Proposal: concentrating on buff spells

[rigrig]

Optimal play for spells like Ozocubu's Armour seems to be keeping them up at all times (in case some monster is hiding around the corner), thus interrupting auto travel/exploration every time they start to run out.
Repel/Deflect missiles have lately been made much less annoying in this aspect, and I'd like something similar for other buffs.
My proposal is that you can 'concentrate' on keeping your buffs active, preventing them from timing out. Only once your concentration is broken would the regular timeout start.
The idea is that you don't have to bother with the tedious task of refreshing your buff spells all the time while exploring, but during actual danger you still risk them running out in the middle of a fight.

Simply casting a concentration spell would make you start concentrating on keeping your spell active.
Concentration would be broken by pretty much anything interesting happening:
* noticing an enemy
* being hit
* avoiding anything by dodging/blocking/repelling/deflecting
* being scared of losing your concentration, e.g. by flying over lava
* randomly losing your concentration, depending on spell success: the idea being that you can keep up low-failure spells (virtually) indefinitely, but you can't re-attempt a difficult spell in safety until it finally works, and then expect it to last.

An alternative to random concentration loss might be a special casting mode to start concentrating on a spell, but with a higher failure rate. Not too fond of that idea though, as it seems to overcomplicate the casting of spells.


Concentrating would cost <spell hunger>/<spell duration> nutrition per turn, so it uses up exactly as much as when re-casting the spell every time it runs out.
Because of this, there would be a new ability to "stop concentrating", in case you're afraid of starvation. (Possibly with a choice of which spells to stop concentrating on? Although that seems a bit overcomplicated.)
Alternatively, it might only work for spells with zero hunger cost, thus also preventing you from keeping up high-level spells all the time. That would complicate things like the staff of energy/Necromutation/mummies though.

Maybe it should also use up a small bit MP each turn, or lower your max MP while active, but this would make it a worse than buffing up in safety and channeling MP back up to full before exploring.

Spells that I think might qualify for this:
* Ozocubu's Armour
* Stoneskin
* Condensation Shield
* Shroud of Golubria
* Flight
* Phase Shift

And possibly also:
* Control Teleport
* Regeneration
* Ring of Flames
* Darkness
* Necromutation

Also, branding weapons:
* Fire Brand
* Freezing Aura
* Poison Weapon
* Excruciating Wounds
* Warp Weapon
* Sure Blade

[Galefury]

Like all other permabuff proposals this would not just affect the interface, but also have a bunch of important gameplay effects. Also there are some unsolved details, but I'm not going into these. Effects on gameplay that I can see right now:

1. You start every fight with all of your buffs up.
2. Not only do you get your buffs, you get them without "paying" turns and MP.
3. Spell success of buff spells becomes less important (cf. RMsl/DMsl).

Almost any version of permabuffs would have these effects, and I don't think these are good. I'm starting to think the best option would be just radically reducing the duration of most buffs.

[rigrig]

1. You start every fight with all of your buffs up.
2. Not only do you get your buffs, you get them without "paying" turns and MP.
3. Spell success of buff spells becomes less important (cf. RMsl/DMsl).

My point is that I'm doing that right now as well, except I have to manually recast them every time a buff is running out (and auto travel/explore keeps halting for them running out).
I realise this might not be the optimal strategy for every species, but for a mummy/bloodless vampire/lichform with channelling, keeping buffs up all the time seems like a no-brainer. (And without channelling it just means even more down-time waiting for MP to regen)
Right now I find it too tedious to keep refreshing them all the time, but I do keep them up all the time while exploring branch endings.

Almost any version of permabuffs would have these effects, and I don't think these are good. I'm starting to think the best option would be just radically reducing the duration of most buffs.

Reducing buff duration would only make this worse, and introducing glow for e.g. Ozocubu's Armour seems a bit too harsh, as it should probably be refreshable mid-battle.

[TeshiAlair]

I personally think all buffs should cause either some glow or have some form of magical-exhaustion effect as to prevent spamming, and in exchange make them slightly better when you do use them. I'd keep duration the same, but have that MP not be restored naturally until the buff wears off.

[Galefury]

My point is that I'm doing that [starting with buffs up at no MP and turn cost] right now as well, except I have to manually recast them every time a buff is running out (and auto travel/explore keeps halting for them running out).

Yeah, that's what people (including me) said about RMsl and DMsl. But it's far from the same.

[argonaut]

How about if you had only one charms slot. If you cast a charm it goes into the slot (replacing anything that's there) and will automatically get refreshed at the normal rate. Buffs can still expire in combat as usual.

Buffs that would use this
* Ozocubu's Armour
* Stoneskin
* Condensation Shield
* Shroud of Golubria
* Flight
* Phase Shift
* D/RMSL
* Battlesphere

This means you only get one buff and have to choose if Ozo's or dmsl is better for defense or if you want Battlesphere up for offense.

It also opens up some design options for races (Ogres or Felids?) or iterms (Amulet of charms) that add a charms slot.

[jejorda2]

I think of a charm as a physical item, like a piece of jewelry. What if you have to unequip a piece of jewelry to use a buff spell? So casting Repel Missiles puts a conjured ring or amulet of Repel Missiles into your first empty ring or amulet slot. Octopodes are the super-charmer species. Others can cast up to three buffs or wear up to three pieces of jewelry. Buffs stay active until [R]emoved or used up from a big hit (Shroud) or a fire attack (Ice buffs.) Depleted Battlespheres follow the player, but don't fire until recharged by casting the spell again.

Ash and Ru get even more costly!

[Siegurt]

There's probably half a dozen threads covering this topic. Probably we don't need to re-hash those conversations again (and again and again)

My take on this (to shorten it) is that simply trying to implement an interface-centric solution on the existing buffs doesn't work without causing other problems, and what really needs to happen is the buffs each need to be replaced with a newly designed spell that it's not optimal to spam constantly, trying to implement a "everything" solution leads to too many horrible and poor results.

Each spell being redesigned on it's own also allows for them to be tested and implemented individually.

Some suggestions I've made in the past along these lines:

Stoneskin becomes "meld with earth" which gives an AC bonus as long as you don't move.
Repel missiles becomes "Wind Wall" which creates some short-lived non-moving clouds which reduce the accuracy of all projectiles passing through.
Ozocubu's Armour becomes "Snow blind" which gives all creatures in LOS an attack malus

etc. etc. (You get the general idea, create replacement charms which effect the immediate environment, or the creatures in it, not being able to carry it with you or cart creatures back to it is the key here)

[Hurkyl]

It's not completely clear to me what the actual opinions are, but it looks like the best is the enemy of the good -- there seems to be a reluctance to actually make a change that roughly maintains the status quo except decreases annoyance, because such a change is not the wonderful ideal solution people are hoping exists.

[njvack]

The challenge with the changes that roughly maintain the status quo is that they generally don't actually maintain the status quo, and have a bunch of edge cases that would need careful thought and design. So implementation isn't trivial and the underlying problems of the buff spells remain.

My sense is that the people who might actually program a bandaid permabuff system have decided it's not the most pressing (or interesting) thing to do. If someone wants to try an implementation and see if the "yes but what about" issues are overrated, well, the code is there. I'd try it it.

But I suspect Siegurt's take is pretty accurate.

[rigrig]

(You get the general idea, create replacement charms which effect the immediate environment, or the creatures in it, not being able to carry it with you or cart creatures back to it is the key here)

That would remove the tedium from trying to keep them up all the time, but seems like quite a nerf for every situation where you can reasonably expect something nasty to be around the corner/downstairs, and you would really like to buff up before facing it.
I suppose the problem lies in the fact that paranoid behaviour tends to be healthy in the dungeon: you can be pretty sure there is a welcoming party on Vaults:5, but it's even safer to assume there's the same kind of party down every stair and around every corner.

I agree that changing the mechanics of a whole bunch of spells at once is probably not a good idea, but it seems there isn't that much difference between some of them, apart from spell school and power.

[and into]

One should not be able to benefit too much from degenerate game play, but making all charms into something like rmsl would be a ridiculous buff to a school of spells that already tends to be too broad in its usefulness. The possible objection I could see from Siegurt's approach, which you kind of raise, is that we don't want to nerf the advantage of a player having advanced knowledge of a tough fight, which you can derive from high stealth, following Ash, and so on. But even if we reformed oz's armor, shroud of golubria, stone skin, and the like, in order to keep them from being spammable, there would still be plenty of buffs you can still put up in advance, when you *know* there is a fight coming: haste, invis, !might, !agility, !brilliance, form spells, god-granted buffs, etc.

[acvar]

The real question is: are buff spells even a good mechanic at all? I mean is it reasonable that a character that has spent all his experience on dodging be out dodged by somebody that split there experience between dodging, spellcasting, air magic, and charms? Is it reasonable for a character that has spent all his experience in a weapon to be out damaged in melee by a character that has split his experience between a weapon and several magic skills to get song of slaying castabel? It does not seam right. I think if a fire mage suddenly got better damage output from his firebolt just for diverting some of his experience into long blades people would rightfully call foul, so why is the reverse so acceptable?

[Sprucery]

It's magic.

[Siegurt]

The real question is: are buff spells even a good mechanic at all? I mean is it reasonable that a character that has spent all his experience on dodging be out dodged by somebody that split there experience between dodging, spellcasting, air magic, and charms? Is it reasonable for a character that has spent all his experience in a weapon to be out damaged in melee by a character that has split his experience between a weapon and several magic skills to get song of slaying castabel? It does not seam right. I think if a fire mage suddenly got better damage output from his firebolt just for diverting some of his experience into long blades people would rightfully call foul, so why is the reverse so acceptable?

Berserking and Heroism'd meleers out-damage song of slaying using characters, and do so without the added risk of making large amounts of noise every turn (In fact I'd argue that SoS isn't worth the XP invested at all, unless you were going to spend that XP on charms anyway). A closer analogy would be Haste, but Haste helps spellcasters as much as meleers (And Yes, a fire mage does suddenly get better damage output by diverting some of his XP into charms so he can cast haste, and no-one calls foul, because Haste, and Song of Slaying for that matter have built-in drawbacks that make spamming them not a good idea, so you use them when the situation calls for it)

So I should revise my earlier statement, It's my opinion that charms should either: 1. Functionally be unable to be used in advance of combat or 2. Have some drawback that makes it unacceptable to spam them (See Swiftness, Haste, Song of Slaying etc.)

It's actually the case that this game doesn't just not discourage training opposing things to make your character more powerful, it actively encourages it, every character should have a diverse skill set, because that's optimal, a melee user getting charms is no different than a fire magic user getting dodging or fighting and it's a good thing.

[KoboldLord]

There's nothing wrong with spending xp to gain an advantage in combat. The only problem comes from the irrelevant busywork of re-activating an ability periodically so you never have to deal with a combat turn without having the advantage you already paid for.

As they exist now, most charms and charm-likes should just activate when you memorize them and stay up until you amnesia them. A few could be dispelled under certain circumstances and then come back after a delay, like the demonspawn ice facet that gives +10AC, if they're the sort of charms that get disrupted by fire or ranged attacks or whatever. In the general case, you paid your xp, and the cost of re-casting them periodically is a clerical chore and not a real expenditure of game resources.

I don't support the proposal to change all charms into ultra-gimmicky abilities that are usually worse than useless unless you have a very specific build revolving around them. We already have transmutation, a whole skill full of overspecialized spells that are best ignored every time you haven't made the choice to make your entire build revolve completely around them. Charms should ideally become less annoying to use, not more so, even if their power is raised to compensate for the additional annoyance.

[Kismet]

There's nothing wrong with spending xp to gain an advantage in combat. The only problem comes from the irrelevant busywork of re-activating an ability periodically so you never have to deal with a combat turn without having the advantage you already paid for.

As they exist now, most charms and charm-likes should just activate when you memorize them and stay up until you amnesia them. A few could be dispelled under certain circumstances and then come back after a delay, like the demonspawn ice facet that gives +10AC, if they're the sort of charms that get disrupted by fire or ranged attacks or whatever. In the general case, you paid your xp, and the cost of re-casting them periodically is a clerical chore and not a real expenditure of game resources.

I don't support the proposal to change all charms into ultra-gimmicky abilities that are usually worse than useless unless you have a very specific build revolving around them. We already have transmutation, a whole skill full of overspecialized spells that are best ignored every time you haven't made the choice to make your entire build revolve completely around them. Charms should ideally become less annoying to use, not more so, even if their power is raised to compensate for the additional annoyance.

Charms could be memorized into their own specific strategic slot... or count for double/triple spell levels... or something.

[nilsbloodaxe]

Ozocubu's Armour becomes "Snow blind" which gives all creatures in LOS an attack malus

While I care very little about flavor arguments, this is basically a Hex. But this brings up an idea, eliminating some Charms, re-purposing the rest as AoE/LOS Hexes, and eliminate the Charm school. Maybe we could even rename Hexes Enchantments or something like that.

[Galefury]

Funny.

By the way, there's a branch called charms-reform that doy is working on. It makes some spells cast from max MP instead of current MP. Currently this seems to be RMsl, DMsl and Flight. Doesn't mean it's definitely going to work like that in the next version, but it's pretty likely.

[Lasty]

The version of charms reform I'm currently pushing is to make more charms be toggleable always-on spells, but have their chance of proccing on each triggering event be proportional to spell success rate and the degree of their effect be proportional to spellpower. Arguably some mana or time penalty is still needed. Some options include:
* Have a chance to drain 1 mp on proc, perhaps modulated by spellpower.
* Have each active charm slow down your MP regeneration.
* (Extreme, probably bad idea) have each active charm add a delay (or chance of delay) to each action, perhaps modulated by Spellcasting or Charms skill.

To take rMsl as an example, instead of casting it and then having it arbitrarily fall off periodically, once you turned it on the effect would apply continuously for the rest of the game until you turned it off. While active, the effect would trigger on between 0% and 100% of the missiles fired at you, scaling linearly w/ spell success. When it triggers on an attack, it would reduce enemy missile accuracy by a random amount whose max value ranges from 0% to 100% based on spellpower (max power 100?) for single targets, or half that for beams.

Under this plan, the Charms skill ends up being a bit like the Armour skill: it's probably something everyone wants some of, but how much you can make use of depends on the loot (spells) you find and how well you've trained the relevant stat.

As a side bonus, since these charms are always-on and effectively "equipped", we could take their effects out of the status display and add a key to list active charms, reducing the burden on the already-overworked status display.

Haste could not (or at least probably should not) be modified to use this system, since it "procs" on every action, and since having Haste scale with spellpower would probably be a huge nerf to low-int characters. There have been enough other posts about what to do with Haste, so I'll exclude it from this discussion.

[johlstei]

One should not be able to benefit too much from degenerate game play, but making all charms into something like rmsl would be a ridiculous buff to a school of spells that already tends to be too broad in its usefulness. The possible objection I could see from Siegurt's approach, which you kind of raise, is that we don't want to nerf the advantage of a player having advanced knowledge of a tough fight, which you can derive from high stealth, following Ash, and so on. But even if we reformed oz's armor, shroud of golubria, stone skin, and the like, in order to keep them from being spammable, there would still be plenty of buffs you can still put up in advance, when you *know* there is a fight coming: haste, invis, !might, !agility, !brilliance, form spells, god-granted buffs, etc.

Would it really be that big a buff? In practice it would mean more players would use the spells to their full capacity, making it a buff to players, but most of the spells can effectively function this way already. (The main difference being take off armour->cast->put armour on, and ditto for brilliance/wizardry/etc).

What if, until a perfect solution is developed, we did something that pretty well approximates the current "optimal" behavior, while keeping the behavior the same. There could literally just be an auto-recast list, that recasts any spell on it automatically any time you explore or travel? Even just a "re-up buffs from my list" spell macro would help. Such a list would 1) continue checking spell failure/power each time, so no armour swap problems, 2) not be a huge buff, since the only behavior I would automate is already possible, it's genuinely just an interface change unlike other proposals.

Obviously, a genuine charms reform is necessary, and there are some good proposals out there, but I've been waiting a long time, my fingers hurt from casting stoneskin. If I wrote a patch that simply adds ui as I described, would devs consider it? (Thank you for changing swiftness, I was the guy who kept that up constantly to save piety, my fingers appreciate it.)

Please point out if I've missed anything - I love the repel missiles change but the armor swap issue is a real one.

[njvack]

Please point out if I've missed anything - I love the repel missiles change but the armor swap issue is a real one.

See Lastly's post, which addresses this quite handily.

[and into]

@johlstei: The problem is degenerate but beneficial tactics with spell-based buffs, and there have been a few different, well-argued suggestions for how to solve it:

1.) Make constant recasting impossible. Reduce duration drastically, possibly also do stuff with glow, noise, etc., so that keeping up several charms outside of combat even just semi-regularly would be a net loss. (what Galefury essentially recommended early in the thread)
2.) Deal with the charms one-by-one, introduce some kind of cost or drawback, which moves charms into more of a form-like design space.
3.) Remove lots if not all charms. Kind of draconian measure, and it would end up removing lots of what makes many backgrounds different. (Ice elementalist would play very differently without oz's armor, for example.)
4.) Permabuff.

Note that many of these approaches could be combined in various ways, and aren't mutually exclusive or anything. But if we were to use approach number 4 as the solution to most, if not all, spell-based buffs, yes, I do think this would introduce balance problems. I didn't think rMsl/dMsl becoming permanent would affect much, but it does. I agree with Galefury (and Lasty it seems) on this point.

Turns are very valuable in combat, so the time it takes to put up some buffs in combat situations is non-negligible. A single good buff can be applied in every meaningful encounter, but three or four buffs? You *can* cast them all, but I struggle to think of real, practical encounters in which putting up all those buffs in the midst of combat would make me feel safer than using those turns to actually kill the stuff. The more (truly) dangerous an encounter is, the more quickly you want it to be over, either by dispatching the enemy or by getting the hell away. Activating multiple buffs uses up time that isn't being used to summon allies, to reposition, to throw down !fog or use other non-buffing consumables, etc., and that's not even counting stuff that directly damages the enemy in question. In the late game (which is the time where non-Trog characters of all stripes will have had the opportunity to get multiple spell buffs), most truly dangerous enemies can drastically alter the conditions of combat based on their abilities, either with summoning, or hexing you, or doing even stranger things like messing with your location or blinking crap all around you. Giving these sorts of enemies extra turns so that you can maximize your EV, AC, etc., may look nice on paper, but it is probably going to force you into panicked retreats far more often than not.

So yes, again, the problem is casting spells outside of combat that give a meaningful advantage (cf. summons before the re-balancing, and permanent allies to this day, particularly necromancy). In this sense, being able to have three or four buffs active more or less constantly would be quite extreme power creep, even if it would diminish the tedium built in to the fact that you currently can pseudo-permabuff yourself if you literally hate fun and play DCSS as if it were a World Tournament Poker Championship, or something. (Note that I'm not saying it is not a problem that you can pseudo-permabuff yourself already; the question is how one fixes that problem without screwing things up in other ways.)

I will also point out that while the MP cost will not matter as much to characters that just pick up a few charms spells and don't use MP for much else, the MP cost of putting up multiple spell buffs does matter precisely to those characters who can most easily pick up charms (those already built for using magic to kill stuff). Again, casting one buff, maybe two, isn't going to drain so much of your MP. But trying to kill something that was worrisome enough that you buffed up, with only half your MP, and you've let this enemy (or these enemies) sit in your LOS for 3 or 4 turns, during which time you have not actually done anything that directly ends the encounter.... That's another story.

So I think something along the lines of what Lasty has suggested, in the way of balance, would be important if we were going to go with the permabuff approach.

By contrast, Galefury's suggestion is the simplest, and could be accomplished with existing mechanics, but may not *completely* solve the problem, and would be a nerf. However, it seems unlikely that the problem can really be addressed while maintaining exactly the same balance that exists currently, with any of the methods summarized above.

[crate]

Some options include:
* Have each active charm slow down your MP regeneration.

I don't think this is a very good option; see what happened with slow healing mut not too far back. Slow regeneration of any sort is not a very effective debuff as long as it's still >0, since during the majority of a crawl game it just means you need to rest closer to stairs (which is, admittedly, a thing you should already be doing in many cases). This isn't a thing that should be even more encouraged, imo.

I think the goal is fine (at least at first glance it seems much less abusable than similar proposals made in GDD in the past), just wanted to comment on this.

So yes, again, the problem is casting spells outside of combat that give a meaningful advantage (cf. summons before the re-balancing)

FWIW this is still entirely possible; the summon cap did absolutely nothing to affect the feasability of walking around with a pre-summoned army. All it did was limit the number of summons you can have in said army from each particular spell. (Similarly summons not attacking out-of-los didn't change anything about the ability to create a permanent summon army, just reduced its effectiveness.) (see: viewtopic.php?f=8&t=14084 )

[johlstei]

I know the problems that exist with my "solution"(namely it's the behavior that charms already have), and the problem with just making them permanent and forgetting about them. (I've read all of the old threads on this for the past ~2 years or so.) It's not really intended to be a solution at all - more of a holdover until the real solution so that one can use charms without hating oneself for not picking Trog. It's not permabuffs though - you're still bound by spell failure, mana, and (I guess) hunger. It's just porcelain around what's already there. It doesn't make anything worse and it makes some things better. I agree, there are a number of good solutions to the problems of charms in general, at least one of which will likely be implemented in the future, but a simple macro to auto-recast could easily happen right now (assuming I do the dev work - I'm obviously not taking that for granted) and other than ui, genuinely change nothing. This problem has existed for a long time, and a lot of solutions have been proposed, I don't think the (quite )difficult design problem needs to continue preempting the simple ui one.

I literally hate fun and play DCSS as if it were a World Tournament Poker Championship, and I'd like to be able to play IE without murdering my fingers, at least in the interim until a real solution is decided upon. (I can probably find a ttyrec of a game where I actually kept up old swiftness and old repel missiles the whole time I explored, not that it really says anything positive about me. I didn't have it in my heart to throw in stoneskin.) It's not a fix, it's accepting that this is what we have in the game right now and that there is no need to preserve the "balance" the spells have currently through tedium - the spells are imbalanced right now, at some point they will be fixed, can't we accept that and automate the no-brainer decision to recast them out of combat until that's no longer beneficial?

[ebering]

I'm willing to bet that something closely approximating your proposal can be achieved currently with some sufficiently advanced clua magic, to save your fingers.

[Siegurt]

Fwiw, there seems to be a perception that my suggestion is to add a drawback to all charms such that you wouldn't want them on all the time. Actually that is one of the two ways I would redo charms spells, and acutally by preference the less used method. My first preference is that most charms just be b redesigned so that it is not useful to travel with them, that they are tied roughly to the locale in which they are cast, this allows you to cast them just out of los of a known fight, but makes it not useful to simply explore wroth them on.

[KoboldLord]

Fwiw, there seems to be a perception that my suggestion is to add a drawback to all charms such that you wouldn't want them on all the time. Actually that is one of the two ways I would redo charms spells, and acutally by preference the less used method. My first preference is that most charms just be b redesigned so that it is not useful to travel with them, that they are tied roughly to the locale in which they are cast, this allows you to cast them just out of los of a known fight, but makes it not useful to simply explore wroth them on.

Does anybody actually like Leda's Liquefaction and want that dynamic applied to all charms? Maybe it's just me, but even with pre-existing investments in earth and hexes, I'd rather have four empty spell slots than use a spell that inflicts a penalty to an area and makes it really annoying for me to move around.

[Siegurt]

Well, ledas suffers from having a drawback and moving with you (as well as causing some interface annoyances)

My version of ledas would put liquified ground in a circle and leave it in place. That ground would stay liquified even when you walked away and would dry up quickly enough that it wouldn't really work to kite things back to it.

Probably it wouldn't cover every square in its radius and the number of squares and how long they stayed liquified would be spell power dependant

[Quazifuji]

My objection to the permabuff system is that I think charms are actually interesting when you're faced with the decision of whether or not to cast them in combat. The opportunity that and_into described creates interesting situations, in my opinion. Deciding whether you're better off casting fireball or reapplying phase shift is a good type of decision to have in the game. Meanwhile, as mentioned, the permabuff system basically turns charms into a passive skill similar to armour or dodging that lets you trade experience and spell levels for a variety of different bonuses depending on what books you've found, which I find less interesting.

One simple way to keep the decision involved in casting charms in combat but remove the optimality of keeping all your charms constantly up while exploring: make it so charms expire quickly (maybe 3-5 turns) when no enemies are in LOS (there was a similar pitch for summons recently that I remember getting a good response). So it would no loner be feasible to keep all your charms up permanently (would consume mana faster than it regenerates when no enemies are around), but charms would still have their current effects when used when a big fight is right around the corner or an enemy's in LOS and you need to decide if the buff is worth the turn and mana (both cases in which there's an actual decision to be made).

[KoboldLord]

My objection to the permabuff system is that I think charms are actually interesting when you're faced with the decision of whether or not to cast them in combat.

That's really where I disagree with you. There are three circumstances where you might consider casting a buff spell in combat. One is where the buff is obviously favorable, such as Haste. You should do obviously favorable things. Another is where the buff is obviously unfavorable, such as Stoneskin. You should avoid doing these sorts of buffs in combat whether you're allowed to do them out of combat or not.The third is where the question is too close to call, in which case you should refer to a tool like fsim or a set of spreadsheets to make the call. Needing to refer to spreadsheets in the middle of play is bad.

It is possible to have a buff system where the choice of whether to activate a buff is interesting and meaningful, but the mp system is a poor fit because mp recharges itself through the passage of game time, which is not an actual cost that has meaning in reasonable circumstances. This particular vision of buff design works better as a set of consumable items or divine invocations, because item stacks and god piety are both things that can run out, and both can only be restocked by continuing to play the game.

It is certainly possible to force buff spells into the mold of your vision, probably by involving an additional resource somehow that can only be restocked through gaining xp or exploring tiles, but in that case, why are we still using the overloaded spell system for the job?

[and into]

Which is why it is a good thing that casting buffs takes time, which is a valuable resource in real fights, where you do not have perfect information. If you literally cast haste every time it meaningfully helped you ("was favorable"), as opposed to a less clear-cut but better metric ("when it is called for"), then you would be less, not more, likely to win the game. That's because you'd be casting haste too early in lots of fights, against something you could take down quickly and safely enough without haste, and this puts you at needless risk if combat does not go according to plan—if, say, something nastier was outside LOS and joined as your haste was expiring. Now haste is off the table, unless you want to eat some bad glow, or can spare a !cancellation and an extra turn to get your second dose of fast. Haste as a spell may well be too good, that's another contentious topic, but even the question "Should I use haste?," is not nearly as simple as you make it out to be.

Repetitive use of the same buffs (regeneration is a big offender here) is not fun, and discouraging degenerate behavior while also trying to make buffs less annoying to use well are both worth doing. But the opportunity cost of casting even very good buffs in combat that don't cause glow, such as phase shift, is not actually negligible, and this should be considered carefully before more buffs are rMsl-ified. (There are many situations in which taking one step or evoking a wand or using a consumable will make combat safer than +8 EV. Of course there are also lots of situations in which I just hit z* and take the +8 EV because I know that will be sufficient, even if it is not necessarily the best move, but in those cases I'm playing lazy.)

[Siegurt]

Indeed, I agree that the turn costs is significant if it happens while creatures are in LOS, and the MP cost can sometimes be significant if paid in combat (Particularly if your primary damage doing ability costs mana)`, and occasionally if spent immediately before combat.

Turn and MP costs aside, there's also not such a strict deliniation between "useful" and "not useful" to cast buff spells. Stoneskin, to use your example, is a bad spell to cast if there is one creature in LOS and you can obviously kill it at that range in a minimal number of turns, If there's a large number of creatures in LOS and your available methods of doing damage aren't significant at the range that the creatures are from you, spending a turn casting stoneskin while waiting for them to come to you is more valuable, particularly if you're fighting in an area where you are more likely to draw creatures to you (Say for example an axe user in spider, where retreat to a better choke point isn't an easy option.) The small bits of damage shaved by the extra AC keep you safer in that case.

Even the "ambiguous" cases as and_into points out can shift radically from more useful to less useful depending on your current positioning and what else you have available to do with that turn (It happens with more than just Haste)

Going the permabuff route with every charm is a blanket solution to something that's not really a general problem, it's not even exclusively a "Charms" problem (although that school does have the most examples of spells that exhibit the problem that needs to be addressed) (For example, Stoneskin, one of the mentioned problem spells, isn't even a charms spell)

It's a problem with a number of spells, and those spells should be fixed, and they should be fixed in ways that are specific to that spell, that make the spell useful, but not too useful, and make it so that spamming the spell out of combat isn't a productive behavior.

There are probably a number of those spells that should be removed, some that are fine and should be left alone, some that should be given the perma-buff treatment, some that should be given some sort of drawback to make it not a good idea to spam them, and in my opinion the bulk of the spells should be morphed into a different kind of spell that's not useful to spam outside of combat. Generally speaking the way I'd do this for many of those spells would be to change them from "do something to the player" to "do something to the environment"

[Galefury]

The version of charms reform I'm currently pushing is to make more charms be toggleable always-on spells, but have their chance of proccing on each triggering event be proportional to spell success rate and the degree of their effect be proportional to spellpower. Arguably some mana or time penalty is still needed. Some options include:
* Have a chance to drain 1 mp on proc, perhaps modulated by spellpower.

I like this. Having spell success modify the proc chance is a great idea and avoids a lot of the usual permabuff problems. Having a MP cost on proc is also nice. Reducing max MP is another option that would have mostly similar effects. Both reduce the amount of MP available to do other stuff. Having a cost on proc is a lot more dynamic, you only pay if the buff is actually doing something for you in a particular encounter, and you need to have enough MP left to fuel your buffs. Max MP cost is more static and predictable, you always pay and the buff is always on. Each has some potential sources of annoyance, and after thinking about it for a while I prefer the cost on proc, mostly because you would probably want to toggle buffs less often than with the static cost.

Whether the proc cost should depend on the spell and whether it should scale somehow is a difficult question. Making it the same for all spells means there is an incentive to get rid of weak low level buffs as the game progresses, because you develop more efficient uses for your MP. Also it is very easy to communicate. On the other hand it may be a bit difficult to balance and reduces the design space. Scaling with spellpower could be good, especially because many buff effects currently do not scale with spell power and cost would be an easy way to make spellpower do something. But it does make the system a lot less transparent. I think it's reasonable to first start thinking about how to change each spell with no scaling and identical proc cost for all spells. If it turns out it's too hard, add different costs, and finally scaling if it's really necessary.

Another problem is defining what counts as using a buff. Using the example of RMsl, did you use it even if you get hit? Did you use it if you could have dodged the attack without having RMsl up? An even more complicated spell is Ring of Flames. It modifies your resistances, makes flame clouds around you, makes you immune to flame clouds, and enhances your fire spells. Does it "proc" when you cast a fire spell? Does it proc if you get hit by a fire attack? What about an ice attack (reduced resistance)? Does it proc if your flame clouds damage something or block a cold projectile? Or does it just cost some mana every few turns? Do you get fewer flame clouds if your spell success is low? The situation is similar for forms, and most of those you probably wouldn't even want to toggle.

Also, toggling a buff should take a turn so you don't want to do it too much in combat.

Certainly a permabuff system would be cool, but just like any other form of "fixing" charms it requires thinking about each spell individually.

[Lasty]

Another problem is defining what counts as using a buff. Using the example of RMsl, did you use it even if you get hit? Did you use it if you could have dodged the attack without having RMsl up? An even more complicated spell is Ring of Flames. It modifies your resistances, makes flame clouds around you, makes you immune to flame clouds, and enhances your fire spells. Does it "proc" when you cast a fire spell? Does it proc if you get hit by a fire attack? What about an ice attack (reduced resistance)? Does it proc if your flame clouds damage something or block a cold projectile? Or does it just cost some mana every few turns? Do you get fewer flame clouds if your spell success is low? The situation is similar for forms, and most of those you probably wouldn't even want to toggle.

I'm not sure it makes sense to make every duration spell an "always-on" spell as per this model -- does anyone currently keep Ring of Flames up at all times, or cast it for every fight? It probably is quite reasonable to leave Ring of Flames as-is; it already has a drawback that makes it dangerous to leave up constantly, and it takes so much mana to cast that it's probably not "optimal" in any sense to continually recast it outside combat.

[Hurkyl]

It's a problem with a number of spells, and those spells should be fixed, and they should be fixed in ways that are specific to that spell, that make the spell useful, but not too useful, and make it so that spamming the spell out of combat isn't a productive behavior.

There are probably a number of those spells that should be removed, some that are fine and should be left alone, some that should be given the perma-buff treatment, some that should be given some sort of drawback to make it not a good idea to spam them, and in my opinion the bulk of the spells should be morphed into a different kind of spell that's not useful to spam outside of combat. Generally speaking the way I'd do this for many of those spells would be to change them from "do something to the player" to "do something to the environment"

That's fine, but planning to implement a hypothetical ideal solution some time in the unforseeable future is not a reason to avoid implementing something now that would make things better now.

Maybe I'm wrong, but I feel like the biggest obstacle to what you suggest is that nobody the solution yet, but the biggest obstacle to making things better now is simply that it's not (commonly believed to be) the ideal solution people hope to find.

[Siegurt]

Well it is also my opinion that all the "fix all the buff spells with this one mechanism" proposals I have seen so far have been a net loss rather than an improvement.

[ragnarokchu]

How about something like this?

Put an internal CD on every single buff so you can't spam it (longer then the duration), make buffs instant cast or the 1st buff per turn is instant cast.

[Siegurt]

1. That disallows recasting during combat, and you could still wait out the timer expiration before exploring (making it more tedious to spam doesn't make it un spammable)
2. Part of the reason it is optimal to spam spells is that sometimes monsters get a turn before you do, making it instant cast wouldn't negate this pre-turn cast advantage, so it would still be optimal to spam.

[Siegurt]

That's fine, but planning to implement a hypothetical ideal solution some time in the unforseeable future is not a reason to avoid implementing something now that would make things better now.

Maybe I'm wrong, but I feel like the biggest obstacle to what you suggest is that nobody the solution yet, but the biggest obstacle to making things better now is simply that it's not (commonly believed to be) the ideal solution people hope to find.

Coming up with ideas to implement is easy. The last time this topic came up, I spawned 7 replacement spell ideas in 10 minutes. Finding a dev who is on board with choosing up said ideas is the challenging bit. (I would do it myself if I had the time, but I work 60+ hours a week already)

The problem with implementing a halfway solution is that it takes a significant chunk of the existing pool of devs willingness and time, and results in a perception that fixing the actual problem is less important, and significantly generates attachment for the interim solution. Plus no one wants to do a bunch of work with the intention that it is ready to be replaced out of the gate.

[ragnarokchu]

1. That disallows recasting during combat, and you could still wait out the timer expiration before exploring (making it more tedious to spam doesn't make it un spammable)
2. Part of the reason it is optimal to spam spells is that sometimes monsters get a turn before you do, making it instant cast wouldn't negate this pre-turn cast advantage, so it would still be optimal to spam.

Technically all it does it allow you to have all your buffs on at the start of the fight like you would normally anyway if you were spamming. It'll be more optimal to pop it as you need it because it'll have the exact same effect if you waited and popped it 24/7. I didn't remove the spamming to have all buffs on, I just made it easy to do and part of the game normally. All the while making it more tactical during combat.

Also recasting during combat is a personal thing, I just think buffs should be more tactical and should be more wary of them being canceled.

[Siegurt]

Technically all it does it allow you to have all your buffs on at the start of the fight like you would normally anyway if you were spamming. It'll be more optimal to pop it as you need it because it'll have the exact same effect if you waited and popped it 24/7. I didn't remove the spamming to have all buffs on, I just made it easy to do and part of the game normally. All the while making it more tactical during combat.

Also recasting during combat is a personal thing, I just think buffs should be more tactical and should be more wary of them being canceled.

Perhaps I misunderstood what you intended then. Instant cast doesn't automatically give you a buff at the start of combat, because when you reveal a square with a monster in it, it may get a chance to act before you get another turn. Taking a turn and some mp to recast a buff is a tactical decision that you deprive people of by imposing a cool down timer.

I am not sure how "you now have to wait before you can recast the spell while exploring" makes it "easy to do"

If I am misunderstanding what you are suggesting, please rephrase it so I can get what you mean.

[ragnarokchu]

Technically all it does it allow you to have all your buffs on at the start of the fight like you would normally anyway if you were spamming. It'll be more optimal to pop it as you need it because it'll have the exact same effect if you waited and popped it 24/7. I didn't remove the spamming to have all buffs on, I just made it easy to do and part of the game normally. All the while making it more tactical during combat.

Also recasting during combat is a personal thing, I just think buffs should be more tactical and should be more wary of them being canceled.

Perhaps I misunderstood what you intended then. Instant cast doesn't automatically give you a buff at the start of combat, because when you reveal a square with a monster in it, it may get a chance to act before you get another turn. Taking a turn and some mp to recast a buff is a tactical decision that you deprive people of by imposing a cool down timer.

I am not sure how "you now have to wait before you can recast the spell while exploring" makes it "easy to do"

If I am misunderstanding what you are suggesting, please rephrase it so I can get what you mean.

Let me try to break it down a little more.

The monster still getting a chance to act before you get another turn still exists, it's just that while it's your move, a buff type spell is instant/cost no turn time (for 1 or any amount of arbitrary buffing spells). Kinda like using heroism from the god power. But instead of instantly being being to cast again, the spell is locked (on cooldown) until your current buff is removed (with maybe a 2-3 turns cd afterwards so you can't cast it right away but that's just me).

That's the basic premises of it, so instead of game-play being optimal by always keep my buffs on 24/7. It becomes more of I can pop all of my buffs when combat is about to start if I wish on equal terms if the buffs was already up before. But I have spend the mana cost/hunger for the fight. The cool-down is to discourage people from spamming it 24/7 because it might be off if they meet a monster all the while knowing they could have it all on anyway if they met that monster later.

It's not more fixing the problem but more making what people do normally anyway part of the gameplay and may allow for more interesting things such as monster removing buffs you mattering more during fights but that may require duration re-balancing in general for my idea. (Since it's thought out assuming buffs wouldn't last like 30 turns)

[Siegurt]

Technically all it does it allow you to have all your buffs on at the start of the fight like you would normally anyway if you were spamming. It'll be more optimal to pop it as you need it because it'll have the exact same effect if you waited and popped it 24/7. I didn't remove the spamming to have all buffs on, I just made it easy to do and part of the game normally. All the while making it more tactical during combat.

Also recasting during combat is a personal thing, I just think buffs should be more tactical and should be more wary of them being canceled.

Perhaps I misunderstood what you intended then. Instant cast doesn't automatically give you a buff at the start of combat, because when you reveal a square with a monster in it, it may get a chance to act before you get another turn. Taking a turn and some mp to recast a buff is a tactical decision that you deprive people of by imposing a cool down timer.

I am not sure how "you now have to wait before you can recast the spell while exploring" makes it "easy to do"

If I am misunderstanding what you are suggesting, please rephrase it so I can get what you mean.

Let me try to break it down a little more.

The monster still getting a chance to act before you get another turn still exists, it's just that while it's your move, a buff type spell is instant/cost no turn time (for 1 or any amount of arbitrary buffing spells). Kinda like using heroism from the god power. But instead of instantly being being to cast again, the spell is locked (on cooldown) until your current buff is removed (with maybe a 2-3 turns cd afterwards so you can't cast it right away but that's just me).

That's the basic premises of it, so instead of game-play being optimal by always keep my buffs on 24/7. It becomes more of I can pop all of my buffs when combat is about to start if I wish on equal terms if the buffs was already up before. But I have spend the mana cost/hunger for the fight. The cool-down is to discourage people from spamming it 24/7 because it might be off if they meet a monster all the while knowing they could have it all on anyway if they met that monster later.

It's not more fixing the problem but more making what people do normally anyway part of the gameplay and may allow for more interesting things such as monster removing buffs you mattering more during fights but that may require duration re-balancing in general for my idea. (Since it's thought out assuming buffs wouldn't last like 30 turns)

So my complaints are all valid. It is still technically optimal to have buffs up all the time, and possible to recast them over and over, just less convenient because you have to hit 5 every time a buff runs out. (It is perfectly fine to propose something that doesn't address the "tedious buff spamming" thing, I guess though I just don't know what this does address though, I don't know that "I can't cast all my buff spells instantly" is a percieved problem.

Also heroism isn't instant. It takes a turn, it isn't optimal to spam out of combat because it costs piety.

[ragnarokchu]

Well like I said it is to fix buff spamming outside of combat, since you can nearly have it all up during combat anyway. You can remove tedious habits by making it almost obsolete if you don't want to completely change up the system. Sometimes just only balancing around the most "meta" extreme cases of habits just makes the game less fun.

[Hurkyl]

Well like I said it is to fix buff spamming outside of combat, since you can nearly have it all up during combat anyway. You can remove tedious habits by making it almost obsolete if you don't want to completely change up the system. Sometimes just only balancing around the most "meta" extreme cases of habits just makes the game less fun.

Engaging in unfun tedious habits makes the game less fun.

Consciously avoiding good play because it's too annoying makes the game less fun.

[ragnarokchu]

Well like I said it is to fix buff spamming outside of combat, since you can nearly have it all up during combat anyway. You can remove tedious habits by making it almost obsolete if you don't want to completely change up the system. Sometimes just only balancing around the most "meta" extreme cases of habits just makes the game less fun.

Engaging in unfun tedious habits makes the game less fun.

Consciously avoiding good play because it's too annoying makes the game less fun.

So how does my suggestion point towards either? Buffing yourself has been the staple since any rpg-like game ever, and it becomes standard like in any other rpg to buff before if you want when the fight starts. The too annoying part if that you constantly buff yourself outside of combat to always "better" to have on, interrupting exploring and walking around/other activities.

We don't need to reinvent the wheel here on the buff system, I'm just suggesting making the too annoying just much less annoying and part of the gameplay experience. If anyone played any RPG ever before they would always think to buff themselves at the start of the fight anyway, it's a common normal thing people expect.

Any other suggestion keep "buff" flavor of buffing spells would either make them non-spells (passives/charms ect), channeling of some sort (the OP) or somehow only have you buff during combat anyway at no actual penalty (my suggestion) But it would near impossible to have people not "meta-game" the crap out of any nuances within a system. So might as well make the next optimal (of like what, 1% less worse) thing what people would do normally.

The only thing I can think of is having spells be a buff that drain mana over time that progressively gives you more glow and as you buff/it expires it stacks a debuff on you.

[damiac]

Well I think Lasty's version is much better than the 'glow punishment' version. It needs testing and tweaking obviously, but it seems to completely eliminate any possible 'tedious but optimal' buff behavior.

[KoboldLord]

Glow won't work as it is currently implemented. It certainly does not prevent you from having Haste up for every action taken in every remotely threatening fight. Since glow decays from the passage of time, glow buffs just mean you have to include in your explorations periodic retreats to the stairwell to clear accumulated glow. This would be a worse problem than the status quo.

Glow could work, however, if it was cleared only by either gaining xp or revealing unexplored tiles. Making that change would not be a small task, however, since glow currently shows up in a whole lot of places that would be detrimental to the game if it also served the purpose of being a meaningful check on the buff system.