Hypothetical: Alternate warlock-like casting *for everyone*

[PhoenixPhyre]

Warlocks get, for most of their career, 2 + 2*SR spell slots of their highest level and no slots above 5th. These refresh an unlimited number of times, contingent on short rests. At higher levels this goes up to 3 + 3*SR and they get Mystic Arcana. This limits their per-encounter "burn rate" and sharply weakens reaction spells, especially low-level ones. In exchange, they get invocations and more powerful cantrips.

"Regular" full-casters have an ever-increasing array of slots of all the levels; a couple classes have ways to partially refresh their slots once per day. This means that they are limited mostly by their action economy, rather than spell slot count, after the first few levels. In "exchange", they have poorer cantrips and fewer (and weaker) ways to do things that don't involve casting leveled spells.

I have a hunch (not more than that) that in 5.5e, the design will become more warlock-like--more emphasis on non-spell-slot abilities, stronger "at will" (whether they're cantrips or not) and fewer overall spell slots. This would, if I'm right, go some way in smoothing out the adventuring day--less incentive to go nova and then force a rest because you're out of stuff. Instead of going from 100% (with slots) down to ~25% (without your big slots), you'd go from 80% (of the present model) when full to 50% when "empty". But I expect them to balance most things around long rests. Which, in my opinion, is backward. Because unless you restrict long rests, you'll still have the present issue, just for everyone.

Instead, what about something like the following:

TL;DR -- you max out at 4 spell slots: 1 of the highest level you can cast (except at level 2) (max 5th), 2 at the next level down, and 1 at the level below that. 6+ would be "Mystic Arcana-like" limited uses per day. Everyone could refresh up to half their slots (minimum 1) twice per day over the course of 10 minutes (decoupled from short rests, which would remain for spending HD and possibly a few other things). Everyone would get stronger at-wills and limited-use (but not spell slot) abilities. Some things could just go straight into the class calculation--mage armor could become an Unarmored Defense-like ability (your AC is 10 + INT + DEX while unarmored/no shield).




Level
1st
2nd
3rd
4th
5th
Special


1
1
-
-
-
-
-


2
1
-
-
-
-
1


3
2
1
-
-
-
1


4
2
1
-
-
-
2


5
1
2
1
-
-
2


6
1
2
1
-
-
3


7
-
1
2
1
-
3


8
-
1
2
1
-
4


9
-
-
1
2
1
4


10
-
-
1
2
1
5


11
-
-
1
2
1
5, 1x6th


12
-
-
1
2
1
5, 1x6th


13
-
-
1
2
1
5, 1x6th, 1x7th


14
-
-
1
2
1
5, 1x6th, 1x7th


15
-
-
1
2
1
5, 1x6th, 1x7th, 1x8th


16
-
-
1
2
1
5, 1x6th, 1x7th, 1x8th


17
-
-
1
2
1
5, 1x6th, 1x7th, 1x8th, 1x9th


18
-
-
1
2
1
6, 1x6th, 1x7th, 1x8th, 1x9th


19
-
-
1
2
1
6, 2x6th, 1x7th, 1x8th, 1x9th


19
-
-
1
2
1
6, 2x6th, 1x7th, 1x8th, 1x9th


Half-casters would behave similarly, except more slowly, and not get the "mystic arcana" equivalents.


The things in the "special" column would be things like Invocations, spell-like abilities, limited non-spell-slot castings of specific spells, etc. Whether fixed by the class or menu-style like Invocations would be up to the class.

You'd still learn more spells after you stop getting spell slots, but the numbers would be adjusted.

The idea would be that you wouldn't have the current geometric growth in capability above low levels, where you both get more resources and each resource gets more powerful. Instead, you'd quickly cap in resources and move to a linear model after that, producing something closer to a logarithmic curve (fast at the start, then slowing down to something roughly constant). Mathematically, probably something closer to log-linear (x ln x).


Yes, this would be a big nerf to full casters, especially in the "short working day" model. That's rather intentional. Because currently, the only real constraint is DMs imposing relatively artificial barriers. And the "right" answer is almost always to burn resources as fast as possible to end encounters fast. And then rest. Which means that those classes that are limited only by action economy (or even less than that[1]) end up distorting the calculus for everyone. Which means that classes that can't burn their resources in an alpha strike rarely get to shine--the nova people go, and then you rest.

The goal would be to limit the alpha-strike capability, while giving everyone some amount of "mid-day" replenishment.

[1] paladins and your "can burn 3 spell slots in two actions" stuff, I'm looking at you :smallbiggrin:.

[Rukelnikov]

I understand the issues you are mentioning, and I can see why you take this path to address them, the Warlock doesn't present these problems, so lets make the other casters more like the warlock. I don't know if this is what they are gonna go for or not in the next edition, but in the current one, I think it'd make casters (especially arcane ones) feel even more samey.

I'd go in the totally opposite direction, not only shouldn't full caster get their spells back in a short rest, they shouldn't even get them in a long rest. They should recover a certain amount of spells slots per rest, but that ammount should be below their maximum, going from "a rest and a half" at lvl 1 to go to full, ending at 10+ days to go from fully depleted to full.

I'd like a greater change to the magic system for the coming edition, like your current ammount of remaining spells affecting your spellcasting in a way, like managing cantrips more like reserve feats from 3.5, where the highest spell level still available to you would determine the potency of your at will spells.

[PhoenixPhyre]

I understand the issues you are mentioning, and I can see why you take this path to address them, the Warlock doesn't present theses problems, so lets make the other casters more like the warlock. I don't know if this is what they are gonna go for or not in the next edition, but in the current one, I think it's make casters feel even more samey.

I'd go in the totally opposite direction, not only shouldn't full caster get their spells back in a short rest, they shouldn't even get them in a long rest. They should recover a certain amount of spells slots per rest, but that ammount should be below their maximum, going from "a rest and a half" at lvl 1 to go to full, ending at 10+ days to go from fully depleted to full.

I'd like a greater change to the magic system for the coming edition, like your current ammount of remaining spells affecting your spellcasting in a way, like managing cantrips more like reserve feats from 3.5, where the highest spell level still available to you would determine the potency of your at will spells.

A slower recovery simply exacerbates the issue. Because now the incentive is to take a week off instead of a day. And fundamentally, in many campaigns, time's "cost" scales sub-linearly. Turns matter, minutes matter less, 10 minutes is less meaningful of a cost than 10x 1 minute segments, and a week is only about twice (roughly) the cost of a day. The goal is to reduce the disparities, so that everyone is getting some resources back continually, while not being able to nova as hard.

And yes, non-casters should have similar resource constraints. None of this "rogues have no resources" and "barbarians only have 2 things that only come back on LR". Everyone should get some benefit from
a) short breaks (thinking around 10 minutes)
b) short rests
c) full rests.

The exact things can vary, but everyone should get something meaningful at each point. And planning around (by default) full recovery every "full rest" (the in-game time of which can vary) makes a DM's life way easier.

[Rukelnikov]

A slower recovery simply exacerbates the issue. Because now the incentive is to take a week off instead of a day. And fundamentally, in many campaigns, time's "cost" scales sub-linearly. Turns matter, minutes matter less, 10 minutes is less meaningful of a cost than 10x 1 minute segments, and a week is only about twice (roughly) the cost of a day. The goal is to reduce the disparities, so that everyone is getting some resources back continually, while not being able to nova as hard.

That may be the case, I'm not sure, then if a week is too few, then it takes months to refill all your spell slots. Off the top of my head I was thinking something like prof slot levels per long rest, at lvl 17, refreshing your 9th level slot takes a long rest and a half, refreshing your 7th, 8th, and 9th slot would take 4 days, but if that's "too fast" just make it slower. This creates the distinction of still being able to go nova, but the cost is proportionally higher.


And yes, non-casters should have similar resource constraints. None of this "rogues have no resources" and "barbarians only have 2 things that only come back on LR". Everyone should get some benefit from
a) short breaks (thinking around 10 minutes)
b) short rests
c) full rests.

The exact things can vary, but everyone should get something meaningful at each point. And planning around (by default) full recovery every "full rest" (the in-game time of which can vary) makes a DM's life way easier.

Well, here I think our disagreement comes from design philosophies, I like the mechanical distinction of having the full spectrum of some classes just not having resources at all like the Rogue, to the other end of the spectrum being full resources, and little "at will" stuff, like Wizard or Sorcerer.

I'd hate it for classes to be a preset object to be filled with specifics. IE, everyone gets a short rest ability at lvl 1, a long rest ability at lvl 3, a second short rest ability at lvl 4, etc.

To me, that's just having 1 class.

EDIT: Taking it a bit further, I do expect more uses for HD in the next edition, and I think it'd be interesting if during the long rest, you could choose to either restore HD or Spells slots, or a combination thereof.

[PhoenixPhyre]

Well, here I think our disagreement comes from design philosophies, I like the mechanical distinction of having the full spectrum of some classes just not having resources at all, like rogue to the other end of the spectrum being full resources, and little "at will" stuff.

I'd hate it for classes to be a preset object to be filled with specifics. IE, everyone gets a short rest ability at lvl 1, a long rest ability at lvl 3, a second short rest ability at lvl 4, etc.

To me, that's just having 1 class.

You don't need to lock it down that far. There absolutely could be variance, but the idea that everyone gets something at each "break" throughout the day is, I think, critical.

------

I'm thinking through something I call "Rogue Equivalent Units" for considering damage. Basically, the damage a rogue who gets sneak attack every turn does, before accuracy (assumed to be equal for everyone) or magic items would be one REU. And classes that have resources should do less than 1 REU without spending resources, and spending resources should bump them up. And thinking about things in terms of actions, not encounters. So a hypothetical class that did 0.5 REU baseline but had 3 daily "offensive" resources would need each daily resource to do (approximately) (1/6*R + 1/2) REU each, where R is the number of rounds expected per day.

The general formula would be (R = rounds, 1/x = at will damage in REU, x > 1, N = resources, Y = damage per resource expenditure): R = (R - N)/x + NY ---> Y = R/N - 1/(N*x) + 1/x. So more resources should mean lower per-resource damage, holding at-will damage constant. And more at-will damage should mean lower number or power of "offensive" resources.

[Rukelnikov]

You don't need to lock it down that far. There absolutely could be variance, but the idea that everyone gets something at each "break" throughout the day is, I think, critical.

------

I'm thinking through something I call "Rogue Equivalent Units" for considering damage. Basically, the damage a rogue who gets sneak attack every turn does, before accuracy (assumed to be equal for everyone) or magic items would be one REU. And classes that have resources should do less than 1 REU without spending resources, and spending resources should bump them up. And thinking about things in terms of actions, not encounters. So a hypothetical class that did 0.5 REU baseline but had 3 daily "offensive" resources would need each daily resource to do (approximately) (1/6*R + 1/2) REU each, where R is the number of rounds expected per day.

The general formula would be (R = rounds, 1/x = at will damage in REU, x > 1, N = resources, Y = damage per resource expenditure): R = (R - N)/x + NY ---> Y = R/N - 1/(N*x) + 1/x. So more resources should mean lower per-resource damage, holding at-will damage constant. And more at-will damage should mean lower number or power of "offensive" resources.

Thinking in terms of actions instead of encounters is the correct thing to do. And having a baseline is also ok. Overall, I like the way you approach this.

However, I do not adhere to the idea that everyone should do roughly the same damage though. A class like barbarian being meaned to 1.25 RDU for example, and a Bard being meaned to 0.75 is not only ok, but desirable IMO.

[PhoenixPhyre]

Thinking in terms of actions instead of encounters is the correct thing to do. And having a baseline is also ok.

I do not adhere to the idea that everyone should do roughly the same damage though. A class like barbarian being meaned to 1.25 RDU for example, and a Bard being meaned to 0.75 is not only ok, but desirable IMO.

The idea is to benchmark. Variance is normal, but you have to decide how much variance is acceptable. And currently, there is no such thing. As it stands, a Champion 2H fighter (no feats and really bad optimization as well) runs well above 1 REU...with basic attacks alone. A SnB fighter is roughly 1 REU steady state.

This analysis also lets you figure out what the "hidden" assumptions of the system are--after doing it for a few cases, it seems that the expectation for paladins is that they'll spend roughly half their slots smiting, mostly their top-level ones (except 5ths). And that a barbarian will have rage roughly half the time. That's all on an assumption that the real baseline for a primary damage dealer is slightly above 1 REU (~1.05 REU or so).

Feats and multiclassing throw all this way out the window.

[Rukelnikov]

The idea is to benchmark. Variance is normal, but you have to decide how much variance is acceptable. And currently, there is no such thing. As it stands, a Champion 2H fighter (no feats and really bad optimization as well) runs well above 1 REU...with basic attacks alone. A SnB fighter is roughly 1 REU steady state.

This analysis also lets you figure out what the "hidden" assumptions of the system are--after doing it for a few cases, it seems that the expectation for paladins is that they'll spend roughly half their slots smiting, mostly their top-level ones (except 5ths). And that a barbarian will have rage roughly half the time. That's all on an assumption that the real baseline for a primary damage dealer is slightly above 1 REU (~1.05 REU or so).

Feats and multiclassing throw all this way out the window.

That's an interesting excercise to do, and some interesting results the paladin and barbarian thing.

Rogue being the baseline makes sense since they are the most steadily increasing ones, but you said you didn't take accuracy into consideration, I'd assume rogues were envisioned as either 2WF or attacking with advantage, so a miss for the rogue would mean a proportionally lower amount of missed damage than a Champion. Paladin is somewhere in the middle, work like a rogue if the smite, and like a champion if they dont.

[PhoenixPhyre]

That's an interesting excercise to do, and some interesting results the paladin and barbarian thing.

Rogue being the baseline makes sense since they are the most steadily increasing ones, but you said you didn't take accuracy into consideration, I'd assume rogues were envisioned as either 2WF or attacking with advantage, so a miss for the rogue would mean a proportionally lower amount of missed damage than a Champion. Paladin is somewhere in the middle, work like a rogue if the smite, and like a champion if they dont.

The actual calculation included accuracy (assuming the same modifiers and same targets) and crits. It also considered TWF vs ranged for rogues, SnB vs GWF (yes, including the fighting styles) for fighters, and, IIRC, only GWF for the paladin. It was...labor intensive, so I didn't do it for most possible things.


------

Doing a napkin-math calculation at 5th level for a wizard:
Assumptions: AoE spells will deal 2 "average damage" amounts (whether by hitting more targets that save or hitting fewer that don't). R = 18 (roughly based on 6 3-round combats). At will damage is firebolt, not modified. Ignoring accuracy. The wizard has

1 REU = 4d6 + 4 (one attack, sneak attack) = 18.
1 firebolt = 2d10 = 11. --> At will damage = 11/18 REU = 1/1.63 (repeating).
Resources: 10. 3 fireballs (including arcane recovery), 3 shatters, 4 chromatic orbs.

So total expenditure is: 3 fireballs, 3 shatters, 4 chromatic orbs, 8 firebolts.

Total damage: 8d6*2*3 + 3*2*3d8 + 4*3d8 + 8*2d10 = 48d6 + 30d8 + 16d10 = 391.

That's 21.7 REU (121%). Which is within the ballpark of what I remember for paladins and fighters. The actual throughput will vary tremendously based on number of targets and how well they save. If you roll saves like I do attacks and have lots of monsters in fireball formation...ouch.

[Rukelnikov]

The actual calculation included accuracy (assuming the same modifiers and same targets) and crits. It also considered TWF vs ranged for rogues, SnB vs GWF (yes, including the fighting styles) for fighters, and, IIRC, only GWF for the paladin. It was...labor intensive, so I didn't do it for most possible things.

Oh ok. Yeah it sounds like a lot of work.

Going back to your idea of 1 5th lvl slot, 2 4th level slots, and 1 3rd leve slot. I'm all in favor of reducing the ammount of effective actions caster get by late tier 2 and onwards.

But I think making everyone a Warlock has its own problems too. Like healing for instance, if you can take one short rest, you can likely take 2, that means a short rest is now a long rest, full hp, full spells (pre T3, and for T3 is "almost" full spells), this makes consequences even fleeter than they already are.

Paladins would smite with great vengeance and furious anger, knowing they get their smite slots back on a short rest.

Point is, I think it may be fixing some issues while creating others of similar scale.

[PhoenixPhyre]

Oh ok. Yeah it sounds like a lot of work.

Going back to your idea of 1 5th lvl slot, 2 4th level slots, and 1 3rd leve slot. I'm all in favor of reducing the ammount of effective actions caster get by late tier 2 and onwards.

But I think making everyone a Warlock has its own problems too. Like healing for instance, if you can take one short rest, you can likely take 2, that means a short rest is now a long rest, full hp, full spells (pre T3, and for T3 is "almost" full spells), this makes consequences even fleeter than they already are.

Paladins would smite with great vengeance and furious anger, knowing they get their smite slots back on a short rest.

Point is, I think it may be fixing some issues while creating others of similar scale.

But you can only do that once per day. You get 2 refreshes period per long rest, each one restoring half resources.

So you get 8 spell slots period. Total. No more. And you don't refresh any of your other special stuff. This is a little bit better than currently--a current warlock (assuming 2 short rests per day) gets 6 or 9 slots (depending on level). But they're not as high level--a level 9 regular!warlock has 9 5th level slots; by this model, they'd have 8: 2 5ths, 4 4ths, and 2 3rd.

And proportionally, a 9th level paladin would have the slots of a 4th level warlock: (2/4/2). A current paladin has 9 slots (4/3/2). So you trade 2 first level slots for an extra 2nd. Total. Per day. And that total doesn't go up with level--a 20th level alternate!paladin has 8 slots per day: 0/0/2/4/2. A 20th level current!paladin has 15: 4/3/3/3/2. And a 5th level slot isn't an improvement over a 4th for smiting.

If we calculate smite dice[1] per day: an alternate!paladin has 10+20+8 = 38 smite dice at level 20. A current!paladin has 54 dice per day. And can burn all of those in a single fight if it stretches long enough, while an alternate!paladin can only burn 19 per fight no matter what.

[1] an alternate concept for paladins that I rather like--instead of having spell slots at all, give them a pool of smite dice that scales with level. They can burn those to smite or to cast spells. Utterly neuters sorcadins that want to smite a lot--since they're no longer smiting from slots, they don't gain more smite slots by going sorcerer, while not really touching a full-class paladin.

[Rukelnikov]

I'll paste this here, cause the EDIt may have gone unnoticed.


Doing a napkin-math calculation at 5th level for a wizard:
Assumptions: AoE spells will deal 2 "average damage" amounts (whether by hitting more targets that save or hitting fewer that don't). R = 18 (roughly based on 6 3-round combats). At will damage is firebolt, not modified. Ignoring accuracy. The wizard has

1 REU = 4d6 + 4 (one attack, sneak attack) = 18.
1 firebolt = 2d10 = 11. --> At will damage = 11/18 REU = 1/1.63 (repeating).
Resources: 10. 3 fireballs (including arcane recovery), 3 shatters, 4 chromatic orbs.

So total expenditure is: 3 fireballs, 3 shatters, 4 chromatic orbs, 8 firebolts.

Total damage: 8d6*2*3 + 3*2*3d8 + 4*3d8 + 8*2d10 = 48d6 + 30d8 + 16d10 = 391.

That's 21.7 REU (121%). Which is within the ballpark of what I remember for paladins and fighters. The actual throughput will vary tremendously based on number of targets and how well they save. If you roll saves like I do attacks and have lots of monsters in fireball formation...ouch.

Ok, so the way you want to differentiate classes is by how spiky the damage function is, while all damage functions integrating to a rather similar ammount.

I think that's ok, but also take into account that is likely some of those spells will go to things other than damage, and that 18 rounds of combat, could also include rounds where the character is unable to act, or has to take a defensive action.

Also, have you done this for your proposed method to how it compares?

If I got it right it'd be something like this:

1 Fireball, 2 Shatters, 1 CO every short rest.

Assuming the "gitpg forum standard" of a sr every 2 encounters, that'd be:

3 FB, 6 Shatters, 3 CO, 6 Firebolts. Even without doing the numbers this would already be higher than the "regular" 5th level wizard you did.

Total damage: 8d6*2*3 + 6(!)*2*3d8 + 3*3d8 + 6*2d10 = 48d6 + 45d8 + 12d10 = 164 + 202.5 + 66 = 432.5

So that'd be 24.02 RDU, so ~33.4% above the Rogue, and ~10% above the regular Wizard.

With the 1 full refresh only it'd be:

2 FB, 4 Shatters, 2 CO, 10 Firebolts.

Total damage: 8d6*2*2 + 4*2*3d8 + 2*3d8 + 10*2d10 = 32d6 + 30d8 + 20d10 = 102 + 135 + 110 = 347

So that'd be 19.27 RDU, so ~7.1% above the Rogue, and ~11.25% below the regular Wizard.


But you can only do that once per day. You get 2 refreshes period per long rest, each one restoring half resources.

So you get 8 spell slots period. Total. No more. And you don't refresh any of your other special stuff. This is a little bit better than currently--a current warlock (assuming 2 short rests per day) gets 6 or 9 slots (depending on level). But they're not as high level--a level 9 regular!warlock has 9 5th level slots; by this model, they'd have 8: 2 5ths, 4 4ths, and 2 3rd.

And proportionally, a 9th level paladin would have the slots of a 4th level warlock: (2/4/2). A current paladin has 9 slots (4/3/2). So you trade 2 first level slots for an extra 2nd. Total. Per day. And that total doesn't go up with level--a 20th level alternate!paladin has 8 slots per day: 0/0/2/4/2. A 20th level current!paladin has 15: 4/3/3/3/2. And a 5th level slot isn't an improvement over a 4th for smiting.

If we calculate smite dice[1] per day: an alternate!paladin has 10+20+8 = 38 smite dice at level 20. A current!paladin has 54 dice per day. And can burn all of those in a single fight if it stretches long enough, while an alternate!paladin can only burn 19 per fight no matter what.

[1] an alternate concept for paladins that I rather like--instead of having spell slots at all, give them a pool of smite dice that scales with level. They can burn those to smite or to cast spells. Utterly neuters sorcadins that want to smite a lot--since they're no longer smiting from slots, they don't gain more smite slots by going sorcerer, while not really touching a full-class paladin.

Oh, ok, I missed the part where the rests only restored half resources

[Skrum]

This would, if I'm right, go some way in smoothing out the adventuring day--less incentive to go nova and then force a rest because you're out of stuff.

I strongly question this premise. Warlocks are the *opposite* of this, IMO. If an encounter is meaningfully dangerous, than warlocks need to use their slots, of which they generally have 2. That is not longevity, that's leaning into nova'ing and then rest dynamic.

What you're really describing is DM's not constantly throwing deadly encounters at the players. Which has it's own problems - creating narrative meaning to lots of lessor encounters becomes challenging, if there isn't a strong narrative and the players aren't really in danger, it becomes very tedious to actually play out that encounter.