have some slots.
My inspiration for this style of recovering spells is actually the charging of capacitors (and similar physical systems), which will charge to a significant portion of their total capacity over a relatively short period of time, but rather slowly approach their maximum by comparison. Obviously one could do the same thing in a normal Vancian system, but in combination with spell stacks which incentivize the player to not "waste" spell slots, I think there is room for some interesting emergent behavior.
Before I do any further analysis I want to suggest possible flavor for such a bizarre system. After all, my usual intuition of how magic "should work" would run toward a spell point system (e.g. casting the 3rd level magic missile in a 5th level stack might expend 3 points, leaving 2 points left for a later casting) or else toward getting rid of the higher level versions of the spell because casting, and therefore "holding on to", the lower level versions should be easier. Instead, I envision preparing a spell in this system as creating a reservoir and dam which holds back and temporarily tames magical potential energy. This energy absolutely wants to escape, and can be represented by objects of different sizes, where "small magic" like low-level spells might be small spheres, while "strong magic" like 9th-level spells are very large spheres. To cast a spell one must create an opening in the dam large enough to release spheres of the corresponding size. In doing so, however, all the smaller spheres simply leak out along with it. In this analogy, therefore, increasing experience enables the caster to create reservoirs (which is hard) large enough for the large spheres, but these largest spheres are actually the easiest to control because they are only lost if one completely demolishes the mental barriers created by spell preparation which are holding them within.
All right, here is a sample spell progression that I could see working in the game.
Code:
Caster ------ Stack Level ------ Max Max rec
Level 1 2 3 4 5 6 7 8 9 Stacks Slots Points Points
1 5 5 5 5 5
2 6 6 6 6 6
3 5 1 6 7 8 7
4 5 2 7 9 11 9
5 4 2 1 7 11 16 12
6 4 2 2 8 14 22 16
7 3 2 2 1 8 17 31 21
8 3 2 2 2 9 21 41 26
9 2 2 2 2 1 9 25 55 32
10 2 2 2 2 2 10 30 70 38
11 1 2 2 2 2 1 10 35 90 47
12 1 2 2 2 2 2 11 41 111 52
13 1 1 2 2 2 2 1 11 46 136 58
14 1 1 2 2 2 2 2 12 53 164 68
15 1 1 1 2 2 2 2 1 12 58 194 74
16 1 1 1 2 2 2 2 2 13 66 230 85
17 1 1 1 1 2 2 2 2 1 13 71 265 92
18 1 1 1 1 2 2 2 2 2 14 80 310 104
19 1 1 1 1 2 2 2 2 3 15 89 355 116
20 1 1 1 2 2 2 2 3 3 17 101 401 131
- Stacks. The total number of stacks (i.e. different spells) that can be learned. Since a totally rested caster could go nova by casting the highest level spell slot from each stack without casting any other spells, this is also the minimum number of spells slots that could be used before running out of spells.
- Slots. The total number of spell slots that could be cast. A caster that conserved spells to the maximum degree (i.e. always using the lowest level slot in each stack first) could cast this many times before being entirely depleted.
- Max points. To estimate the total power available to a caster I've assigned a point value to spells slots, equal to the level of the slot. (This assumes spells scale linearly by level, such that a level 9 spell is approximately 9 times as powerful as a level 1 spell. Hopefully this won't be a terrible assumption in 5e.) This column reports the maximum possible points of a fully rested caster who chooses to cast in the most efficient possible way. Except at the very lowest levels one would not expect to observe this in play, but it does measure maximum potential.
- Max rec points. This is the maximum possible points recovered in a single day or extended rest, assuming one recovers a number of slots equal to the number of stacks. Note that to achieve this value one must have essentially expended all high-level slots, and also choose not to recover any low-level ones (sacrificing breadth for more potential power). In other words, typically one would recover fewer points worth of power than indicated by this value.
- Also note that if a caster chooses to go nova and casts all their spells without expending any lower-level slots first, then the slots column also calculates the points expended by that effort. Since this is the least efficient way to cast spells, the slots column could also be considered a "minimum points" column.
Some interesting things about this progression (which would likely hold for other similar-looking ones) is that, for better or worse, low-level play is essentially as it always has been in D&D. Maybe a few more spell slots, but when you have 6 level 1 spell slots and get 6 slots to fill after a rest the resource management is unchanged from classic Vancian design. One can see that as level increases, the absolute maximum fraction of points one can recover after a rest decreases, so that managing spell slots has a different feel at high levels. While I'm not in love with this attribute, it certainly does mean a character would have a chance to ease into the resource management.
One thing that changes with level is the fraction of power that can be recovered in a single night. At 5th level a completely-expended caster could recover about 75% of its total power (assuming points as defined earlier are an adequate measure) after a rest. At 11th level that fraction is closer to 50%, and at 20th 33%.
Clearly the total time to recover all spell slots after completely expending them increases with level. It requires 2 days starting at 3rd level, 3 days starting at 7th, 4 days starting at 11th, 5 days starting at 15th, 5 days at 13th, and 6 days at 16th. Essentially any spell progression using this system will have such a character.
Another thing that changes is how much power is wasted by not casting efficiently. It's an extremely rough measure, but we can see that by 8th level a caster who casts all his spells in the least efficient manner possible will have wasted about 1/2 of his power compared to someone who does it in the absolute most efficient manner. By 20th level the amount wasted would increase to about 3/4. Now, casting in an efficient manner becomes wildly impractical later on (and by how much I couldn't yet guess), but even so it is clear that always casting a spell at maximum power could have major consequences.
The number of distinct spells that can be prepared is fairly moderate, in the same ballpark as 4e. Compared to 3.5 (where a 20th level wizard might easily have 50-60 slots) and BECMI/RC (where a 20th level wizard would be in the neighborhood of 35) this is modest. This caster, of course, will have much greater flexibility in how to use those spells, perhaps even cast more overall during a day, because they can be cast multiple times by showing a little restraint. And although the maximum number of slots (~100!) a character could cast would suggest some wicked book-keeping issues, the fact of the matter is that only two pieces of information are needed for each spell stack: the level of the stack, and the lowest level slot still available.
So what virtues might this system have? I think it gives a lot of freedom to the player to be reckless or to be very careful without needing to grind the game to a halt. The careful player will conserve spells, use the least level of spell necessary, and generally be able to stretch their resources for quite a while, especially over a many-day adventure. A more naturally reckless player might have previously been tempted to blow his spells more quickly since they're all coming back the next day anyway, but because they don't like "wasting" perfectly good spell slots show more restraint than in a standard Vancian system. Where they might have previously blow all their highest level spells, now they might choose to use the versions 2 or 3 levels below first, which significantly dampens the negative effects of going nova within a single encounter. Finally, the character that can't help but go for the biggest spells available will still be able to contribute significantly the next day despite the more limited recovery of spells. I hope that these incentives could diminish the 15-minute work day while leaving in really interesting long term resource management.
I have a fear that at some tables it would simply lead to the 15-minute work week because the player feels compelled to nova and then insists upon resting up to full, despite never using most of the low-level spell slots anyway. I suppose this is simply the risk of using incentives rather than force. On the other hand, is the 15-minute work week really any worse than the 15-minute work day? Narratively it is, of course, but at the table when you handwave time you can handwave as much as you want. And maybe some DMs that merely tolerated an extra overnight rest would put their foot down on this issue if it became necessary. Either way, at low levels this system doesn't really do anything for solving the 15-minute day compared to a normal Vancian system, and it is in this regime where the 15-minute day might be the largest problem. In that respect this system (or at least the specific progression I gave above) may simply have failed to meet my goals.
An aspect I like is that it feels like a character can construct a lot of thematic emphasis and signature spells despite the relatively few spell stacks. A 9th level spell stack is much more important to a character than a single 9th level spell slot would be in the usual Vancian system because it provides so much potential depth to that ability. If one wants to be a fire mage memorize different fire spells in your 3 tallest stacks and you will have a lot of flexibility despite committing relatively few stacks to them. One can also do other cool things with them. For example, if the wizard used this system one might choose to implement school specialization by specifying that one of the highest-level stacks must always be assigned to a spell from the appropriate school, but then using traditional Vancian casting on its slots rather than the proposed system. It wouldn't increase the actual maximum potential with the class, but for that spell stack one would never have to worry about wasting slots.
There are a couple big caveats for using this system in 5e that wouldn't necessarily be present in a different game designed with it in mind.
1) I've basically assumed spells have a version from level 1 to 9, otherwise all those lower-level slots are wasted. As much as I might wish it, I highly doubt 5e will ever introduce a 1st level spell that scales smoothly to meteor swarm. Personally, I think that would be the perfect way to give this caster passive and/or at-will effects. For example, if fireball is still a 3rd level minimum spell perhaps it could grant a +1 damage bonus to all fire spells or a control fire cantrip, or the like. This sort of thing is tricky, though, because it could lead to all sorts of stacking issues, etc. Maybe just a bonus to its own stack?
If all spells were accessed with this system one could simply make sure the base version is good enough to account for the "missing" slots, but that strikes me as singularly awkward, especially for balance purposes.
2) I've also tacitly assumed that there is a version of every spell at every level above its base, which is by no means guaranteed in 5e. That's less an issue than number 1, in my opinion -- if occasionally one gets a 6th level effect for a 7th level slot I wouldn't cry. If lots of spells only have huge jumps (like from 1 to 4 to 9) then it might be an important issue.
3) Players with excellent resource management skills could completely overshadow other players in the same class, heck even a player with an identical character. A savvy 20th level caster might manage to cast 17 7th, 8th, and 9th level spells. An unsavvy one might only be able to cast 8, but at the end both might have the same level of depletion in the corresponding stacks. Granted, the savvy player was taking a risk that he would need that higher level spell later instead of immediately, but when that gambit pays off it really pays off. This is probably my least favorite aspect of the system, and I usually have a pretty high tolerance for systems with a lot of room for system mastery. For those who hate system mastery... well it's probably a deadly flaw. I'm not sure it can be easily mitigated without changing the entire character of the system.
I hope those of you who read this wall of text enjoyed it! I welcome comments on the mechanical aspects and the fluff justification. I don't see 5e straying this far from standard Vancian casting, but I sure had an interesting time thinking about it, so thanks for reading.
Last edited: