Magic system (Or, Stargate starts something again...)

[Ra-Tiel]

[...] Would it work?

With the parameters you set forth, unlikely.

Assuming your "mana die" follows the same rules as the normal hit die, your numbers are odd by a bit out of whack. ;)

At first level, you'd get MAX(4d4) plus the casting ability mod (assuming a 16). That'd give 19 mana points (MPs) right at level 1.

At 20th level, you'd be looking at (assuming appropriate and very conservative increases to the casting ability) 16 + 19*4d4 + 20*5 = 16 + 76d4 + 100 = 16 + 190 + 100 = 306 MPs.

More than double of what you wanted.

[...] the problem [...]

Is actually twofold, imho.

First, it suffers from the same problem 3.0 psionics did. Fixed damage. Your fireball (for example) stays at 5d6, while the monsters get tougher and tougher. You're effectively forcing your casters into an arms race, as their spells quickly loose so much punch, that they basically have to rely only on their top-tier spells in combat.

But second, your fixed damage scales not linearily. That means, that it's far more cost efficient to use battlefield control spells to lock down the enemy, and then ping him to death with low level spell spam. Because so far you didn't mention anything about the costs (at least as far as I can tell), let's assume it follows the same guideline as the system from WotC: a spell of level X costs a number of MPs equal to 2*X-1.

If you then analyze the costs per d6, you'll very quickly find out that the higher level spells are extremely inefficient. At spell level 1, 1d6 damage costs 0.33 MPs. At spell level 9, 1d6 damage costs 0.76 MPs.

Therefore, you create a very hard dilemma for casters. Either they are effective, or efficient. But both at the same time is impossible. Well, perhaps there is a sweet spot somewhere in the middle where the cost/damage ratio has an optimal value, but currently I'm too tired (and can't be arsed :P ) to calculate all the numbers.

[Stargate525]

The Problem:

The problem with this model above is that, while it does perform rather well at higher levels, in fact making casting classes rather weaker in raw damage output, at first and second levels the damage potential there is gigantic. 3d6 damage at first level is unheard of, and this is only compounded by the fact that they'll be able to pull this off multiple times. Of course, we could merely chalk that up to going nova, as they won't be able to replenish said spells quickly or at all.

I'm more inclined to let it sit, but I wanted to know what you guys thought.

[Stargate525]

First off, curse you for posting while I am!

With the parameters you set forth, unlikely.

Assuming your "mana die" follows the same rules as the normal hit die, your numbers are odd by a bit out of whack. ;)

At first level, you'd get MAX(4d4) plus the casting ability mod (assuming a 16). That'd give 19 mana points (MPs) right at level 1.

At 20th level, you'd be looking at (assuming appropriate and very conservative increases to the casting ability) 16 + 19*4d4 + 20*5 = 16 + 76d4 + 100 = 16 + 190 + 100 = 306 MPs.

More than double of what you wanted.

bloody hell, you're right. In fact, I was about to go and fix that and put it back at a d12. The numbers should lower by quite a bit. I have no idea what that string of numbers means, otherwise I'd recalculate.

First, it suffers from the same problem 3.0 psionics did. Fixed damage. Your fireball (for example) stays at 5d6, while the monsters get tougher and tougher. You're effectively forcing your casters into an arms race, as their spells quickly loose so much punch, that they basically have to rely only on their top-tier spells in combat.

You're forgetting metamagic, but I agree. Other than using non-fixed damage and the multitude of problems that brings to the table, how is this problem solved?

But second, your fixed damage scales not linearly. That means, that it's far more cost efficient to use battlefield control spells to lock down the enemy, and then ping him to death with low level spell spam. Because so far you didn't mention anything about the costs (at least as far as I can tell), let's assume it follows the same guideline as the system from WotC: a spell of level X costs a number of MPs equal to 2*X-1.

It doesn't. Or didn't. it was equal to the spell level, and 0-level spells were free. That would only exacerbate the problem.

If you then analyze the costs per d6, you'll very quickly find out that the higher level spells are extremely inefficient. At spell level 1, 1d6 damage costs 0.33 MPs. At spell level 9, 1d6 damage costs 0.76 MPs.

Yes, but that's also what happens to the WOTC spells, except it does it within level. A caster has to pay extra mp to get his spells to do more damage, on the order of 1mp per d6 (usually). your efficiency there is messed all to bits.

Therefore, you create a very hard dilemma for casters. Either they are effective, or efficient. But both at the same time is impossible. Well, perhaps there is a sweet spot somewhere in the middle where the cost/damage ratio has an optimal value, but currently I'm too tired (and can't be arsed :P ) to calculate all the numbers.

I understand. But is that dilemma in and of itself a bad thing? A car is usually fast or efficient, not both. Food is either cheap or good tasting. And the old programmers maxim of 'fast, cheap, or good, choose any two.' Where does it say that casters have to have their cake and eat it too?

[Ra-Tiel]

First off, curse you for posting while I am!

:P

bloody hell, you're right. In fact, I was about to go and fix that and put it back at a d12. The numbers should lower by quite a bit. I have no idea what that string of numbers means, otherwise I'd recalculate.

I think it originated when you doubled the "7.5" to "15" and decided that while "d12" was close enough, you wanted a "4d4" progression. ;) When you doubled the first number, you also doubled the final result.

And that "string of numbers" is the listing of averages for a level 20 caster with 4d4 MPs per level and Int 20. ;)

You're forgetting metamagic, but I agree. Other than using non-fixed damage and the multitude of problems that brings to the table, how is this problem solved?

Well, I don't know what other options you include for your casters. Perhaps feats that work based off "School Focus", and increase damage dealt by +1d6 with a basic feat and by +2d6 with an advanced feat? Or certain class abilities that let them "overcharge" spells (increase die step by one category if they take spelllevel d6 non-lethal damage, or by two steps if they take spelllevel d6 lethal damage) or something like that.

It doesn't. Or didn't. it was equal to the spell level, and 0-level spells were free. That would only exacerbate the problem.

Hmmm... let's see. A spell of level X costs X MPs.

This gives us at first level the same cost. 1d6 for 0.33 MPs. The higher level spells however change. A level 9 spell gives us 1d6 for 0.69 MPs. A bit better, but still more than twice as expensive.

Yes, but that's also what happens to the WOTC spells, except it does it within level. A caster has to pay extra mp to get his spells to do more damage, on the order of 1mp per d6 (usually). your efficiency there is messed all to bits.

True. However, when using such a system one-shot direct damage effects indeed are inefficient. Damage over time or summoning effects, however, really shine. I don't know if wall of fire, blade barrier or mage's sword exist in your setting. These would be the spells holding the crown of efficiency, so to speak.

I understand. But is that dilemma in and of itself a bad thing? A car is usually fast or efficient, not both. Food is either cheap or good tasting. And the old programmers maxim of 'fast, cheap, or good, choose any two.' Where does it say that casters have to have their cake and eat it too?

No, I was just pointing it out. This dilemma could have only become apparent in actual game and not before. I'm not saying it's inherently a bad thing. I'm basically for anything that brings casters based on 3.5 in line a bit. However, you'd have to take a careful look at the numbers to prevent casters being less effective than desired.

[Stargate525]

:P

:-/

I think it originated when you doubled the "7.5" to "15" and decided that while "d12" was close enough, you wanted a "4d4" progression. ;) When you doubled the first number, you also doubled the final result.

well you can't exactly have a 1d16, now can you? :P The reason I doubled it is that that then makes the original number the average of a 1dX, where X is the doubled number.

And that "string of numbers" is the listing of averages for a level 20 caster with 4d4 MPs per level and Int 20. ;)

ah.  :weirdo:

So under 1d12, the number becomes the following;
12 + 19d12 + 20*5
12 + 66 + 100
178

That's more like it, but still too high.

1d10

10+19d10+20*5
10+55+100
165

Much closer.

1d8
8+19d8+20*5
8+44+100
152

That's as close as we'll get. This is kinda neat. Now, theoretically, one could simply assign a varying die type for magic to emulate how focused one is. Rangers and Bards get a d8, while the others get a d10 or d12.
 

Well, I don't know what other options you include for your casters. Perhaps feats that work based off "School Focus", and increase damage dealt by +1d6 with a basic feat and by +2d6 with an advanced feat? Or certain class abilities that let them "overcharge" spells (increase die step by one category if they take spelllevel d6 non-lethal damage, or by two steps if they take spelllevel d6 lethal damage) or something like that.

I'm building this system with d20 mechanics in mind, but basically from the ground up. I'm open to almost anything. The crystals come to mind here; since they provide bonuses to the casting as well, we might be able to get some damage increase from them, or as a class feature.  

True. However, when using such a system one-shot direct damage effects indeed are inefficient. Damage over time or summoning effects, however, really shine. I don't know if wall of fire, blade barrier or mage's sword exist in your setting. These would be the spells holding the crown of efficiency, so to speak.

Like I said above, ground up. the spells that will be in this are the spells that I'll be putting in this. That said, I'd like to work damage over time and area effects as a function of the damage output stated above, to better allow players and DMs other than me to see why and how I got those numbers.

No, I was just pointing it out. This dilemma could have only become apparent in actual game and not before. I'm not saying it's inherently a bad thing. I'm basically for anything that brings casters based on 3.5 in line a bit. However, you'd have to take a careful look at the numbers to prevent casters being less effective than desired.

Which is why I'm bouncing these ideas off of you guys. What do you think so far? Is it doing too much to hamper spellcasters, are we on track, or are we not doing enough?

[Ra-Tiel]

:-/

:D

well you can't exactly have a 1d16, now can you? :P The reason I doubled it is that that then makes the original number the average of a 1dX, where X is the doubled number.

True. But the average of 4d4 is not 7.0. It's 4*2.5 = 10. That's one reason why your numbers are higher than anticipated. The other reason is that you underestimated the impact of the casting ability modifier.

I've been toying around with a similar "mana die" idea several years ago (paladin/ranger: d4; bard: d6; druid: d8; cleric/wizard: d10; sorcerer: d12 - or something alone these lines). What I found out was that for a dedicated caster, the mana die only represents about 1/2 to 1/3 of the available mana points. The rest comes from the ability score associated with the character's casting.

And an Int of 20 @ ECL 20 is extremely conservative and completely ignores the existance of things like inherent bonuses (various books, wishes) and enhancement bonuses (headband of intellect, anyone?). If you include those, you'll get an ability score closer to 30, which would mean +200 MPs alone from the ability score @ ECL 20.

ah.  :weirdo:

So under 1d12, the number becomes the following;
12 + 19d12 + 20*5
12 + 66 + 100
178

That's more like it, but still too high.

1d10

10+19d10+20*5
10+55+100
165

Much closer.

1d8
8+19d8+20*5
8+44+100
152

That's as close as we'll get. This is kinda neat. Now, theoretically, one could simply assign a varying die type for magic to emulate how focused one is. Rangers and Bards get a d8, while the others get a d10 or d12.

Naaahhhh, not really. ;)

The average of d12 is 6.5, so your first example would have 12 + 19*6.5 + 20*5 = 235.5 MPs.
The average of d10 is 5.5, so your second example would have 10 + 19*5.5 + 20*5 = 214.5 MPs.
The average of d8 is 4.5, so your third example would have 8 + 19*4.5 + 20*5 = 193.5 MPs
 

I'm building this system with d20 mechanics in mind, but basically from the ground up. I'm open to almost anything. The crystals come to mind here; since they provide bonuses to the casting as well, we might be able to get some damage increase from them, or as a class feature.  

You could make something similar to the warlock's scepter from CArc. Crystals that are specific to each school of magic, come on three strengths, have limited charges/per day uses, and add a fixed value to damage of a spell from that school.

Least augmentation crystal: adds 1d6 damage.
Lesser augmentation crystal: adds 2d6 damage.
Greater augmentation crystal: adds 4d6 damage.

Each crystal has 50 charges/can be used 3 times per day/inflicts 2 points of Cha penalty per use on the user for 24 hours/whatyouwant.

A crystal attuned to evocation cannot be used to augment a necromancy spell, etc.

Like I said above, ground up. the spells that will be in this are the spells that I'll be putting in this. That said, I'd like to work damage over time and area effects as a function of the damage output stated above, to better allow players and DMs other than me to see why and how I got those numbers.

Well, you could always make DoT effects require a "sustain" or "upkeep" cost. Either the caster is forced to spend a full round action concentrating on the spell, or the spell drains its initial costs for as long as the caster does not dismiss it or its duration runs out.

The first option would prevent casters from spamming low-level DoTs on a target to get a high damage per round ratio, while not wasting their reserves.

The second option would allow casters to repeatedly cast DoTs on a target, but they will bleed their MPs like no tomorrow. Additionally, I would highly recommend adding a "(D)" tag to each damaging spell that has a duration other than "instantaneous" or "permanent", to allow casters to dismiss a spell they don't need any longer.

Which is why I'm bouncing these ideas off of you guys. What do you think so far? Is it doing too much to hamper spellcasters, are we on track, or are we not doing enough?

Well, to be honest, I'm anxiously waiting for 4E, so I'm not really putting any more effort into my own 3.5 stuff. I'm still offering suggestions to others, however, which I hope help them a bit.

Therefore I can't say I'm in a position to actually judge your system so far. Especially not without knowing what spells are in and which are out (yes Gate, Polymorph, Black Tentacles, I'm looking at you! :-/ ).

;)

[Stargate525]

:D

:poke:

True. But the average of 4d4 is not 7.0. It's 4*2.5 = 10. That's one reason why your numbers are higher than anticipated. The other reason is that you underestimated the impact of the casting ability modifier.

right. I can't seem to get anything math related correct...

Naaahhhh, not really. ;)

The average of d12 is 6.5, so your first example would have 12 + 19*6.5 + 20*5 = 235.5 MPs.
The average of d10 is 5.5, so your second example would have 10 + 19*5.5 + 20*5 = 214.5 MPs.
The average of d8 is 4.5, so your third example would have 8 + 19*4.5 + 20*5 = 193.5 MPs

...And this is an example of it. they're still almost twice as high as I want them. The easy way out would simply be to double the amount of spell points it takes to cast a particular level.

You could make something similar to the warlock's scepter from CArc. Crystals that are specific to each school of magic, come on three strengths, have limited charges/per day uses, and add a fixed value to damage of a spell from that school.

Least augmentation crystal: adds 1d6 damage.
Lesser augmentation crystal: adds 2d6 damage.
Greater augmentation crystal: adds 4d6 damage.

Each crystal has 50 charges/can be used 3 times per day/inflicts 2 points of Cha penalty per use on the user for 24 hours/whatyouwant.

A crystal attuned to evocation cannot be used to augment a necromancy spell, etc.

Interesting concept. Lemme think.

Well, you could always make DoT effects require a "sustain" or "upkeep" cost. Either the caster is forced to spend a full round action concentrating on the spell, or the spell drains its initial costs for as long as the caster does not dismiss it or its duration runs out.

The first option would prevent casters from spamming low-level DoTs on a target to get a high damage per round ratio, while not wasting their reserves.

The second option would allow casters to repeatedly cast DoTs on a target, but they will bleed their MPs like no tomorrow. Additionally, I would highly recommend adding a "(D)" tag to each damaging spell that has a duration other than "instantaneous" or "permanent", to allow casters to dismiss a spell they don't need any longer.

This I like. I like it alot.

Well, to be honest, I'm anxiously waiting for 4E, so I'm not really putting any more effort into my own 3.5 stuff. I'm still offering suggestions to others, however, which I hope help them a bit.

Don't hold your breath.

Therefore I can't say I'm in a position to actually judge your system so far. Especially not without knowing what spells are in and which are out (yes Gate, Polymorph, Black Tentacles, I'm looking at you! :-/ ).

first impressions?

[Stargate525]

This seems somewhat stupid. Why am I shoehorning myself into spell levels with a points system? What is the inherent bonus to it? Why can't every spell have its own unique cost based on what it does, rather than it being standardized and forcing some to become better or more efficient than the rest? Is there any inherent benefit to a spell level system?

Have I been going about this all wrong?

[LordVreeg]

I dropped spell levels decades ago.

Spell levels were created so that Vancian systems could do 'X' amount of that level spell per day.  
In point systems, it is unneeded.

As I recently said in Snakefing's thread, the cost/scoring system is intrinsically tied to the differentiation system.  IN Vancians, that means, " I can cast this many levels of this type spell".

In mana systems, spell levels become less useful, unless you need them for 'scoring/cost' purposes.

[snakefing]

Well, there are two things that a spell level can do for you:

1) How many spells of this type can you cast? This is handled directly by a spell point system, so it isn't relevant, unless you are using spell level as a proxy for spell point cost.

2) How good do you have to be in order to learn this spell? In the D&D system, this is controlled by the first level where you gain a spell slot for that spell level. In a system that doesn't have spell slots, you'd need some other mechanism to control this. For example, you could rate each spell with a minimum spellcaster level to use it as well as a mana point cost. Or have a spell failure chance that is higher for difficult spells but goes down with the spellcaster level.

[Stargate525]

The only one of those two that apply is the second, and I'm not certain that that's a bad thing in an of itself. Say we allow a caster to choose any spell he wants, even ones he can't yet cast. Since the number you get to pick originally is limited, choosing one that's too points-heavy or uncastable is a poor choice, as it will quickly drain your reserves. Sure the spell might be an encounter-breaker, but it's an encounter-breaker once, and completely negates the caster for the remainder of the day (several, if he dipped beyond what 1 day's worth of mp gain can get him).

[snakefing]

True enough. The only downside is that relying on encounter frequency as a balancing mechanism may not work in certain types of campaigns. (That is, if you are running a campaign in which serious encounters only happen ever few days, the caster with one killer spell per day may not be hampered much.)

Or, consider how different your BBEG encounter will be depending on whether your caster still has her one killer spell per day, or not. If you scale your BBEG on the assumption she'll have her killer spell, but she doesn't, TPK. If you scale the BBEG on the assumption she doesn't have it, but she does, total anti-climax.

Not saying it can't work, but you might find out one of your players will try it, and the DM ends up constantly having to work around it.

[Ra-Tiel]

[...] 1) How many spells of this type can you cast? This is handled directly by a spell point system, so it isn't relevant, unless you are using spell level as a proxy for spell point cost.

To a certain degree that is true. However, in a point based system you face a different problem.

Let's take a look at the power curve of casters (manifesters) in a slot-based system (like vancian casting) and in a point-based system (like psionics).

First, the slot-based system:

Each and every spell the caster releases depletes his resources and drains him of some of his daily power. The process of losing power is gradually, and the caster is actually aware of it.

Second, the point-based system:

The caster remains at his full power potential until he finally runs out of enough points to cast his most powerful spell. The process of losing power is almost instant, and often surprising to the caster.

Additionally, the point-based system requires reworking of almost every spell in the d20 system. Spells like Wail of the Banshee, Time Stop, or Gate were just not designed with the possibility in mind to be cast 20 times per day - or even more often *shudder*. Just compare the high-end spells to the high-end psionic powers and how many spells are significantly stronger, cheaper, or lack any psionic counterpart at all.

2) How good do you have to be in order to learn this spell? In the D&D system, this is controlled by the first level where you gain a spell slot for that spell level.

And therefore it was a direct function of class level in a spellcasting class.

In a system that doesn't have spell slots, you'd need some other mechanism to control this.

Just take the same progression - for example, wizard's spells per day vs psion's powers known. Both follow the same paradigm of "you gain access to spell/power level X at class level 2*X-1".

For example, you could rate each spell with a minimum spellcaster level to use it as well as a mana point cost.

Which would most likely default back to class level.

Or have a spell failure chance that is higher for difficult spells but goes down with the spellcaster level.

No, please don't do that. You remember the BG2 add-on "Throne of Bhaal"? You remember the new mage specialist class introduced then, the chaos mage? You remember how horribly broken it was by allowing to cast level 9 spells at character level 6 but with the "annoyance" of a wild surge (chaos magic thingy) occuring and sometimes lightly screwing up? For all that is holy and just, please don't do such a thing with this system.

The only one of those two that apply is the second, and I'm not certain that that's a bad thing in an of itself. Say we allow a caster to choose any spell he wants, even ones he can't yet cast. Since the number you get to pick originally is limited, choosing one that's too points-heavy or uncastable is a poor choice, as it will quickly drain your reserves. Sure the spell might be an encounter-breaker, but it's an encounter-breaker once, and completely negates the caster for the remainder of the day (several, if he dipped beyond what 1 day's worth of mp gain can get him).

Which is why I personally would not follow the design pattern of "one huge point reserve for the whole day" but rather the pattern "one small point reserve for each encounter". This quite elegantly prevents the caster of blowing his whole daily reserve in a single encounter, aka "going nova".

If you make all casters have a reserve (per encounter) of 1 point per class level + 1 point per casting ability mod, this should work out quite nicely. Taking ability modifiers (including appropriate stat booster) within a reasonably range into account would give us the following ranges of points a caster could utilize per encounter:
- Level 01: 01 (Int 11) - 06 (Int 20)
- Level 05: 07 (Int 14) - 11 (Int 23)
- Level 10: 13 (Int 17) - 18 (Int 26)
- Level 15: 20 (Int 20) - 24 (Int 29)
- Level 20: 28 (Int 27) - 33 (Int 36)
(Yeah, I am to lazy to make a table out of that :P )

Or even make it only +1 point per 2 class levels if that numbers seem too high.

[Stargate525]

...Which is why I personally would not follow the design pattern of "one huge point reserve for the whole day" but rather the pattern "one small point reserve for each encounter". This quite elegantly prevents the caster of blowing his whole daily reserve in a single encounter, aka "going nova".

But his daily reserve is in fact a slight fraction of the power he actually has available. I'm beginning to realize more and more that I want the option of going nova to still be there; it's a viable and unique option available under this system. However, if your spellpoints regain slowly, instead of the entire bank renewing every day, going nova will be amazingly devastating, but leave the caster crippled, potentially for weeks on end.

If you make all casters have a reserve (per encounter) of 1 point per class level + 1 point per casting ability mod, this should work out quite nicely. Taking ability modifiers (including appropriate stat booster) within a reasonably range into account would give us the following ranges of points a caster could utilize per encounter:
- Level 01: 01 (Int 11) - 06 (Int 20)
- Level 05: 07 (Int 14) - 11 (Int 23)
- Level 10: 13 (Int 17) - 18 (Int 26)
- Level 15: 20 (Int 20) - 24 (Int 29)
- Level 20: 28 (Int 27) - 33 (Int 36)
(Yeah, I am to lazy to make a table out of that :P )

Or even make it only +1 point per 2 class levels if that numbers seem too high.

And under this system, how does one account for non-encounter magic?

[Ra-Tiel]

But his daily reserve is in fact a slight fraction of the power he actually has available.

Wha? If a caster's daily reserve is only a "slight fraction" of his total power, what in all Nine Hells is his full power?  :huh:  "Mana points per week", or something? :P

I'm beginning to realize more and more that I want the option of going nova to still be there; it's a viable and unique option available under this system.

Going nova is imho neither viable nor unique.

It's not viable because it heavily involves metagaming (How do the casters know they won't face a second, equally difficult encounter before rest?) or abuse of certain spells (most often spells like teleport to move to safety, or rope trick to create a save resting place on the fly), and disrupts the flow of the game. "Nova, rest, nova, rest, nova, ...".

It's not unique because it's been done to death with vancian casting and with psionics. Actually, psionic characters are just easier to design for going nova, while vancian casters (especially wizards) are still the undisputed masters of going nova (Time stop, Quicken Spell, greater metamagic rod of quicken, Sudden Quicken Spell, polymorph into a creature that gives you a second standard action per round, and so on and so forth).

However, if your spellpoints regain slowly, instead of the entire bank renewing every day, going nova will be amazingly devastating, but leave the caster crippled, potentially for weeks on end.

And you think that this is better in exactly what way? I don't know if you've played the older The Dark Eye computer games, but they used exactly this way. Once your casters were drained in battle, you had to rest several days (up to two weeks) to completely refresh their mana pool (*). Considering that at least the second game in the series had a limited ingame time (you had to finish the main quest within 2 ingame years or you automatically lost), you can imagine that this was not quite as interesting as it sounded.

(*) Or use up some of the rare potions and herbs that restored mana points, and which you usually tried to preserve for combats.

Under a slot-based system, I could imagine a system like this resembling a mechanic you want:
- You don't regain all your spell-slots after a full rest.
- Instead, you regain a spell slot after resting for a number of hours equal to the slot's level.
- You regain spell slots of different levels simultaneously (e.g. if you rest for 4 hours, you regain 4 level 1 slots, 2 level 2 slots, 1 level 3 slot, and 1 level 4 slot).
This requires some extended resting, without leaving the casters grounded for weeks at a time. Also, you can easily determine how many slots are regained by counting the full hours of uninterrupted rest.

However, you're not using spell-slots so this is out for you.

And under this system, how does one account for non-encounter magic?

You could follow a similar mechanic as they did with 4E and encounter powers. Encounter powers are (afaik) usable out of combat after taking a "short rest" (aka 5 minutes break).

In a system with a point reserve based on encounters, one could always assume that out of combat you regain 1 point per minute or something. For smaller stuff it's usually not worth the bookkeeping, but if you keep casting higher-level spells repeatedly, the costs can quickly sum up and leave you dry.