What's the fun of having both Daily & Encounter recharges?

[OgreBattle]

So much digital ink has been spilt over 1 encounter workdays vs "ha ha vietnamese kobolds constantly harass you into the night no rest!!" and so on. All these different discussions hinge on some D&D characters not having a weardown mechanic other than HP, others recharging every short rest, and others requiring 8 hours rest.

It's the root of "this breaks my suspension of disbelief that an athlete requires rest after maximum muscle effort" type arguments too, people attach roleplay and 'realism' to the rest mechanics.

So what are yalls preferences for D&D games where different PC's recharge at different times, as a Game Master and as a Player?

[Krusk]

I think it provides for a relatively interesting resource management scenario. If everyone recharges at the same time with the same mechanic, there's an obvious best time. If everyone is different, it drives towards discussions around "should we keep going" and can make optimal tactics harder to determine and fosters debate. People solved 4e very quickly on this front. if everyone is different, there are a lot of variables, and it will change group to group, and potentially encounter by encounter.

I don't care if fighters can only do their super move 1/day.

[Kaelik]

Resource management provides three functions:

1) Provide reasons for people to wait and not advance and/or a resource to measure attrition against.

2) Provide decision making opportunities to the players so that combat isn't a solved state.

3) Set up opportunities to leverage asynchronous power in favor of the enemies or for the PCs to set up opportunities to leverage asynchronous power against enemies.

If you have only one resource management system in your game it moderately negatively impacts 3) because the only possible asynchronous usage is just "using your abilities to kill them before they can use theirs" or "a new set of enemies shows up when the PCs are low" which is much more limited than the possibilities if every class has a different resource management system.

It also majorly reduces 2). You just aren't going to figure out the resource management system that all players are going to enjoy working under better than any other. It doesn't exist. So having one system means almost everyone playing the game could enjoy the game more if their character had a different resources management system.

There is still a downside to giving each class a different one, maybe some people would enjoy Encounter Wizards but all you have are Daily Wizards, maybe some people would enjoy Rage Fighters, but you have at will Fighters. Ect. The more resource management systems you have the more limited the class selection is for people who enjoy certain resource management systems more than others. So you face the trilemma of:

A) Everyone gets the same resource system and everyone has maximum class choice, but also no resource management choice.

B) Every class has a different resources management system, but everyone has to balance class choice against resource management choice where one pole is "I still hate my resource management as much as I did under A) but I choose from all the classes" and the other is "I love my resource management, but I choose from one or two classes that have it."

C) You make 5 classes, Wizard, Rogue, Fighter, Cleric, Barbarian. Then you pick five management systems, Daily, Encounter, At Will, Rage, Winds of Fate. Then you fill out the grid so you have a Daily Wizard, an Encounter Wizard, an At Will Wizard, a Rage Wizard, and a Winds of Fate Wizard. Ect. Everything is five times as much work for the same thematic space and you get mad and quit when you have to make a Winds of Fate Wizard.

I personally choose B). I think B) is best because it at least allows people to trade off across the spectrum based on what they care about, but the reason I choose B) was because I enjoy designing classes around different resource management systems.

[Foxwarrior]

There's a fourth function: resource management systems can add a lot to the fluff/mechanics bond of a power source. Brewing daily potions, being a psychic who's ominous all of the time but only mad some of the time, a wacky gambler mage, should all feel different and resource management systems will keep them feeling different even when some of the actual effects of their abilities overlap.

[jt]

As a DM it's nice when players have resources that line up with the thing I'm designing. If they have N daily powers, then I get a cheap source of tension when I give them nearly N daily problems. They have to consider when it's worth rationing, and are more likely to have to resort to tools that aren't quite aligned with their problems, both of which are interesting. But I'm never actually designing days, so daily is a bit awkward and not a perfect fit. It'd be better for me if they had per-dungeon powers instead, but obviously that's not very well defined outside of a dungeon crawl, and also a mess flavor-wise.

Encounter powers are actually lined up with things I design, but I'm not designing encounters as a series of round by round challenges (and the encounter-building tools I'm given don't frame it that way either), so they're not as good at generating rationing or misaligned tools problems. I'm stuck hoping the system organically generates interesting uses of these. I rarely end up having to care about the difference between players' per-encounter and at-will abilities.

As a player, rationing dailies is interesting. I found rationing encounter powers interesting for a while, but it increasingly became clear that the answer is always to nova everything. I have the most fun with a well-designed, flexible at-will power. The at-wills that can maintain interest for an entire session tend to be either magic (minor illusion, thorn whip, familiars) or racial specials (aarokocras can fly), which makes it awkward that D&D wants at-will to be the special domain of the martial classes, which it gives the most boring boring at-will abilities.

[Thaluikhain]

In Dragon Warriors, all magic-users regain their magic at a certain time of day, and it's different for each type.

At first glance, that seems cool and thematic, at second you've gone and set it up so that the best time to do anything is different for each type of caster. That looks like a good way to stop the game while the player with the Sorceror argues with the player with the Mystic.

Also, the magic comes back instantly, which means it could be suddenly in the middle of a scenario, and maybe even in the middle of an encounter. Which seems not great.

(This doesn't apply to the Elementalist because they are terrible, but you won't want to play one because they are terrible for other reasons anyway)

[Foxwarrior]

As a DM it's nice when players have resources that line up with the thing I'm designing.

Oh yeah, I kinda forgot about this because the metasystems that make long-term resource management matter are so vague most of the time they barely matter. Technically Gempunks tries to make it so the distinction between things taking 6 hours and 15 minutes to recharge can come up by defining an "encounter" as more like "a conflict with a usually slightly weaker force than the player party, even if the enemies sneak off to heal halfway through and come back tomorrow" and then sets the expectation that you should be starting each encounter with basically full resources, although that sort of long-term hit and run tactic hasn't come up very often.

I imagine if you defined each quest as an "X Day race to save the world from destruction" then you could get the misaligned tools effect with daily power rationing. It would somewhat limit your quest design freedom though.

[hogarth]

As a DM it's nice when players have resources that line up with the thing I'm designing. If they have N daily powers, then I get a cheap source of tension when I give them nearly N daily problems.

Do your players know in advance that they will see N encounters in one day? If not, then it seems to me that the tension would be lost (in my experience, that usually translates to hoarding daily abilities in order to avoid running out of gas too early, and then going supernova when it's clear the day is almost over).

[deaddmwalking]

For our homebrew almost every ability uses the same limited resource - spells and rounds of rage or special attacks/abilities deplete mana; rest recovers mana. Most characters need 3+ hours of rest to recover fully. Players have a much more limited resource called action points; some of them can be traded for regaining mana faster, but they can also be traded for 'not dying' so people don't want to run out of those. This allows us to take a 1 hour rest and continue with reduced capability but we can still do our important abilities; longer rests mean that we can probably have 3 or 4 encounters of moderate difficulty before we're 'on fumes'. What works best about it is that we don't have different players fighting about a short rest or a full rest - it's really just a question of how hurt someone is and what we think the minimum amount of resources we need.

When most of the party is low/out of action points, people agree that a full rest is required; when people are only slightly wounded or we have enough mojo left to handle the healing without depleting all our mana, we continue on.

[jt]

I imagine if you defined each quest as an "X Day race to save the world from destruction" then you could get the misaligned tools effect with daily power rationing. It would somewhat limit your quest design freedom though.

Might work with one of those settings where dungeons are a transient magical phenomenon instead of the result of bad stuff slowly accumulating in a ruin. A dungeon suddenly appears so somebody has to get in and destroy its core within 7 days, or it'll explode and flood the countryside with monsters. That's just the way the world works, that's why we have adventurers.

I'm personally not a huge fan of those settings in the first place, but if it has that sort of handy game design hook it's more tempting.

Do your players know in advance that they will see N encounters in one day? If not, then it seems to me that the tension would be lost (in my experience, that usually translates to hoarding daily abilities in order to avoid running out of gas too early, and then going supernova when it's clear the day is almost over).

I give a lot of clues about how much day is left, in terms of distance to obvious campsites, size of dungeon, that sort of thing. I also push the characters hard enough that they need to spend resources in every encounter.

There's definitely still some hoarding / nova going on. Like I said, daily doesn't actually line up that well with anything.

It also distorts some encounter design. Any combat in a town ends up having to be way harder than one a dungeon crawl, since it's so easy to nova next to an inn. Which is pretty metagamey, but more fun than constantly exiling characters from town and trying to assassinate them at night, so I do it anyway.

[MGuy]

I pretty much agree with everything kaelik posted except that while I prefer B, want to do B, and will lean more for playing games that are like B, A is easier for me to design around by far.

[Zinegata]

So what are yalls preferences for D&D games where different PC's recharge at different times, as a Game Master and as a Player?

In general, outside of RPGs games with recharge mechanics have dropped all pretense of hourly or daily timing, in favor of recharging based on the next encounter.

I agree with this approach because if you're aiming to have balanced combat encounter, then you want to know all the resources the PCs will actually have available. You don't have to guessimate how much was stripped away prior to that encounter.

It also makes it much more likely for players to actually use the majority of their abilities. A lot of players end up trying to "save" their one-per-day ability for a "real emergency", but in practice what occurs is that they either don't use it at all.

[OgreBattle]

Resource management provides three functions:

1) Provide reasons for people to wait and not advance and/or a resource to measure attrition against.

2) Provide decision making opportunities to the players so that combat isn't a solved state.

3) Set up opportunities to leverage asynchronous power in favor of the enemies or for the PCs to set up opportunities to leverage asynchronous power against enemies.


A) Everyone gets the same resource system and everyone has maximum class choice, but also no resource management choice.

B) Every class has a different resources management system, but everyone has to balance class choice against resource management choice where one pole is "I still hate my resource management as much as I did under A) but I choose from all the classes" and the other is "I love my resource management, but I choose from one or two classes that have it."

C) You make 5 classes, Wizard, Rogue, Fighter, Cleric, Barbarian. Then you pick five management systems, Daily, Encounter, At Will, Rage, Winds of Fate. Then you fill out the grid so you have a Daily Wizard, an Encounter Wizard, an At Will Wizard, a Rage Wizard, and a Winds of Fate Wizard. Ect. Everything is five times as much work for the same thematic space and you get mad and quit when you have to make a Winds of Fate Wizard.

I personally choose B). I think B) is best because it at least allows people to trade off across the spectrum based on what they care about, but the reason I choose B) was because I enjoy designing classes around different resource management systems.

Been considering how most video games go with A, an energy or stamina meter as a universal mechanic, which I think Zinegata is mentioning.

So Kaelik's type A resource. So here's another consideration...

Same encounter resource mechanic, but different mid-encounter recharges

Such as restoring all encounter points with a 2nd wind, but another class recharges one point at a time with one of their actions. This encourages one class to alpha strike and recharge everything just once, and another class can still spike their power but will recharge slower but able to do this in an extended fight. Kinda inspired by 9Swords warblade spending an action to recharge everything vs swordsage recharging one at a time

[Emerald]

In my view, every class should have a different resource system. That's literally why OD&D was designed with classes, after all, to provide different play experiences based on different resource schedules.

That doesn't mean you can't have multiple classes using the same general kind of resource system (like wizards, sorcerers, and beguilers all being arcane casters with per-day spells and similar capabilities), just that each class should have their own spin on it (like how wizards are prepared, sorcerers are spontaneous-chosen-spells, and beguilers are spontaneous-fixed-spells). It's a setup like the wizard and wu jen that should be avoided, where the resource system is completely identical (Int-based prepared spells from a spellbook with identical numbers per day and numbers acquired) and they just have slightly different spell lists for no good reason beyond "but it's Asian!", in which case one can just be a subclass or selectable class feature (or whatever similar doodad the system uses) of the other.

So I suppose that's a vote for something between Kaelik's options B and C, where by default Wizard = Daily Spells but if you want to add a Wizard-themed class that's not Daily-based or a Daily-based class that's not Wizard-themed that's fine so long as they're sufficiently distinct to be worth existing next to the Wizard.

A lot of computer RPGs choose option A, but that's more due to programming limitations and a broad acceptance of ludonarrative dissonance among players, not something you necessarily want to replicate on tabletop. The Mass Effect games, for instance, have biotics, supertech, and combat training all work on a cooldowns-and-toggles system because it makes coding everything easy and means the Adept, Engineer, and Soldier all have roughly equal complexity, but you might want a tabletop ME RPG to use different systems for the three (like Shadowrun does for magic, decking, and combat) because you might want e.g. Sentry Turrets to be modeled as equipment and Adrenaline Rush to use some sort of stamina system, or whatever.

I agree with this approach because if you're aiming to have balanced combat encounter, then you want to know all the resources the PCs will actually have available. You don't have to guessimate how much was stripped away prior to that encounter.

This assumes, of course, that you're balancing each encounter in a vacuum. The whole point of Vancian casting in AD&D was that the DM doesn't care about balancing a single encounter, they care about balancing an ancient crypt (or haunted forest or...) as a whole. The party headed down into the dungeon and the magic-user and cleric had to decide which fights were worth breaking out the big spells for and which were worth leaving to the fighters and thieves, and if they planned poorly or weren't judicious enough with their resources, too bad.

There are certainly plenty of downsides to this approach, which have been rehashed in every Vancian casting discussion ever, but there are benefits as well. A purely encounter-based design framework can't really handle common scenarios like "The [encounter-appropriate number] castle guards in this room called for reinforcements from the guards down the hall and now there are [3× encounter-appropriate number] guards in the room!", and if neither the party nor the baddies can do hit-and-run tactics or the like (because all defensive buffs and healing refresh per-encounter) to weaken a stronger opponent, then the range of opponents that can be faced is constricted quite a bit.

It also makes it much more likely for players to actually use the majority of their abilities. A lot of players end up trying to "save" their one-per-day ability for a "real emergency", but in practice what occurs is that they either don't use it at all.

This is mostly dependent on the frequency, type, and quantity of per-day abilities in question. A crusader who has a single Smite Evil per day is going to be reluctant to pull it out "too early," but a blackguard who has 7 Smite Goods per day is going to be more willing to use them, a paladin who has 4 Smite Evils, 3 Turn Undeads, and 5 spells is going to be fine using daily abilities early and often, and a cleric who has 30 spells and 6 Turn Undeads is going to resort to daily abilities first last and always with no issue.

In my experience, 1/day abilities tend to be hoarded but 3+/day abilities tend not to be, because the latter aren't all-or-nothing things so missing with them, using them early in the day, etc. isn't nearly as risky. The easy solution, then, is to hand out per-day abilities with enough uses that it feels less like a Final Fantasy healing potion and more like a per-every-couple-encounters ability.

[Ice9]

I like the idea of different classes have different resources systems, but in practice "You need to have X encounters per day if you want any balance" and "Recharge is based on purely metagame factors, if you go take a three month break in the middle of 'a plot' or 'a dungeon' then it doesn't count" are both shitty.

So like, different mechanics within an encounter? Sure. Having longer-term recharge cycles? Fine, if it applies to all classes. Having different longer-term recharge cycles? Maybe.

Basically I want to support certain things more than I want fancy resource management:
* Sometimes you just fight the one thing in a given day, and that's fine.
* Recharging is comprehensible IC.
* Reasonable level of intra-party balance, regardless of pacing.

But if a recharge mechanic can support those and still vary between classes then all the better.

[Kaelik]

"Resource Management System" is also more than just at will, encounter, daily.

Various "at will" systems like Winds of Fate, Toggling, and "at will from a small list but you change the list with some time" are all different resource management systems. And for Daily the difference between the Kaelik Cleric style "buffs but then you expend them for a big effect" versus Wizard "your spells are literally worthless if you don't cast them" are also different mechanics.

Various approximately encounter level trade offs like recharge and cooldown and use up and rest are also different.

But fundamentally the first question you need to ask is what you are balancing these systems around, and I just did not choose "an encounter" or "X encounters a day" as the standard of balance.

[PhoneLobster]

I think fundamentally the problem with these specific schedules and mixed schedules is a need fairly to acknowledge the real practical applications. What these rules really mean and what real groups really do with them.

The idea that most resource renewal schedules that exceed the main formally managed short term "combat" time scale of the game are even a firm formal mechanic is for most actual rules sets and basically all actual game tables basically a bunch of wank.

Because whatever the pretense in practice the way basically all RPGs manage longer term time passing and more importantly, what number of encounters of what potential resource cost interrupt it, generally all boil down to the GM just decides what they feel like. And maybe if they can't decide they might roll a non binding dice for inspiration, and maybe if the system or adventure feels helpful it will offer a non-binding table or something to roll on.

And sure players can decide to do things that might mean another encounter or not, but generally the GM gets final say by means of how many waves of surprise goblin ninjas they feel like throwing at a party as it is trying to ask the GM to please hit the time progress montage button until they arbitrarily get their resources back by means of saying "we camp" or similar.

What? There are (rarely) some rules, limitations or at least guidelines about the number of waves of Goblin ninjas that can turn up? Well, rarely, also bullshit, and also so what most groups won't know and most of those that do won't care.

Out of combat time is, fairly reliably, bullshit time. Out of combat resource renewal schedule is by extension, fairly reliably, bullshit resource renewal schedule.

That requires acknowledgement in order to hope properly manage it, basically no one does that.

So any proper management of that sort of resource schedule, and any particular desirable outcome, whether you randomly decide you desired attrition warfare or reliable easy resource renewal is basically happening by accident at the coincidental whim of random GMs. And who can blame them when all the "rules" did was name drop a few real world time units some of the time and left them with an unmanageable only semi-formally defined mess full of holes and non-binding freedom amounting to little more than a pretense that absolutely anything at all helpful had been provided for them to work with in the first place?

[Zinegata]

Been considering how most video games go with A, an energy or stamina meter as a universal mechanic, which I think Zinegata is mentioning.

Not really, my point is everyone refreshes at the same time in most modern games. Not that they have the same resource system.

In Monster Hunter for example there are no “days” or “encounters”. Instead everyone in the party accepts a common mission with an arbitrary timer. Everyone is at full strength at the start of said mission.

Because the thing here is you don’t want to start a mission where some players may still be gimped due to damage from a previous mission. You want the encounter to be balanced based on all party members being at 100% in the beginning.

Same encounter resource mechanic, but different mid-encounter recharges

I’ll be blunt here - I think that TTRPGs having different refresh times for various abilities and powers had always been a mistake. This includes Vancian casting. Because the pace of the game ends up being dictated by these arbitrary refresh rates instead of the pace of the action.

The main reason why you have these different refresh rates had always been to force players to time their usage. If a Wizard can cast unlimited Fireballs, then obviously they’ll just spam it every encounter. If they don’t have unlimited fireballs, then they’ll try to save it and think carefully when to use it. The thing is, it also results in the “lets rest before going into the final boss” issue

Other games by contrast use different means to make players time the usage of their abilities.

For instance - a lot of modern action games have a “light attack” and a “heavy attack”. If pure damage-dealing output is the only concern, then people should just spam heavy attack.

But in most games spamming light attack is valid, because a light attack completes faster than a heavy attack and leaves the player able to dodge an enemy counterattack. Heavy attacks are instead saved for moments when the opponent is staggered or can’t fight back, and pure damage output can be spammed without risk.

Hence, the “timing” of ability usage in games with light / heavy attacks isn’t based on any arbitrary refresh rate. Instead its based on the enemy state. If the enemy is still actively defending himself, spam light attack. If they’re now helpless, finish them with heavy attacks.

Likewise, in Gloomhaven you can’t just spam your best abilities - because you have to wait for your character to draw that card out of your character’s deck. Its arbitrary and random - but it does force players to think of how to best handle the situation using the cards and abilities they have in-hand instead of simply spamming fireball. It thus turns into a turn-by-turn optimization puzzle.

In short, there are other ways to make the players think about timing of abilities and how to optimize their actions; without being constrained by the refresh per day model.

[Zinegata]

I agree with this approach because if you're aiming to have balanced combat encounter, then you want to know all the resources the PCs will actually have available. You don't have to guessimate how much was stripped away prior to that encounter.

This assumes, of course, that you're balancing each encounter in a vacuum. The whole point of Vancian casting in AD&D was that the DM doesn't care about balancing a single encounter, they care about balancing an ancient crypt (or haunted forest or...) as a whole. The party headed down into the dungeon and the magic-user and cleric had to decide which fights were worth breaking out the big spells for and which were worth leaving to the fighters and thieves, and if they planned poorly or weren't judicious enough with their resources, too bad.

There are certainly plenty of downsides to this approach, which have been rehashed in every Vancian casting discussion ever, but there are benefits as well. A purely encounter-based design framework can't really handle common scenarios like "The [encounter-appropriate number] castle guards in this room called for reinforcements from the guards down the hall and now there are [3× encounter-appropriate number] guards in the room!", and if neither the party nor the baddies can do hit-and-run tactics or the like (because all defensive buffs and healing refresh per-encounter) to weaken a stronger opponent, then the range of opponents that can be faced is constricted quite a bit.

It also makes it much more likely for players to actually use the majority of their abilities. A lot of players end up trying to "save" their one-per-day ability for a "real emergency", but in practice what occurs is that they either don't use it at all.

This is mostly dependent on the frequency, type, and quantity of per-day abilities in question. A crusader who has a single Smite Evil per day is going to be reluctant to pull it out "too early," but a blackguard who has 7 Smite Goods per day is going to be more willing to use them, a paladin who has 4 Smite Evils, 3 Turn Undeads, and 5 spells is going to be fine using daily abilities early and often, and a cleric who has 30 spells and 6 Turn Undeads is going to resort to daily abilities first last and always with no issue.

In my experience, 1/day abilities tend to be hoarded but 3+/day abilities tend not to be, because the latter aren't all-or-nothing things so missing with them, using them early in the day, etc. isn't nearly as risky. The easy solution, then, is to hand out per-day abilities with enough uses that it feels less like a Final Fantasy healing potion and more like a per-every-couple-encounters ability.

I don’t disagree at all - but the issue I’m highlighting here is that most games have moved to an encounter-centric refresh for a simple reason:

It is much, much easier to balance multiple individual encounters - each where the power level of the party remains the same with no variation due to attrition - than it does to balance an entire adventure consisting of multiple encounters that are supposed to gradually chip away at a party’s resources.

Indeed, the mindset of a lot pf adventure designs tends to be contradictory to the actual evolution of the party’s power level. The party starts off at full strength and gradually gets weaker. But most adventures usually start off with weaker enemies that gradually become tougher.

As a result - unless DMs adjust on the fly or allow reset rests after a particularly damaging encounter - most adventures end up being a “poor gets poorer” type situation. Bad luck with the early goblin encounter - costing you more healing and spells than anticipated - means the next battle is now even harder unless the players decide to just postpone the adventure and reset their power level.

[OgreBattle]

Y'all ever played a game where the opposition party's use of dailies and encounters was taken into account when they fight the PC's?

[PhoneLobster]

Y'all ever played a game where the opposition party's use of dailies and encounters was taken into account when they fight the PC's?

No.

Not saying it's impossible, not saying I haven't like wow, barely ever, done something loosely like that as a GM. But thinking about it. No. No GM has ever in any system ever done that in any game I've played in. And I feel like I could play another decade or more keeping an eye out and it would still be unlikely for me to see it happen.

I think that's kinda indicative of a large part of the whole "bullshit time" problem.

And what's your plan here?

Hell. What's your statement here?

Is this is a proposed solution to "bullshit time"?

Trying to counter "but the GM just decides how many encounters happen during bullshit time scale" with "but the GM could decide they are weaker encounters". Would be kinda a non solution.

If that's the case I'm not even sure it's at the level of calling it a clearly distinct concept from the thing it's meant to solve.

[hogarth]

Y'all ever played a game where the opposition party's use of dailies and encounters was taken into account when they fight the PC's?

On rare occasions in published modules I'll see a stat block for a spellcaster saying "NPC has already cast spells X and Y today". But otherwise, no.

[Krusk]

Y'all ever played a game where the opposition party's use of dailies and encounters was taken into account when they fight the PC's?

If I'm designing an NPC spellcasting villain, I'll make a note of their spellcasting tactics. For example, if they have mage armor that lasts 6 hours, I might say they cast it at dawn, and every 6 hours on the dot (reducing spells available in the encounter if they fight at 10PM). I might also say hey cast it at the slightest suspicion of an encounter, but don't bother while walking around their house until they have showered, read the paper, and walked into the office (so to speak). I just recently wrote up one who was specifically super arrogant and lazy. His mage armor lasts IIRC 11 hours and I said it was a 100% chance he casts it in the morning, but a 75% chance he forgets to recast later in the day when it wears off until something jogs his memory.

If they have other spells with long durations or long impact, I usually make notes of those as well. "She can cast plant growth 3/day as an SLA, so she does 3 every morning to make her forest just obnoxious thick." So I know she can't do it in a fight.

[OgreBattle]

Im trying to figure out what my plan for 'bullshit time' is, but most boss battles I've played through or read about in adventure paths have a fresh boss in their home and the PC's are on the offensive through multiple encounters to reach them.

Say if the players had some NPC minions already assaulting the evil castle so the big boss already burned some resources in their encounters. Or the PC's are the ones defending the Good Castle as the big bad storms in. Or both Good & Evil guys are chewing through the other's armies to clash in the climax of the battlefield.

Perhaps it can be a reward for info gathering, sneaking and so on. The big bad is in the middle of crafting their spooky army or sending forth plague, their resources have already been committed so they got less death rays to shoot at PC's.

[jt]

I usually remove a few spells from an enemy caster's list as "already used," but I've never had them try to pace themselves against a player party (or really had a situation where the enemy should do that). Mostly the players face casters in their lair with ample warning, sometimes the players are on the receiving end of a scry and die.

I've never actually seen players plan to sneak up on an enemy caster and attack them during their day to day routine. If they did, I guess I'd give the enemy a fair number of escape spells, then fill the rest of their list with already-used slots and irrelevant economic utility spells.