Purpose of the Charms school

[goodcoolguy]

People often say the max mp solution and its variants have a problem that mp doesn't matter to a certain class of characters. There's a big problem with this argument: If it's true of max mp costs, it's even more true of the current situation. That means that this is actually a point where the max mp solutions are clearly better than the current situation or whatever incremental improvements to it are being discussed.

Another one that comes up in connection with charms and duration spells is that it's somehow bad for spells to act like equipment. This argument doesn't hold up well to basic scrutiny. Leaving aside charm and duration effects that are literally provided by equipment, e.g. rmsl, ozo's armor, which were not viewed as a particular problem until quite recently, there's the problem that there are several spells in this game that do direct damage at melee range and dozens of spells that fire either a single target direct damage projectile or penetrating bolt at monsters --exactly-- the way those pieces of equipment we all know as "weapons" do. The main difference is that the magic versions of these effects are much more fiddly from a UI perspective, have a more burdensome impact on your character in terms of equipment and skill choices, and allow you to switch among several barely different attack types you might want to use on every turn with no cost.

[dpeg]

Thread has become long, I'll just add a bit here and there.

Idea for Tactical cost: "MP = 0" - casting the buff sets your MP to 0 and keeps it there until it expires. Preferably convert all that 'lost MP' into the effect of the buff.

This is a strong clamp, but I'd say it is too strong: spells are partly about MP management. This mechanic disables that, with the slightly annoying caveat that players will do other spells (such as a Transmutation, or Mephitic Cloud, say) before the one buff they're allowed to cast. In other words, I think this change would move Charms too far away from "spells in Crawl".

By the way, it occurred to me that MP depletion has a drawback as well: there will be cases when the battle is over, but your Charm spell(s) is/are still on, causing loss of MP. This could be easily counteracted by some way to stop buffs, but I dread the interface cost. On the other hand, you have to regain MP/HP after the battle anyway, so it could be just as painless of healing off poison. I still think that the link "duration of effect" ~ "MP cost during duration" is strong and flavourful.

A recap of downside ideas that don't seem to fall victim to hyperoptimal tedium or other issues:

1. Ultra-short duration
2. (Rapid) MP-depletion on seeing enemy, infinite duration until low MP
3. Malmut Nuke: Rewrite Contam so it's now "per turn chance to temp tiered malmut" at any level of glow. Give lots of Charms immediate Contam that dissipates very soon after the buff is over.
4. Incentivise exploration between casts. Either a hard "can't cast this yet" limit or by taking the above Charm-malmut and declaring it wears off by seeing new tiles.
5. Requires kills for benefit
6. Buff lasts forever until some consciously avoidable In-Combat Condition, possibly turning buff into a nerf when triggered
7. Continuous loud noise

I don't like 4. -- this is something that evocables do, not spells, in my opinion. Option 5. is fine, but (like 1.) will only work for some spells. Option 7. is what Song of Slaying does; there has been a thread about more songs, and I think that would be cool to try out.
I am missing an option: "Casting the spell causes MP loss while the effect is active. (MP depletion)."

[VeryAngryFelid]

Other options:
1) No MP regen as long as any buff spells are active
2) No HP regen as long as any buff spells are active
3) -Tele as long as any buff spells are active
4) -Potion as long as any buff spells are active
5) -Scroll as long as any buff spells are active
6) -Invocations as long as any buff spells are active

[Psieye]

Thread has become long, I'll just add a bit here and there.

Sounds like a time for a new thread, I've filled a lot of this one with recaps.

I am missing an option: "Casting the spell causes MP loss while the effect is active. (MP depletion)."

That's 2, with the tweak to avoid "I take 2 steps then rest until full reset".

[DracheReborn]

I am missing an option: "Casting the spell causes MP loss while the effect is active. (MP depletion)."

That's 2, with the tweak to avoid "I take 2 steps then rest until full reset".

You can't actually do that right now with Infusion, since Infusion has both a duration and MP depletion on hit.

Also, the MP depletion can just be a flat cost per turn, whether or not the buff is actually activated.

[dpeg]

DracheReborn: Yes, this is what I have in mind.

[Psieye]

Also "spell only has any effect of cast in combat" (effect on kill is one way to achieve this, but not the best imho)

The face value of 'buff only does anything when in combat' hit the dead end of "I tow a popcorn along to make sure I'm 'in combat' before jumping into the big fight which could happen at any time, therefore I must be 'in combat' the whole time."

Psieye:

2. (Rapid) MP-depletion on seeing enemy, infinite duration until low MP

dpeg:

I am missing an option: "Casting the spell causes MP loss while the effect is active. (MP depletion)."

Psieye:

That's 2, with the tweak to avoid "I take 2 steps then rest until full reset".

You can't actually do that right now with Infusion, since Infusion has both a duration and MP depletion on hit.

This option was not supposed to represent Infusion.

Also, the MP depletion can just be a flat cost per turn, whether or not the buff is actually activated.

MP-depletion regardless of combat incentivises the scummy behaviour of "I cast my buff, take 2 steps, press 5 until buff wears off and I'm back at full MP, then repeat". See above post.

[Siegurt]

Also "spell only has any effect of cast in combat" (effect on kill is one way to achieve this, but not the best imho)

The face value of 'buff only does anything when in combat' hit the dead end of "I tow a popcorn along to make sure I'm 'in combat' before jumping into the big fight which could happen at any time, therefore I must be 'in combat' the whole time."

Actually i meant "the effect is only useful against things that are on screen when the spell is cast"

As an imaginary off the cuff example:

  Code:

Mirrored image:
 
Charms/translocations 5

When cast, the player creates after images of himself in the eyes of every creature that can see him, monsters may attack these duplicates instead of the player, these images are only visible to creatures that can see the player when the spell is cast.

When struck these false images disappear, and any movement on the part of the player dispels the false images immediately.


Obviously that is pretty rough being just off the top of my head, but that is the sort of idea i was meaning.

[Psieye]

Actually i meant "the effect is only useful against things that are on screen when the spell is cast"

Hmm, this sounds like "mass hex" than a buff. But maybe that is a good thing: a LoS-wide, unresistable hex which grants a relatively-tiny modifier to combat.

[Siegurt]

Actually i meant "the effect is only useful against things that are on screen when the spell is cast"

Hmm, this sounds like "mass hex" than a buff. But maybe that is a good thing: a LoS-wide, unresistable hex which grants a relatively-tiny modifier to combat.

Yes, I was really hoping for "Charms effect the environment" and "Hexes effect your enemies" as the dichotomy there, but that spell is a bad example of "effects the environment"

[Psieye]

Actually i meant "the effect is only useful against things that are on screen when the spell is cast"

Hmm, this sounds like "mass hex" than a buff. But maybe that is a good thing: a LoS-wide, unresistable hex which grants a relatively-tiny modifier to combat.

Yes, I was really hoping for "Charms effect the environment" and "Hexes effect your enemies" as the dichotomy there, but that spell is a bad example of "effects the environment"

Hmm, then how about a more literal take on 'affect the environment'? Like Zin Sanctuary or Fedhas Sunlight?

[Siegurt]

[quote="Siegurt"]
Actually i meant "the effect is only useful against things that are on screen when the spell is cast"

Hmm, this sounds like "mass hex" than a buff. But maybe that is a good thing: a LoS-wide, unresistable hex which grants a relatively-tiny modifier to combat.

Yes, I was really hoping for "Charms effect the environment" and "Hexes effect your enemies" as the dichotomy there, but that spell is a bad example of "effects the environment"

Hmm, then how about a more literal take on 'affect the environment'? Like Zin Sanctuary or Fedhas Sunlight?[/quote]
Well the problem with the literal take is (generally speaking) it tends towards luring if it doesn't have any other restrictions (for example clouds were awkward for this reason until "expires when out of los" was added, which doesnt 100% solve the problem (it is probably still technically optimal to spam clouds at the edge of your LoS until you are sure nothing will wander into them, then move a couple steps and repeat) That isn't to say that the more literal form isn't useful as an example set, only that it is incomplete. Otoh, converting "haste" into "summon torpor snail" sounds pretty awesome to me

[Nekoatl]

1) Instead of adding a fixed amount of contamination at each initial spellcast that wears down over time, change contamination to grant a larger amount of contamination at spellcast that does not wear down while the spell remains active. Grant contaminating spells an ability toggle, to turn off both the spell effect and contamination at will.

2) Add the above contamination-while-active effect to all Charms deemed problematic.

3) Next, change contamination-granted mutations from MUTCLASS_NORMAL into MUTCLASS_TEMPORARY....

This sounds really interesting and fun to me. I'd suggest the following in conjunction:

A) Make all such spells last until deactivated (either explicitly by the player or alternately by some spell-specific failure condition).
B) Rather than an ability toggle, make a second casting of the spell deactivate the effect at no cost in MP or time.

Having a contamination cost to leaving buffs active would render having to occasionally spend MP on recasts to be a redundant cost, therefore the need to recast should be removed as a matter of convenience. Removing the need to recast spells to refresh their duration opens the UI space to have recasts perform the alternate function of deactivating the effect instead. Using the spellcast UI instead of the ability UI means there's no need to choose a new ability hotkey to go along with each such spell.

Also, if the goal is to discourage players from simply leaving buffs active all the time, I'd consider having the contamination ramp up gradually rather than appear full-force from the moment of casting. This would allow players to load up on buffs right before facing a very dangerous opponent and deactivate them immediately after the fight without facing the risk of high contamination. The MP cost of casting buffs becomes less relevant as MP is recovered over time, so having the contamination cost of leaving them active become more relevant as contamination builds over time seems like a good trade to me.

[Psieye]

1) Instead of adding a fixed amount of contamination at each initial spellcast that wears down over time, change contamination to grant a larger amount of contamination at spellcast that does not wear down while the spell remains active. Grant contaminating spells an ability toggle, to turn off both the spell effect and contamination at will.

2) Add the above contamination-while-active effect to all Charms deemed problematic.

3) Next, change contamination-granted mutations from MUTCLASS_NORMAL into MUTCLASS_TEMPORARY....

This sounds really interesting and fun to me. I'd suggest the following in conjunction:

A) Make all such spells last until deactivated (either explicitly by the player or alternately by some spell-specific failure condition).
B) Rather than an ability toggle, make a second casting of the spell deactivate the effect at no cost in MP or time.

Having a contamination cost to leaving buffs active would render having to occasionally spend MP on recasts to be a redundant cost, therefore the need to recast should be removed as a matter of convenience. Removing the need to recast spells to refresh their duration opens the UI space to have recasts perform the alternate function of deactivating the effect instead. Using the spellcast UI instead of the ability UI means there's no need to choose a new ability hotkey to go along with each such spell.

Also, if the goal is to discourage players from simply leaving buffs active all the time, I'd consider having the contamination ramp up gradually rather than appear full-force from the moment of casting. This would allow players to load up on buffs right before facing a very dangerous opponent and deactivate them immediately after the fight without facing the risk of high contamination. The MP cost of casting buffs becomes less relevant as MP is recovered over time, so having the contamination cost of leaving them active become more relevant as contamination builds over time seems like a good trade to me.

I like the "just recast to switch off the otherwise infinite duration buff" idea.

Slow accumulation of contam over time sounds good on paper, but Invis show the opening given to hyper-tedium play: "I cast my buffs, take some turns, I switch off my buffs, I press 5, I repeat". That tedium is not worth the tension of "oh shit, this fight lasts longer than I planned for, do I ditch my buffs?"

By making contam a constant, front-loaded thing that immediately goes away when switching off buffs, you make the downside of buff-spam unavoidable. Also note: the contam would be doing something immediately, not just when you reach a fixed threshold.

[Implojin]

Nekoatl, Psieye: Thank you for your feedback.

Re: Tempmut charm duration:

Making the spellcast button into a toggle makes sense to me *if and only if* Charm spell durations become infinite.

If the durations do not become infinite, then you need to reserve the spellcast button for refreshing duration, which is the current behavior. I agree that using the same button to disable spells as was originally used to cast those spells would be the most intuitive UI approach, but being able to repurpose that key hinges upon those spells not needing to be recast.

I do not think that making glowCharm durations infinite would be a good idea. If you think this would be a good idea, I haven't yet seen the reasoning sufficiently supported.


Some problems that would arise in a ruleset with perma-tempmut-Charms:

1) A hapless new player casts a Charm. It begins mutating their character. The player does not realize this. Several hundred turns later they either die to a mysteriously weakened character, or they press 'A' and wonder why they have a page+ of malmutations.

This problem is effectively rate-limited by Charms wearing off after a duration. The UI aspect is not addressed by a duration component remaining attached here, but at least the negative effects would be limited to happening (for spell duration) after intentional recast.

The UI concern here might be alleviated by something like mprf(MSGCH_WARN, "BIG OBVIOUS TEMPMUT GLOW DEBUFF DANGER MESSAGE, PRESS 'A' TO CHECK YOUR MUTATIONS"); but I don't particularly like the solution of simply dumping a text message whenever a Charm is activated. Crawl's combat message log is inconsistent when it comes to the relevance of the messages it supplies: Some of them are critical not to miss, some of them are solely for flavour, and as the game is currently written it is largely up to the player to determine which is which. It feels like rcfile MSGCH_ filtering, force_more_message, and flash_screen_message, are intended to address this, but as has been mentioned in previous threads, rcfile modifications are beyond the ken of the majority of players to begin with. The risk here is that even if you print a spammy combat log message whenever a glow aura is activated, it might still be ignored by a playerbase that is accustomed to turning a blind eye to combat log dross.

A better solution to this might be to create a visual aura around the player that changes its color saturation and radius according to current glow. That would be very obvious and very hard to miss, although with this approach you risk giving players the impression that the aura affects enemies rather than the player character.


2) One of the underlying intentions of the tempmut change is to encourage careful selection of which Charm spells to activate in any given situation. Giving players the UI support to turn on every spell *and leave them on* for infinite duration risks giving players the impression that glowCharms are meant to be used that way. They aren't. My approach here is admittedly heavy-handed: By continuing to force players to recast glowCharms manually, we force them to express intent every time they want to opt-in to nuGlow.

(Notably, the annoyance factor of this behavior is at least *no worse* than current DCSS behavior. In current Crawl, there is a strong motivation to cast *every* Charm you have before *every* fight. With nuGlow, players will still probably want to cast *at least one* Charm before *nontrivial* fights, and they might want to cast *multiple* Charms before *very dangerous* fights.

This behavior isn't much of a UI improvement for veteran players, but it is still a UI improvement over the status quo that simultaneously creates gameplay depth:

Players will still have to recast their Charms many times per game, and they might still plausibly become annoyed with those repetitive recasts when compared to a theoretical "turn it on and leave it on" approach. However, they will have to cast them fewer times in total, and it will be to their benefit to think about when it's most appropriate to cast each Charm, instead of mindlessly activating their Charms every time they engage a monster.)

3) Additionally, infinite durations at their simplest implementation would effectively bypass MP costs. Removing MP costs from spells is a topic worthy of its own thread, and it has been discussed at length elsewhere. Suffice it to say that MP cost annulment was not the intent of the glowCharms changes as written above.


Re: Ramp-up contamination:

No. This design would run counter to the intent of the glowCharms changes, as Psieye has commented above. Encouraging "toggle and kill within ramp-up turns" playstyles would not result in fun gameplay.

In my opinion, if we were to adjust ramp-up time at all, it would likely be most appropriate to front load 1 guaranteed mut at each glow 'tier', with further chance of muts on subsequent activated turns, to make the choice of activating a Charm be guaranteed to produce a counterbalancing effect.

[Nekoatl]

Making the spellcast button into a toggle makes sense to me *if and only if* Charm spell durations become infinite.

If the durations do not become infinite, then you need to reserve the spellcast button for refreshing duration, which is the current behavior. I agree that using the same button to disable spells as was originally used to cast those spells would be the most intuitive UI approach, but being able to repurpose that key hinges upon those spells not needing to be recast.

Agreed. An alternate approach would be to have a single ability that could be used to deactivate any allowed spell. E.g., press a -> e -> M where 'e' is the hotkey the "End spell" ability has been assigned to and 'M' is the hotkey for the "Repel Missiles" spell.

I do not think that making glowCharm durations infinite would be a good idea. If you think this would be a good idea, I haven't yet seen the reasoning sufficiently supported.

If there's a risk of getting a new temp malmutate each turn a spell is active, and the player has the ability to end the effect early, then the optimal play is always to end the effect as soon as it is no longer needed. Leaving the spell active to randomly timeout on its own would be a trap for new players who might not look to see how to manually end the spell if they assume that it's just supposed to wear off on its own. Ideally, there would be a note about this in the spell descriptions, perhaps set aside like Vehemut notifications to draw the eye, but players may not read the descriptions unless they notice something unexpected about the spells' behavior.

1) A hapless new player casts a Charm. It begins mutating their character. The player does not realize this. Several hundred turns later they either die to a mysteriously weakened character, or they press 'A' and wonder why they have a page+ of malmutations.

If a player didn't notice they were mutating at all, that would be a serious problem. I think that regardless of whether the buffs had a limited or infinite duration, there should be a magenta TMut status displayed in the list of buffs/debuffs below the character stats. I think this, in conjunction with the existing notice of contamination, would be sufficient to draw a character's attention to investigate and learn to deactivate the buffs.

2) One of the underlying intentions of the tempmut change is to encourage careful selection of which Charm spells to activate in any given situation. Giving players the UI support to turn on every spell *and leave them on* for infinite duration risks giving players the impression that glowCharms are meant to be used that way. They aren't. My approach here is admittedly heavy-handed: By continuing to force players to recast glowCharms manually, we force them to express intent every time they want to opt-in to nuGlow.

Full disclosure: I would play at least one character where I left as many charms active as possible just for laughs. But, I get what you're saying. I just don't think this benefit outweighs the costs, unless you want to make the random timeout the only way (aside from cancellation) to remove the effects, in which case the glow could be balanced around the assumption that the buffs will remain in effect longer than needed/desired in most cases. The worst approach, in my opinion, would be to to have both an option to cancel and a random duration, because then you have to balance for the players who cancel buffs and penalize players that haven't figured out that there's an option to do that.

Also, I don't see a problem with making the penalty light enough that leaving one charm active for extended periods of time would have only a negligible glow effect, as the player would still have to choose which one charm that should be (assuming they find more than one).

(Notably, the annoyance factor of this behavior is at least *no worse* than current DCSS behavior. In current Crawl, there is a strong motivation to cast *every* Charm you have before *every* fight.

Worse, there's an advantage to recasting buffs whenever they start to wear off to avoid the pre-fight loss of MP (I've actually played characters this way, and it became *very* annoying, especially while autotravelling). Personally, I'd like to see all charms and transmutations last until explicitly cancelled, regardless of whether or not they become tied to contamination, although for this to work, there does need to be a mechanism to prevent (or at least discourage) leaving all charms active all the time.

3) Additionally, infinite durations at their simplest implementation would effectively bypass MP costs. Removing MP costs from spells is a topic worthy of its own thread, and it has been discussed at length elsewhere. Suffice it to say that MP cost annulment was not the intent of the glowCharms changes as written above.

I thought the purpose of giving the temporary mutations as a penalty was supposed to address this issue? If not, it seems more practical to just go with a reserved MP system like Path of Exile's and call it a day (Xom dislikes this idea).

Re: Ramp-up contamination:

No.

Yeah, upon further consideration, I don't like the idea either.

[Hurkyl]

Before we go too far down this path... is this fun? Self-inflicting temporary mutations to get various effects sounds more frustrating than enjoyable. (unless you're in the mood for randomness)

Also, it doesn't sound like charms; it sounds more like chaos magic or some sort of sacrifice-flavored magic (unfortunately, "wild magic" already means something in crawl). I think a reflavoring is in order if we go down this path. And the new flavor might help inspire ideas about how to redesign the school.

[Psieye]

Yes, some cool one-word name for "Faustian bargaining" sounds like the necessary flavour re-dressing to get a radical mechanical shift swallowed.

As for whether it's fun... it seems our goals have turned to "make optimal (ab)use more fun" even if that sacrifices the fun of casual usage. By carefully designing the tiers of malmut, we can restrict the harm done to honest fun.

[Hurkyl]

Ah! I think the focus on trying to prevent people from keeping a charm active all the time is a red herring -- instead, the focus should be on the problem* that people want to use their charms every fight.

I imagine if people usually want to not use a charms in fights, then it would probably follow that they won't want to keep it up all the time either.

*: I call it a problem for the sake of argument

[archaeo]

The problem with all of these ideas is that they don't really do much about the actual problem since all that's really being fiddled with is the functional duration, as no one will keep charms on all the time if they temporarily mutate you or drain your MP or give you some very dangerous status. Given that you'll still want to turn them off, you've only really added another button press to the process, and while you may succeed in preventing stacking (which goes a long way!), it won't really address the "these spells aren't much like Crawl spells" issue or the problems with alternate charms favoring melee killdudes over "caster" killdudes.

Pretend the rest of this post is a spirited defense of my previous charms-as-equipment proposal.

[Nekoatl]

Yeah, the malmutate sounds like fun in a zany sense, but might be a dealbreaker for some players, and thematically, it doesn't seem very charms-like. What about this:

1) Charms (or any self-cast spell with a duration) last until cancelled, either by the player, by a failure condition specific to the spell, or by spell collapse (explained below).
2) The casting cost of such spells is paid constantly while active by reserving that many MP.
3) If the spells have a hunger cost to the character casting them, the player's metabolism rate increases accordingly while the spells are active.
4) MP is no longer given based on level, but only by investment in Spellcasting/Invocations/Evocations.
Edit: 5) Spell power influences the potency of ongoing spells' effects so that investing in magic skills continues to provide some benefit all the way to 27.

Thus, there is no tedium of having to recast, players who wish to keep one or more charms active may do so, but must invest to get enough MP points to do be able to reserve their combined casting costs. Players who further wish to use MP for invocations, rings, and other spells, would need to invest more deeply. Dedicated melee characters would no longer have a huge MP pool for free, so the more charms they want to utilize, the more deeply they'd have to hybridize into the relevant skills. I think this preserves the costs associated with magic use in general while allowing players a streamlined method of changing their active spells only when they want to, hopefully eliminating or at least minimizing tedium from recasts.

Spell collapse:

Spells with infinite duration are occasionally checked for spell failure as if they had just been cast. If the check fails, the effect ends, and the normal consequences of a miscast occur as if they had just failed to cast the spell. This keeps spell failure chance relevant.

[duvessa]

3) If the spells have a hunger cost to the character casting them, the player's metabolism rate increases accordingly while the spells are active.

so you should toggle them off when resting?

[Nekoatl]

3) If the spells have a hunger cost to the character casting them, the player's metabolism rate increases accordingly while the spells are active.

so you should toggle them off when resting?

I imagine players progressing from not being able to cast a spell at all, to having to decide when they should have it up or not, to never having to worry about it again as they invest more deeply in magic skills. Often, especially in the early game, a little extra hunger while resting is irrelevant anyway.

[Psieye]

Well the problem with the literal take is (generally speaking) it tends towards luring if it doesn't have any other restrictions

Hmm, so an "enchant whole floor" conceptualisation. Something to think about, yes.

if people usually want to not use a charms in fights

Good observation, but any idea how to achieve this in the framework of spells?

"these spells aren't much like Crawl spells"
alternate charms favoring melee killdudes over "caster" killdudes.

Hmm, charms which explicitly favour caster killdudes over melee killdudes... it's certainly an unexplored direction. Will need to think on this.

1) Charms (or any self-cast spell with a duration) last until cancelled, either by the player, by a failure condition specific to the spell, or by spell collapse (explained below).
2) The casting cost of such spells is paid constantly while active by reserving that many MP.
3) If the spells have a hunger cost to the character casting them, the player's metabolism rate increases accordingly while the spells are active.
4) MP is no longer given based on level, but only by investment in Spellcasting/Invocations/Evocations.

Spell collapse

Spell hunger (and food in general) has been a contested topic for a long time. I think it's best we avoid treading on that ground.

2 is functionally "Permanent MP reduction" which has been discussed. You address this with 4 but let's just get to the core of it:

"A Level N spell costs N mana and takes up N spell slots. No exceptions." - Is this rule hindering more than helping? I'm going to ignore my software developer instincts screaming "wtf huge refactoring ticket" for now and say this should be discussed.

The need for MP to serve so many different uses stretches it as is, also mandating that mp costs be homogenised by spell level further distorts the situation. The tie works for Conj, Summ, Transl, Hexes, etc but Charms is an outlier. Maybe there is a way to solve Charms while sticking to this framework, but I want to know we actually do want this framework. For instance, the "permanent MP reduction" argument fails because the automatic assumption is that the amount of reduction is the casting cost - what if it required 25 MP reduction for maintenance?

[Nekoatl]

2 is functionally "Permanent MP reduction" which has been discussed.

I suggest this because I've changed my perspective from "permanent MP reduction doesn't impact dedicated melee characters because they have lots of MP anyway, so permanent MP reduction is a problem" to "permanent MP reduction doesn't impact dedicated melee characters because they have lots of MP anyway, so dedicated melee characters having lots of MP for free is a problem".

Also, I think it's important that the MP regeneration rate should not be reduced by reducing the maximum MP, because that would disproportionally hurt characters that rely primarily on casting spells. Reducing max MP rather than reserving some MP would also lessen the absorption power of spirit shields, which I'm not sure is warranted. This is why I'm trying to draw a distinction between MP reservation and max MP reduction.

"A Level N spell costs N mana and takes up N spell slots. No exceptions." - Is this rule hindering more than helping? I'm going to ignore my software developer instincts screaming "wtf huge refactoring ticket" for now and say this should be discussed.

There is an exception, in that Fireball + Delayed Fireball give a 5 slot discount. Other exceptions could be made, should the need arise, whether or not MP reservation becomes a thing.

The need for MP to serve so many different uses stretches it as is, also mandating that mp costs be homogenised by spell level further distorts the situation. The tie works for Conj, Summ, Transl, Hexes, etc but Charms is an outlier. Maybe there is a way to solve Charms while sticking to this framework, but I want to know we actually do want this framework. For instance, the "permanent MP reduction" argument fails because the automatic assumption is that the amount of reduction is the casting cost - what if it required 25 MP reduction for maintenance?

I don't see any reason why casting costs necessarily need to match spell levels, but I can't think of an example of a spell that I would recommend have a different cost, either. Similarly, there's no reason why MP reservation necessarily has to match casting cost, but it seems like the obvious choice from a balance perspective. I don't see a justification for having any spell reduce a player's available MP by 25, especially if it's only being suggested because characters get 23 free points of MP (before scaling) by levelling up.

At this point, if there's a reason why reserving MP by the casting cost of an ongoing spell doesn't seem balanced, I'm inclined to dig deeper into exactly why that is. If it's a symptom of a more fundamental problem that is merely more obvious with charms than with other schools of magic, I say let's identify and correct that problem. If it's an objection to characters investing in magic only to be able to cast some charms and leave them up indefinitely, I say if they obtained the spell and invested enough in magic skills to have 1% or less failure and no spell hunger, then they've earned the right to do that, but if you disagree, then perhaps adding some sources of cancellation from monsters would satisfy?

As for MP serving different uses, I also find that somewhat questionable. It's always bothered me that Spellcasting, Invocations, and Evocations all increase max MP, but not in conjunction with each other. I would support a push to tie MP to Spellcasting only, and remove MP from Invocations and Evocations (both the gain from skills and costs from abilities). My vine stalker monks would be sad, but they'd get over it.

[sooheon]

I love the MP reservation as cost for charms idea. It really fits my idea of charms, as semi-permanent augmentations to martial prowess, which requires skill and concentration to maintain. There seems to be two main arguments against this:

1. this would make charms OP for melees, for whom the opportunity cost of fixed mana is less.
2. charms need to be a tactical decision, having them on all the time is against the design of this game, therefore we need costs such as glow.

I believe #1 can be fixed by making charms have meaningful upkeep failure rates. Just like repel missiles and shroud of golubria failing on hit, make this the case for other charms as well, and scale failure rate steeply with spellpower. When charms are dispelled in this way, the reserved mana should obviously not be refunded. For melees to use charms meaningfully, they have to invest even more heavily than light armor wearers, or face having to constantly recast charms (at high failure), and waste time.

#2 is a more delicate problem, and I'm still ruminating over everyone else's input. My personal take is that "meaningful choice" of charms could be handled by a similar mechanism as above: piling on charms upon charms takes a toll on your concentration, affecting upkeep failure rate even for the best practitioners. This would be immediately visible on the `I` screen (or by coloring of the charms under character ui), and players would learn that having all learned charms on at all times is not viable. They would still learn to adjust charms based on the situation, but with less tedium. Recasting charms would be when either the charm has failed during combat (meaningful), or when you decide you *really* want a different one (also meaningful).

[Psieye]

my idea of charms, as semi-permanent augmentations to martial prowess, which requires skill and concentration to maintain.

This theme has come up a number of times in the Charms debate. It makes me think Okawaru's 2 buff abilities (which use the old Duration thing but with piety as limiter) could take some of the ideas we're generating.

making charms have meaningful upkeep failure rates. Just like repel missiles and shroud of golubria failing on hit, make this the case for other charms as well, and scale failure rate steeply with spellpower. When charms are dispelled in this way, the reserved mana should obviously not be refunded. For melees to use charms meaningfully, they have to invest even more heavily than light armor wearers, or face having to constantly recast charms (at high failure), and waste time.

There are two things here:
A) Charms have chance to fail on hit
B) The probability of "fail on hit" scales with spellpower

Detail B) runs into the issue of "that's only interesting in the midgame. Get to the endgame and that investment will be an inevitability." Haste was bad because "no-brainer strategic decision". In other words, if the Charm is universally good for every melee character, demanding steep XP investment is not good enough for the cost. I think heavy armour melee dudes would have been happy to put all that XP into getting Haste online even if there was no other spell in the school.

Detail A) with a stronger penalty than "you don't get an MP rebate" looks promising. Like "if you get hit while this buff is up, you get corroded or temp malmut'd". Alternatively, "Charms are guaranteed to fail if a monster does a (reaching allowed) melee attack on you".

2. charms need to be a tactical decision, having them on all the time is against the design of this game, therefore we need costs such as glow.

#2 is a more delicate problem, and I'm still ruminating over everyone else's input. My personal take is that "meaningful choice" of charms could be handled by a similar mechanism as above: piling on charms upon charms takes a toll on your concentration, affecting upkeep failure rate even for the best practitioners. This would be immediately visible on the `I` screen (or by coloring of the charms under character ui), and players would learn that having all learned charms on at all times is not viable. They would still learn to adjust charms based on the situation, but with less tedium. Recasting charms would be when either the charm has failed during combat (meaningful), or when you decide you *really* want a different one (also meaningful).

Certainly we've been trying to tackle "I stack every buff I know for every fight". The mechanism of "let me pick from a set of mutually exclusive buffs" is Transmutation territory. Hence why we tried to use revamp'd glow which was met with "is that fun?" Also note that Haste was OP even if no other Charm was used.



Hmm, I lack time right now to link the following ideas into a coherent flow so take them as parallel suggestions. For now, I will stay within "Level N spell costs N mp".

1 ---------
Old Q: "Why is MP reduction/reservation ineffective?"
A: "Because melee dudes with fat MP bars that serve no other purpose."
New Q: "Then why not use this cost for effects only casters will care about?"

"Charms effect the environment"

Example: "Enchant floor: Battlestorm"
- Until you leave floor, infinite duration for perma-reduce MP
-- reduce, not reserve. We want the MP regen penalty and all
- While floor is enchanted, whenever an enemy is hit by a conjuration, create 'battlestorm' cloud that lasts 1 turn
-- where 'conjuration' = Vehumet's list of approved 'conjuration-like' spells, not just spells which actually have Conjurations as a requirement
- Whenever you cast a conjuration, any 'battlestorm' clouds deal damage to whatever is standing in them

Example: "The Kill Radius"
- Until you leave floor, infinite duration for perma-reduce MP
- Enemies at exactly range 4 take +20% damage from conjurations
- Enemies at range 3 or less take -50% damage from conjurations

2 ---------
"Avoid buffs that you'd always want to switch on given costs, i.e. the buffs themselves have downside in their very effects"

Example: "Entropy Zone"
- Until you leave floor, infinite duration for perma-reduce MP
- Anyone taking direct damage (from any non-DoT, non-Cloud source) gets corroded

Example: "Aura of Harm"
- Silence-style shrinking radius. Initially range 4, becomes range 0 towards end.
- Enemies in aura get -4 EV and take +20-30% (spellpower dependent) extra damage from all sources.
- You get -4 EV and take +20% extra damage from all sources.

3 ---------
"Whether to put XP into charms is interesting at mid-game. If the buffs are equally applicable in the endgame, they become a no-brainer."

Example: "Leech from your future self"
- Only castable if low XL
- Cast to get +10 max HP for 500 turns
- On reaching XL=20, lose -2 max HP for each time you cast this spell

[Nekoatl]

2. charms need to be a tactical decision, having them on all the time is against the design of this game

In this respect, MP reservation has an advantage over the current MP-on-cast paradigm because with high mana regen, you can stagger a lot of charms and effectively keep more of them up, but with MP reservation you have a hard limit on the total cost that can only be increased by investing in a larger MP pool. If items are used to do this, presumably some other benefits from items must be sacrificed. If skill investment is used to do this, the character progressively transitions from a meleedude to a melee/caster hybrid. Once a character has already invested enough into magic skills to be able to maintain every charm all the time, ideally there should be some non-charm spells that would be tempting choices for the marginal skill xp investment. With enough spells, then, there should always be meaningful choices to make.

[Siegurt]

Well the problem with the literal take is (generally speaking) it tends towards luring if it doesn't have any other restrictions

Hmm, so an "enchant whole floor" conceptualisation. Something to think about, yes.

I wasn't suggesting "enchant the whole floor" that sound awful to me personally. I was suggesting that "other restrictions" are needed in addition to locale specifics to prevent luring being optimal. For example when you walk out of the area, the effect goes away (which is what clouds do) is fine.

I think designing a good charm is *hard* attempting to slap a single rule over the whole school is sloppy, and always misses something.

I think a good charm:
1. Should be useful in some situations (enough that it is worth the strategic costs) and be a problem in others (so you dont want it up all the time or in every fight)
2. Should not use waiting more as a deterrent for keeping it up all the time
3. *should* include a stategic power limitation (aka spellpower should not be irrelevant)

Some possible mechanics that *might contribute to* the creation of a good charm include: locale specific, consumes a non renewable resource, monster specific (works only on critters in los when cast), is only beneficial against certain types of critters (selecting by attack type is probably better than by, say, race)

I think perma charms (in all the variations) are awful, anything that becomes "the new baseline for character power" is bad (this is just as bad game-design wise as casting the charms over and over, it is just now automated so it's more convenient bad design), and if it has drawbacks, so you have to toggle it on and off, then it's better off having a duration anyway.

All of that makes designing charm spells pretty darn hard, designing 6-8 for a full school where they dont horribly overlap with one another is harder (maybe some can have "better" high level versions, to help with the design space)


So for a thought excersize in investgating the "only beneficial against certain types of critters" further, I think a good place to start would be to look at "what problems do we want to have charms help solve" and try to come up with a list that breaks up the game into parts, and given we have parts a,b, and c, a charm might help with a, but hurt you with c, and another might help with c, but hurt you with b (i think the best design would have them be not 1:1 reciprocal)

For example, we might have "ranged weapons", "melee weapons", and "conjurations" as a,b, and c. So there might be a clear winner for usefulness against one type of critter, and you could probably do 2 without problems, but of you start adding a third type in there and you can't cast all your charms without leaving you worse off against something. I think those actual categories might be too broad, and i think some offense oriented charms would be good to throw in there, that was just by way of example, and also the "drawback" for a spekl could include several benefit categories, or not precisely overlap "benefits" at all, but I think that a beneft category list is a place to start.

That proposal might legitamately need to be split into it's own thread.

[Hurkyl]

Haste was bad because "no-brainer strategic decision". In other words, if the Charm is universally good for every melee character, demanding steep XP investment is not good enough for the cost.

I bet that's not true. If your not zigrunning, is your CPA wearing MiFi really going to learn Haste if it's a lvl 9 triple school spell?

"that's only interesting in the midgame. Get to the endgame and that investment will be an inevitability."

This is a red herring -- it's true of every skilling decision in crawl. If charms is going to be singled out, we need a reason that makes it different, or it prompts the question of why we aren't giving things like evokables, conjurations, melee weaponry, and armor the same scrutiny.

(non-rod evokables can be excused due to the fact you can't use them frequently, but there are still rods...)

[dpeg]

Haste was bad because "no-brainer strategic decision". In other words, if the Charm is universally good for every melee character, demanding steep XP investment is not good enough for the cost.

I bet that's not true. If your not zigrunning, is your CPA wearing MiFi really going to learn Haste if it's a lvl 9 triple school spell?

You are right, this is not literally true. What people mean is that many otherwise not-magic-using characters can, and should, pick up some spells at a point. Possibly up to Haste, but before Blink etc. are worthwhile targets. (This bit about the Haste spell is slightly misleading, because there's probably enough other Haste sources. That does not make Haste spell removal bad, and I hope we can retire the wand for the subsequent release.)

"that's only interesting in the midgame. Get to the endgame and that investment will be an inevitability."

This is a red herring -- it's true of every skilling decision in crawl. If charms is going to be singled out, we need a reason that makes it different, or it prompts the question of why we aren't giving things like evokables, conjurations, melee weaponry, and armor the same scrutiny.

The huge difference between Charms and, say, Conjurations is easy to see: if you pick up conjurations then spellpower matters a lot, and "just a bit of Conjurations" (e.g. being barely able to case these spells) is *much* less of help than "just a bit of Charms". If you're a melee-oriented character, then you get a much bigger, sudden power boost from enabling certain spells than from other skill investments.

To be more specific: if I have a meleeing character, then I'll improve my total performance only a tiny bit if I learn low-level conjurations. I'll improve my character by training some more of Armour/Dodging/weapon, but this is *continuous* process. By picking up Reflect Missiles, on the other hand, I can suddenly improve performance in one big (discrete) step.

That does not imply Charm spells are bad, but I think it's underlying a lot of the discussion here. In particular, it works this way because success and spell power matter much less for out-of-combat spells than for spells you have to cast during battle.

[Siegurt]

Haste was bad because "no-brainer strategic decision". In other words, if the Charm is universally good for every melee character, demanding steep XP investment is not good enough for the cost.

I bet that's not true. If your not zigrunning, is your CPA wearing MiFi really going to learn Haste if it's a lvl 9 triple school spell?

You are right, this is not literally true. What people mean is that many otherwise not-magic-using characters can, and should, pick up some spells at a point. Possibly up to Haste, but before Blink etc. are worthwhile targets. (This bit about the Haste spell is slightly misleading, because there's probably enough other Haste sources. That does not make Haste spell removal bad, and I hope we can retire the wand for the subsequent release.)

"that's only interesting in the midgame. Get to the endgame and that investment will be an inevitability."

This is a red herring -- it's true of every skilling decision in crawl. If charms is going to be singled out, we need a reason that makes it different, or it prompts the question of why we aren't giving things like evokables, conjurations, melee weaponry, and armor the same scrutiny.

The huge difference between Charms and, say, Conjurations is easy to see: if you pick up conjurations then spellpower matters a lot, and "just a bit of Conjurations" (e.g. being barely able to case these spells) is *much* less of help than "just a bit of Charms". If you're a melee-oriented character, then you get a much bigger, sudden power boost from enabling certain spells than from other skill investments.

To be more specific: if I have a meleeing character, then I'll improve my total performance only a tiny bit if I learn low-level conjurations. I'll improve my character by training some more of Armour/Dodging/weapon, but this is *continuous* process. By picking up Reflect Missiles, on the other hand, I can suddenly improve performance in one big (discrete) step.

That does not imply Charm spells are bad, but I think it's underlying a lot of the discussion here. In particular, it works this way because success and spell power matter much less for out-of-combat spells than for spells you have to cast during battle.

I suggested, and even wrote a patch for, haste depending on spellpower (the speed boost being small at low power level, and up to the current speed at max spell power) at the time the complaints were that it behaved differently from consumable haste making it bad, which maybe having two different types of haste is bad, i don't know, i would've rather the potions and wands go, and that haste be spell only and power dependant (rightfully reserving such a powerful effect for lots of xp investment) but as there was way too much haste i definitely agree something had to go, and without changing the spell, it was the most obvious choice.

There is no reason at all the current charm set couldn't be rewritten so that it was more xp hungry, beyond duration, the effects themselves could be made spellpower dependant (some already are)

[dpeg]

Siegurt: Yes, that too. Ultimately, we'd need to settle down and look at Charm spells (current, obsolete, and some of the proposals). Not sure if this will happen, I don't have the time for that endeaveour right now. It's a tricky question, and perhaps Charm will disappear nonetheless.

[Nekoatl]

@Siegurt: I mostly agree with everything you just posted, however:

I think it's important to distinguish between charms and ongoing spells. Song of Slaying is an excellent example of a charm spell that would not be appropriate as an ongoing spell, and Darkness is an excellent example of a non-charm spell that would be appropriate as an ongoing spell.

An MP cost should always be a downside in itself. If it's not, that's a problem to be addressed. In theory, the value sacrificed by the MP cost + any additional downsides should equal the value provided by a spell in the expected use case. If a fixed amount of MP is always best spent on keeping a specific spell up, then either the value provided by that spell is too great for the value sacrificed by paying that MP or there simply aren't enough alternative sources of value on which that MP could be spent. Additional downsides always make a spell more interesting, but it shouldn't be necessary to have one for each spell. As long as a spell isn't optimal in all situations, it doesn't need to be explicitly bad in some situations. Also, for every situation that a spell is specifically designed to not work in, there should be some alternate way of dealing with that situation to compensate for the tactical vulnerability of the spell's downside.

I also think it's a mistake to think that charms only need to compete against each other for MP allocation. If we're saying that a character is far enough along that skill xp is no longer a meaningful cost for charms magic, then it's also true that skill xp is no longer a meaningful cost for any other school of magic, and so charm spells should be competing for MP allocation with all schools of magic. For example, leaving Phase Shift (RIP) on permanently might seem like a good idea, but against floating eyes, Airstrike would offer much more value for the same MP. There should be many other examples of situations where spells other than charms would be more optimal than charms, just as there should be situations where charms are more optimal.

A player might create a theme character where they just load up as many charms as possible and leave them on all the time, but that wouldn't be optimal play, and I think that's fine. I love Crawl mostly because it's a game that allows players to choose from a wide variety of diverse character themes and play styles. Sticking with long blades on a tengu monk that finds an amazing artifact polearm on D:3 is not optimal play, but it's fun, and isn't that the most important thing? If players want to live the fantasy of a magic swordsman, I think there's nothing wrong with that, as long as it doesn't wreck the game balance for players who want to maximize their win rate.

I think the core problem with charms right now is that some of them simply don't cost enough for the power they offer. For example, on conjurers I'll often memorize Repel Missiles when I find it but not bother to skill it up at all because I can just cast it out of combat as many times as necessary, and then rest until my MP recovers before I continue exploring. Essentially, it's a powerful effect that only costs 2 levels of spell memorization. If, instead, I had to allocate 2 points of reserved MP, get the failure rate down in order to avoid it randomly collapsing on me, and invest in spellpower before the repulsion effect became significant, it would no longer be an automatic pick. Charms wouldn't be so difficult to design if there were systems in place to make ongoing spells accountable to the same limitations that one-shot spells are.

[Siegurt]

@Siegurt: I mostly agree with everything you just posted, however:

I think it's important to distinguish between charms and ongoing spells. Song of Slaying is an excellent example of a charm spell that would not be appropriate as an ongoing spell, and Darkness is an excellent example of a non-charm spell that would be appropriate as an ongoing spell.

An MP cost should always be a downside in itself. If it's not, that's a problem to be addressed. In theory, the value sacrificed by the MP cost + any additional downsides should equal the value provided by a spell in the expected use case. If a fixed amount of MP is always best spent on keeping a specific spell up, then either the value provided by that spell is too great for the value sacrificed by paying that MP or there simply aren't enough alternative sources of value on which that MP could be spent. Additional downsides always make a spell more interesting, but it shouldn't be necessary to have one for each spell. As long as a spell isn't optimal in all situations, it doesn't need to be explicitly bad in some situations. Also, for every situation that a spell is specifically designed to not work in, there should be some alternate way of dealing with that situation to compensate for the tactical vulnerability of the spell's downside.

I also think it's a mistake to think that charms only need to compete against each other for MP allocation. If we're saying that a character is far enough along that skill xp is no longer a meaningful cost for charms magic, then it's also true that skill xp is no longer a meaningful cost for any other school of magic, and so charm spells should be competing for MP allocation with all schools of magic. For example, leaving Phase Shift (RIP) on permanently might seem like a good idea, but against floating eyes, Airstrike would offer much more value for the same MP. There should be many other examples of situations where spells other than charms would be more optimal than charms, just as there should be situations where charms are more optimal.

A player might create a theme character where they just load up as many charms as possible and leave them on all the time, but that wouldn't be optimal play, and I think that's fine. I love Crawl mostly because it's a game that allows players to choose from a wide variety of diverse character themes and play styles. Sticking with long blades on a tengu monk that finds an amazing artifact polearm on D:3 is not optimal play, but it's fun, and isn't that the most important thing? If players want to live the fantasy of a magic swordsman, I think there's nothing wrong with that, as long as it doesn't wreck the game balance for players who want to maximize their win rate.

I think the core problem with charms right now is that some of them simply don't cost enough for the power they offer. For example, on conjurers I'll often memorize Repel Missiles when I find it but not bother to skill it up at all because I can just cast it out of combat as many times as necessary, and then rest until my MP recovers before I continue exploring. Essentially, it's a powerful effect that only costs 2 levels of spell memorization. If, instead, I had to allocate 2 points of reserved MP, get the failure rate down in order to avoid it randomly collapsing on me, and invest in spellpower before the repulsion effect became significant, it would no longer be an automatic pick. Charms wouldn't be so difficult to design if there were systems in place to make ongoing spells accountable to the same limitations that one-shot spells are.

If by "ongoing spell" you mean "a spell that has no duration and expires only when cancelled (voluntarily or involuntarily)" see my comment a few posts back about "perma charms" it applies whether you call them charms or ongoing spells.

[Nekoatl]

If by "ongoing spell" you mean "a spell that has no duration and expires only when cancelled (voluntarily or involuntarily)" see my comment a few posts back about "perma charms" it applies whether you call them charms or ongoing spells.

More or less. It's more precise to say that I mean a spell that a character maintains, implying ongoing effort to pay for the ongoing effect.

I think perma charms (in all the variations) are awful

I disagree with both the conclusion and the reasoning.

anything that becomes "the new baseline for character power" is bad (this is just as bad game-design wise as casting the charms over and over, it is just now automated so it's more convenient bad design)

I'm not sure exactly what you mean by this, but every interpretation I can imagine either equally condemns every piece of equipment a character might find and wear, among other core game mechanics, and/or assumes that once a spell is turned on there would never be an advantage to turning it off (which I explained is a bad assumption in my previous post).

if it has drawbacks, so you have to toggle it on and off, then it's better off having a duration anyway.

Having drawbacks doesn't mean that a spell would have to be toggled on and off, it would just mean there could be an advantage to doing so... and expiring after a duration would only be better if the following conditions were met:

* The spell would have to be worth toggling on and off at least half as frequently as it would be worth casting if it timed out instead.
* The mechanisms by which an ongoing spell were paid for would have to be somehow abusable in a similar fashion to players maintaining several cast-and-expire buffs and near max MP by perpetually staggering their spell casts.
* Ongoing spells would have to be paid for in a way that would allow players to "cheat" the MP cost for their spells as per Sif Muna's new Divine Energy ability.
* The effects of the spells would have to be functionally equivalent aside from the nature of their expiration, which would be a wasted opportunity.

[dpeg]

Alright, I've re-read the whole thread. Here are some points I collected for myself.

• I don't want permanent (or on-going) spells. If we happen to have spells with a duration, then that duration should be reasonably short (say, between 1 and 50 turns). Yes, that means that I want to modify or remove Repel/Deflect Missiles.
• I don't like MP=0 as a condition for buff spells. In general, if Charms is going to survive as a spell school, then its members should be reasonably spell-like. For me, that included MP-management, and MP=0 disables that. Similarly, I don't like downtimes, or having to enable re-casting through exploration or experience. These are mechanics that can work, but they should not be tied to spells.
• I really like Implojin's model: casting buff spells apply temp-glow, building up to temporary mutations (which can be purged with experience, as usual). It might be possible to move the current contamination mechanic for spells under this umbrella as well (and keep mutation glow disabling invisibility), but that's a separate step. (Edit: actually, temporary mutations are somewhat of a timer, but they have two big advantages: spells are related to contamination already; temporary mutations already exist. I think that's a very neat combination, that'll make it natural on buff spells.)
• Durations should be fixed (for each spell). Spell power should always affect the buff itself; this is possible for almost all of these spells (Darkness is an exception). Durations need to be short enough that you recover less MP than the spell costs. I think it is important that players cannot (and don't want to) cancel buff spells.
• If we fail to salvage Charms as a spell school, and also independently, I like archaeo's proposal to put charms as permanent effects on equipment.
• In my opinion, any definition of "buff" includes a duration during which the buff is active. I think that can be favourably systematised by doubling on the MP cost (not for all spells, but for example for Repel Missiles): you pay the spell's level worth of MP to cast the buff, and then the same amount will tick down during the duration. If MP runs out, e.g. from eye of draining or casting other spells, the buffs stops immediately. (This is like poison taking a continuous tax on HP.) Note that simply paying double MP upon casting is not the same thing, because then you'd almost never get all buffs to stop on 0 MP. This is the weakest point in the list (because it may make players carefully count MP), so it comes last.

On top of this, each spell should be looked at individually, I noted for myself:

• Regeneration should increase the regeneration rate for kills during the duration, a la the DS mutation Powered by Blood. (Shard)
• Darkness needs a shorter duration, and would profit a lot from glow. (Sandman)

There have been some interesting spell proposals in the thread, but I am done for now.

[Psieye]

A player might create a theme character where they just load up as many charms as possible and leave them on all the time, but that wouldn't be optimal play, and I think that's fine. I love Crawl mostly because it's a game that allows players to choose from a wide variety of diverse character themes and play styles. Sticking with long blades on a tengu monk that finds an amazing artifact polearm on D:3 is not optimal play, but it's fun, and isn't that the most important thing? If players want to live the fantasy of a magic swordsman, I think there's nothing wrong with that, as long as it doesn't wreck the game balance for players who want to maximize their win rate.

Issue is, there are different 'flavours' of fun. The tavern community puts more emphasis on "fun by proving competency" over "fun by roleplaying". It is not enough to think about "as long as it doesn't wreck game balance for optimisers". Emphasis is also placed on "it must not make gameplay tedious for optimisers" - which is impossible to completely satisfy but is a strong direction of influence.

want to modify or remove Repel/Deflect Missiles.

don't want permanent (or on-going) spells.
casting buff spells apply temp-glow
Durations should be fixed (for each spell). Spell power should always affect the buff itself
pay the spell's level worth of MP to cast the buff, and then the same amount will tick down during the duration.
If MP runs out, buffs stops immediately.

New RMsl proposal:
- Lasts 7 turns exactly
- Gives 1 glow (with temp malmut revamp so any level of glow can mut)
- Costs additional 2 MP after cast
- Optional: When monster attempts melee at you, RMsl immediately expires

charms as permanent effects on equipment

Add "scroll of brand armour". Since there is demand for "interesting strategics choices for XP in midgame", make these effects scale with Armour skill.

There have been some interesting spell proposals in the thread, but I am done for now.

For convenience, I've collected my proposals in this thread here:

viewtopic.php?p=295532#p295532
viewtopic.php?p=295537#p295537
viewtopic.php?p=295307#p295307
viewtopic.php?p=296153#p296153

[dpeg]

Psieye: Many thanks, for the whole thread in fact! And for injecting sanity, here and elsewhere.

[Implojin]

I have begun tinkering on a WIP branch for tempmut glow, it can be found in my github repo under the charms_glow branch here:
https://github.com/Implojin/crawl

This branch is not yet cleaned up to the point where gameplay feedback will be meaningful (and I am not yet convinced that this is the way to go over equipment-Charms, further gameplay testing after achieving a meaningful balance will be sorely needed), but the initial implementation groundwork is there. I will update this post once the branch is further along (time and interest permitting), or possibly start a new thread if someone becomes interested in hosting a later version of the branch for online playtesting.

During my local testing, I have found that the effect of low levels of spell miscast glow resulting in tempmuts is actually pretty fun to play with (and thematic!), and whether or not the charms_glow approach is taken, I think contam-as-tempmuts would be something interesting to add to the game.

Some of my current todos for the branch:

Spoiler: show

charms_glow branch todo:

1) Add an expiry timer to all perma-charms that use glow. Permaglow on permacharms just leads to
muts piling up and tedious toggle behavior, this doesn't work.
1a) Remove ABIL_DISPERSE_CHARMS once expiry timers are in place.


3) Add a Charms skill amplifier to all glowy-Charms spell effects. These spells should have their
effects increased at higher school skill, and probably also decreased at lower school skill.

4) Add the glowy-Charms effect to every Charms spell[??]. Currently, the thematic linkage doesn't
work very well -- Skald's starting book only has two "problematic" spells to which I have
attached the glow effect. This glow effect would play better, I think, if it affected *every*
Charm spell.
(The downside to this is that some spells are decently-balanced already and may not be in need of
further glow adjustment. I think this might be justifiable under the principle of increasing the
player's impetus not to cast several charms before every combat, though.)


7) Fixup interaction with:
!cancellation
Undead player races
Zin
General cleanup

[duvessa]

tempmuts as glow sounds awfully mean to new players, since they have to learn what about 20 mutations do in order to use charms in an informed way. Note that the current system only mutates you when you use charms "wrong".

[VeryAngryFelid]

tempmuts as glow sounds awfully to old players too IMHO. I cannot imagine power of charms which would make me risk having a chance of even temp teleportitis or berserkitis, especially provided I will have to train Charms to get the mutations. It is not like there are no other ways to spend XP.

[Psieye]

tempmuts as glow sounds awfully mean to new players, since they have to learn what about 20 mutations do in order to use charms in an informed way. Note that the current system only mutates you when you use charms "wrong".

Good point on new players. Something to carefully design (and document) around - e.g. maybe narrow down the list of possible (tiered) charms-malmuts.

risk having a chance of even temp teleportitis or berserkitis

Tiered malmuts-from-charms. We could have teleportitis/berserkitis as final-tier temp malmuts or even impossible to get from Charms. It's a Balance Knob question.

[VeryAngryFelid]

Tiered malmuts-from-charms. We could have teleportitis/berserkitis as final-tier temp malmuts or even impossible to get from Charms. It's a Balance Knob question.

Then we should list possible mutations in descriptions for every spell (I assume different charms will have different list). Otherwise we will have yet another reason to use spoilers.

[Psieye]

Tiered malmuts-from-charms. We could have teleportitis/berserkitis as final-tier temp malmuts or even impossible to get from Charms. It's a Balance Knob question.

Then we should list possible mutations in descriptions for every spell (I assume different charms will have different list). Otherwise we will have yet another reason to use spoilers.

Simpler to have all charms share the malmut tiers by default (some may be allowed unique malmuts). Charms descriptions should tell user to refer to "?/" to look up specifics.

[archaeo]

charms as permanent effects on equipment

Add "scroll of brand armour". Since there is demand for "interesting strategics choices for XP in midgame", make these effects scale with Armour skill.

FWIW, the devs have pretty consistently shot down any brand armour scrolls for a variety of reasons.

Personally, I'm not sure I see the utility in making charms-as-equipment effects scale with Armour skill; I'm not sure who's demanding interesting midgame decisions, but making one of the core skills for the vast majority of characters even better doesn't strike me as particularly interesting. I also haven't ever really found charms skill to have an interesting effect on the spells beyond acting as an XP sink, and I think having some of these spells as armour brands would necessarily limit them enough that floor RNG would be sufficient limitation on their use.