Make charms reserve life.

[Steel Neuron]

In order to remove the tedium of recasting, a MP reservation system has been proposed for charms and rejected for a variety of reasons, but we could just stop at the simplest: If you weren't planning to use MP, you are not paying a price by having less max MP. This makes MP reservation only matter for casters and broken on characters that are more liable to use charms in the first place, like strong melee fighters who branched into the school.

However, HP reservation is a universal price. Paying with a percentage of your max HP for the privilege of having extra regen, evasion, speed, etc is a meaningful decision and an interesting way to supplement your build in exchange for extra risk.

It can be easily justified in terms of theme (sustaining that magic takes a toll on your health), and particular charms can carry other associated effects if necessary (such as swiftness switching out temporarily when hit, for example).

[Elitist]

Having less HP lowers your regeneration rate. I'd only request that your regeneration rate be based on whatever buffs you have + your maximum HP's regen rate.

Losing HP for more speed isn't a good trade. In general, speed is so much more important than HP that experienced players consider Spriggan and Felid among the better races in the game despite having among and the absolute worst HP apts in the game, and Centaur tied generally with Troll for the very best.

For the rest of the charms, certainly. I think the ones with drawbacks built in (song of slaying/infusion/Spectral Weapon) should be exempt, but I'd love to try this for a rework to charms. I'd like the thought of having tons of passive status effects going on. Specifically, I'd like to see how ugly a permanent spectral weapon would act like.

[Psieye]

Hmm, certainly in the last big thread on Charms, reserving HP was not discussed.

What I recall from that thread was the reluctance to keep buff spells as an extension of equipment, i.e. "find it, invest some EXP+spell levels and you more-or-less permanently have it on". Yes, this was in direct conflict with opinions that simply want the flavour of "buff up a lot then jump in" regardless of the meta-consequences. Will probably come down to "someone try coding this up and be prepared to face rejection".

[Leszczynek]

First let's see how this works in Path of Exile:
1) Buffs are reserving MP, and they almost all do it by percentage, so e.g. it always takes away 25% no matter how much max MP you have.
2) Buffs/auras cost a lot so you only have at most a few on at any given time.
3) You use MP for basic attacks. Well, not exactly, but you will always use your skills on everything and those use MP.
4) The game expects you to have a good mana leech/regen so that you can sustain continuous use of skills even at very low max MP (after buff/aura reductions).
Obviously it can't work the same in Crawl because points number 3 and 4 are completely out of the question. Especially no point 3 means that MP reservation is indeed meaningless for a typical melee fighter.

So is HP reservation good instead? I'm afraid it might make many charms either 100% or 0% useful (either always or never on), but the evaluation may still vary between characters because of their total HP, Charms apt, worn armor and so on. The xp argument isn't very strong though as even old ogres could cast Haste if they really wanted in extended; unless going for a challenge run, you will be able to cast any but highest level charms if you wish to. It will ultimately depend mostly on HP lost vs value received from the buff.

I think the idea is worth testing out in an experimental branch at least, but I suppose it depends on the devs willing to do brainstorming on numbers and some bigger coding.

[StupidBerserker]

How about giving a temporal Deterioration mutation on each cast of charms?

When the mutation is removed, then the charm effect also disappears.

Deterioration has three ranks, so we could stack three charms effect every time. It's a nice limitation to avoid casting of all available spells.

[Sprucery]

One suggestion which has been also mentioned before is that there could be a limit to how many charms you can have on at a time. This limit could even include other buffs. The most drastic example: If you could only have 1 buff on at any time, you would really have to consider which one to use. E.g. haste, might or deflect missiles?

[Steel Neuron]

One suggestion which has been also mentioned before is that there could be a limit to how many charms you can have on at a time. This limit could even include other buffs. The most drastic example: If you could only have 1 buff on at any time, you would really have to consider which one to use. E.g. haste, might or deflect missiles?

I don't like this solution, it lacks a few key things:

• It renders the choice of "not using charms" inexistent. If exclusivity is the only price you pay, it's always objectively better to at least have one charm. There is no opportunity cost aside from training Charms to whatever arbitrary point you need to in order to use a powerful charm.
• There is no concept of risk reward, just opportunity cost, so the optimal "solution" would be reached easily as it would be applicable for all players, regardless of level of skill. If the best combination for a HuFi is decided to be Regen, Haste and DMsl, your choice as a player disappears and you're only grinding towards collecting the pieces. Spells aren't like items in that you make do with what you find; in a sufficiently long game you'll end up finding all of these spells and thus trivializing the choice.

With HP reservation, a skilled player may want to push their luck further and win with very low HP levels in exchange for a powerful build.

HP reservation should be tied to Charms level and could be flat or percentage based depending on the type of charm.

[ThreeInvisibleDucks]

Felids with HP-40% really benefit from Repel Missiles. Having additional HP hit would render most charms unusable.

I don't quite get what's so terrible in recasting RM if it expires. Is pressing another button apart from [o] and [Tab] really that inconvenient?

I personally enjoy melee builds with lots of support spells, I find playing them more interesting when it's not just about mashing 2 keys.

[PowerOfKaishin]

Felids with HP-40% really benefit from Repel Missiles. Having additional HP hit would render most charms unusable.

I don't quite get what's so terrible in recasting RM if it expires. Is pressing another button apart from [o] and [Tab] really that inconvenient?

I personally enjoy melee builds with lots of support spells, I find playing them more interesting when it's not just about mashing 2 keys.

Give Felids the ability to take from MP. If they want to stack on all of the buffs they can as a physical character, that's fine, since the meaningful choice for them is giving up either lots of magic reserves, or Trog.

[archaeo]

I don't quite get what's so terrible in recasting RM if it expires. Is pressing another button apart from [o] and [Tab] really that inconvenient?

I personally enjoy melee builds with lots of support spells, I find playing them more interesting when it's not just about mashing 2 keys.

It's not the fact that you're pressing another button that bothers me, it's the fact that each charm is like a piece of equipment you can put on, but it's cursed to fall off every few dozen turns, so you have to put it on before every battle. The rest of my equipment just stays on, because that's how equipment works.

This kind of mechanic is interesting when there's a tight action economy and there's a significant opportunity cost to casting a buff. Because of the size of Crawl maps and the player's ability to reset encounters with relative easy, this doesn't exist in Crawl, and the few times where it might matter, casting a buff is a mistake compared to using a consumable that will let you reset the encounter entirely.

This argument about "mashing 2 keys" is a distraction and has literally nothing to do with whether or not charms belong in the game.

[Siegurt]

I don't quite get what's so terrible in recasting RM if it expires. Is pressing another button apart from [o] and [Tab] really that inconvenient?

I personally enjoy melee builds with lots of support spells, I find playing them more interesting when it's not just about mashing 2 keys.

It's not the fact that you're pressing another button that bothers me, it's the fact that each charm is like a piece of equipment you can put on, but it's cursed to fall off every few dozen turns, so you have to put it on before every battle. The rest of my equipment just stays on, because that's how equipment works.

This kind of mechanic is interesting when there's a tight action economy and there's a significant opportunity cost to casting a buff. Because of the size of Crawl maps and the player's ability to reset encounters with relative easy, this doesn't exist in Crawl, and the few times where it might matter, casting a buff is a mistake compared to using a consumable that will let you reset the encounter entirely.

This argument about "mashing 2 keys" is a distraction and has literally nothing to do with whether or not charms belong in the game.

This is why the appropriate solution to charms is to rework each of them to only be useful to cast while in combat, where the opportunity cost matters.

[KoboldLord]

This is why the appropriate solution to charms is to rework each of them to only be useful to cast while in combat, where the opportunity cost matters.

I consider this the worst possible solution. I don't want to spend turns 'powering up' at the beginning of every single combat in the game. That is the problem I want fixed. Forcing me to spend more player-attention on micromanaging my charms is worse than just straight-up removing them.

Personally, I consider charms defaulting to always-on as the best solution. Putting xp into several other skills gives your character passive benefits, so it seems perfectly workable to do it with one more skill. Balance can be adjusted as a separate problem. Failing that, however, total removal is preferable to any of the other solutions I've heard proposed.

[archaeo]

This is why the appropriate solution to charms is to rework each of them to only be useful to cast while in combat, where the opportunity cost matters.

We have them. They're called hexes. You could arguably make some of the charms into decent hexes, if you wanted to.

Personally, I consider charms defaulting to always-on as the best solution. Putting xp into several other skills gives your character passive benefits, so it seems perfectly workable to do it with one more skill. Balance can be adjusted as a separate problem. Failing that, however, total removal is preferable to any of the other solutions I've heard proposed.

We've got always-on charms, too. It's equipment and brands. The charms that can't be made into hexes can just be made into new brands or new items. The fact that you have to sink XP into using charms would have more merit if there were any charms that were even half as good as equipment now (that a player could reasonably be expected to learn in the 3-rune game, anyway).

[Siegurt]

This is why the appropriate solution to charms is to rework each of them to only be useful to cast while in combat, where the opportunity cost matters.

We have them. They're called hexes. You could arguably make some of the charms into decent hexes, if you wanted to.

The distinction I would make between hexes and charms, is that hexes are resisted by monsters (with MR), charms are not (you cast them on yourself or your environment)

If we wanted to return to having one school called 'enchantment' and not distinguish between the two, that'd be all right too, but I personally prefer having a single specific school of spell that's resisted by MR. I think makes things clearer

[ONIchinchin]

But the tedium of recasting is a non-issue, it's a mere necessity of spells not only from Charms to activate their effects. Do you actually consider bumping enemies for XP "tedium" too? Ridiculous.

[Hellmonk]

So in other words you prefer the charms system that exists now instead of a system that has a better interface and (potentially) better balance because you enjoy mashing za zb zc out of combat every time you want to fight a yak. Suit yourself, I guess.

[ONIchinchin]

I'm not that bad of a player to unironically require buffs to kill a yak

[Hellmonk]

Cool, so you intentionally play suboptimally because it's too tedious to play optimally. Sounds like the current charms system is bad, then.

[Steel Neuron]

Cool, so you intentionally play suboptimally because it's too tedious to play optimally. Sounds like the current charms system is bad, then.

This sums up the problem perfectly.

[bel]

If you didn't require buffs to kill a yak with high probability, casting a buff is suboptimal, for sufficiently high probability, because you waste a turn and MP.

Also, very small differences in probability do not matter because the threat detection system is not that accurate.

The current charms system is indeed bad, though. The key point, in my opinion, is that reducing the max MP for active charms is functionally identical to the current system. It may or may not be ideal, but it is definitely no worse than the current status quo. Therefore, the only reason it hasn't been implemented, to my mind, is status quo bias.

[nago]

Cool, so you intentionally play suboptimally because it's too tedious to play optimally. Sounds like the current charms system is bad, then.

Well sounds more like it is balanced around the player instead of the game system

[Dioneo]

The best way of reducing tedium and such would probably be to redesign charms from scratch, such that there exists no effect which the player always wants active. They should all come with some significant cost or drawback which makes them all situational and at least potentially risky/detrimental if spammed recklessly. If we can't think of enough such spells to justify charms as a whole, distribute them among other spell schools or simply cut them entirely.

I'm not sure the one-size solution of hp reservation would accomplish this, certain effects will still be more valuable than hp and certain effects will almost never be worth it, which reduces the possible design space for charms. The relative costs are also very unevenly distributed among different species, given the differences in hp, making it even more difficult to ensure that charms remain reasonably balanced across the board.

[archaeo]

The current charms system is indeed bad, though. The key point, in my opinion, is that reducing the max MP for active charms is functionally identical to the current system. It may or may not be ideal, but it is definitely no worse than the current status quo. Therefore, the only reason it hasn't been implemented, to my mind, is status quo bias.

It isn't functionally identical at all. If you use MP to kill things or you're a VS, using buffs becomes significantly more costly, especially given that Sif/Veh/antimagic bite can't give you more than your max MP. If you don't need MP to kill things, the max MP cost is completely toothless. In exchange for reducing the tedium cost, you increase the imbalance between playstyles, a trade-off that doesn't seem worth the trouble of coding a new way of casting spells, and you make skills more narrowly useful, given that choosing to focus on charms necessarily limits your ability to use other spells and vice versa.

[amaril]

One problem w/ always-on charms: if you normally wear heavy armor, you are encouraged to remove armor in order to cast your charms. This is currently the case w/ rmsl, and it is particularly bad there because rmsl times out so fast at low spellpower that in order to play optimally you have to remove armor and rebuff almost every fight...

[Siegurt]

What if Charms had an XP charged timeout like the elemental evocables.

You certainly couldn't keep them up all the time, and blowing them all in a single combat would leave you without any of them for a long period, when you did face something, you'd want to use the charms you'd get the most use out of, and you wouldn't want to waste them by pre-casting them when they weren't needed.

I actually like that idea enough to make it it's own thread.... ->

[bel]

The current charms system is indeed bad, though. The key point, in my opinion, is that reducing the max MP for active charms is functionally identical to the current system. It may or may not be ideal, but it is definitely no worse than the current status quo. Therefore, the only reason it hasn't been implemented, to my mind, is status quo bias.

It isn't functionally identical at all. If you use MP to kill things or you're a VS, using buffs becomes significantly more costly, especially given that Sif/Veh/antimagic bite can't give you more than your max MP. If you don't need MP to kill things, the max MP cost is completely toothless. In exchange for reducing the tedium cost, you increase the imbalance between playstyles, a trade-off that doesn't seem worth the trouble of coding a new way of casting spells, and you make skills more narrowly useful, given that choosing to focus on charms necessarily limits your ability to use other spells and vice versa.

It is functionally identical because of the following argument.

------------------------------------------------------
Suppose the buff in question costs 5 MP. In the current system:

• If you have the buff active in a fight, you spend the 5 MP to cast it.
• If you don't have the buff active in fight, you don't spend 5 MP to cast it.

Therefore, in either case, you have the same MP during the fight as you would have based on the "active buffs reduce max MP" case. The choice of playstyles etc. are not relevant for the argument. It doesn't matter whether you cast spells or hit things with an axe, the point is that you have the same MP to use in a fight.
-------------------------------------------------------

Now consider the cases where the numbers are not exactly identical. I claim that either none of the cases are important, or they lead to degenerate behaviour. "Not important" means "you were fighting popcorn" or "your MP wasn't meaningfully constraining to your character during the fight" (the former is really a special case of the latter).

• MP regeneration during a fight: you gain so much MP during the fight that you would have gone above your (reduced) max MP. I assert without argument that this isn't important.
• Sif/Veh, channeling/MP restoration cannot go above your max MP. This means that you gained MP equal to "MP spent plus 5 MP" so that you were at or near full MP. So either you were fighting popcorn, or you were not constrained by your MP during the fight. Therefore. this case isn't important.
• Channeling. The only case where this matters is you cast a buff, then channel to full MP. I assert that this is degenerate behaviour.
• Anti-magic weapons. My guess is that you're already not using the MP for anything important, but it can be handled by tweaking if needed.
• Vine Stalker: it can take a tiny nerf. I don't feel it's really important, but I think this post is already too long.

[Jeremiah]

Make all charms work like irresistible hexes that affect all monsters in LOS.
So for example Ozocubu's Armour increases AC (and therefore reduces damage.)
Instead, make it apply an equivalent damage debuff to all enemies.
But, as it only affects enemies that were in LOS at the time it was cast, you can't have it "always on" while exploring etc, and it becomes a spell you essentially only cast in combat, like conjurations or whatever.

[Rast]

They should all come with some significant cost or drawback

Like having to spend xp in Charms, Spellcasting, and other schools?

[ONIchinchin]

Cool, so you intentionally play suboptimally because it's too tedious to play optimally. Sounds like the current charms system is bad, then.

So whenever someone intentionally refuses to use consumables, spells of any kind, Evocations, Ranged Weapons and god abilities regardless of the situation and their respective uses is a proof that they have a problem because of tedium and are bad?

[Steel Neuron]

Cool, so you intentionally play suboptimally because it's too tedious to play optimally. Sounds like the current charms system is bad, then.

So whenever someone intentionally refuses to use consumables, spells of any kind, Evocations, Ranged Weapons and god abilities regardless of the situation and their respective uses is a proof that they have a problem because of tedium and are bad?

Hey. don't choke on all that straw!

Everything you listed has a cost (even if it's only opportunity cost, unlike charms) and is not optimal to cast outside of battle, on anticipation of finding monsters. Using those things is not tedium, it is a tactical decision.

Charms are not a decision, they're a chore.

[Dioneo]

Like having to spend xp in Charms, Spellcasting, and other schools?

A more immediate cost or drawback, your xp-allocation has no bearing on your tactical decisions in combat, which I suggest that charms should have.

[archaeo]

@bel: Most of your counterarguments sound somewhat reasonable at 5 MP, but begin to look more questionable as you pile on the charms, something that I think you'd be incentivized to do if you plowed enough experience into charms to have level 5 spells. We also really haven't gotten into the meta discussion about the additional mechanical complexity created by this new spellcasting mechanic, how it plays with DDs, how it plays with antimagic weapons, how it plays with armour, etc., etc. It's a complicated change, and to say that "status quo bias" is the only reason not to make it looks incorrect.

When I was on my big food removal kick, gammafunk told me it would be a good idea to try and chart out all the stuff that would need changed. You might consider doing that for this idea if you'd like to convince the devs you're right. If you can address these concerns, maybe you'd have a winner.

@Dioneo: I think you're right that strategic opportunity costs aren't the point here, but I can't help but notice that there are more people calling for "charms drawbacks/costs" than people proposing what these new spells would look like.

In my opinion, "charms with drawbacks" will nearly always just look like hexes. You can imagine some charm-like effects that aren't hex-y; I remember somebody, maybe Siegurt, once proposed a "wind wall" that would replace rmsl/dms? But in general, I think that if we want spells you can't cast in safety/before combat, we want spells that target monsters and not the player.

I still think charms/tmut-as-equipment/jewelry/brands is the only proposal that actually addresses the various problems here, but it seems as though dpeg was the only one who ever thought that was an idea worth pursuing.

[Hellmonk]

Cool, so you intentionally play suboptimally because it's too tedious to play optimally. Sounds like the current charms system is bad, then.

So whenever someone intentionally refuses to use consumables, spells of any kind, Evocations, Ranged Weapons and god abilities regardless of the situation and their respective uses is a proof that they have a problem because of tedium and are bad?

If you had infinite potions of might and intentionally refused to quaff might while out of combat and not mighted, you would in fact be playing badly. I'm sure you can draw the conclusion wrt charms that you can cast hundreds of times per game for no meaningful cost on your own.

[Siegurt]

@bel: Most of your counterarguments sound somewhat reasonable at 5 MP, but begin to look more questionable as you pile on the charms, something that I think you'd be incentivized to do if you plowed enough experience into charms to have level 5 spells. We also really haven't gotten into the meta discussion about the additional mechanical complexity created by this new spellcasting mechanic, how it plays with DDs, how it plays with antimagic weapons, how it plays with armour, etc., etc. It's a complicated change, and to say that "status quo bias" is the only reason not to make it looks incorrect.

When I was on my big food removal kick, gammafunk told me it would be a good idea to try and chart out all the stuff that would need changed. You might consider doing that for this idea if you'd like to convince the devs you're right. If you can address these concerns, maybe you'd have a winner.

@Dioneo: I think you're right that strategic opportunity costs aren't the point here, but I can't help but notice that there are more people calling for "charms drawbacks/costs" than people proposing what these new spells would look like.

In my opinion, "charms with drawbacks" will nearly always just look like hexes. You can imagine some charm-like effects that aren't hex-y; I remember somebody, maybe Siegurt, once proposed a "wind wall" that would replace rmsl/dms? But in general, I think that if we want spells you can't cast in safety/before combat, we want spells that target monsters and not the player.

I still think charms/tmut-as-equipment/jewelry/brands is the only proposal that actually addresses the various problems here, but it seems as though dpeg was the only one who ever thought that was an idea worth pursuing.

I didn't originally propose wind wall (although i did propose a number of alternate mechanics for charm replacements) however i did champion it.

[bel]

@bel: Most of your counterarguments sound somewhat reasonable at 5 MP, but begin to look more questionable as you pile on the charms, something that I think you'd be incentivized to do if you plowed enough experience into charms to have level 5 spells. We also really haven't gotten into the meta discussion about the additional mechanical complexity created by this new spellcasting mechanic, how it plays with DDs, how it plays with antimagic weapons, how it plays with armour, etc., etc. It's a complicated change, and to say that "status quo bias" is the only reason not to make it looks incorrect.

When I was on my big food removal kick, gammafunk told me it would be a good idea to try and chart out all the stuff that would need changed. You might consider doing that for this idea if you'd like to convince the devs you're right. If you can address these concerns, maybe you'd have a winner.

Your example of food is quite unfortunate, since food still remains in the game

However, your point is reasonable. I have a question: I read somewhere that there was some sort of discussion on charms reform among devs (they seem to have removed Phase Shift but not done much else). Can I read these discussions somewhere, so I can actually address the concerns rather than try to construct straw men to knock down (the latter is much more fun, but my feeling is that it won't be as productive). My opinion is that none of the concerns are really dealbreaking, but I can't address them if I don't know what they are.

[Rast]

Like having to spend xp in Charms, Spellcasting, and other schools?

A more immediate cost or drawback, your xp-allocation has no bearing on your tactical decisions in combat, which I suggest that charms should have.

There isn't a cost or drawback to having a big axe or wearing a pile of randarts.

I agree that there should be limits or costs on charms. Just not drawbacks, for the most part.

[Dioneo]

There isn't a cost or drawback to having a big axe or wearing a pile of randarts.

There also isn't any "maintenance" required to having an axe or randart, but charms currently must be re-cast which leads to tedium if the player has no reason to not have them constantly activated. My suggestion was intended to remedy this tedium since I feel that if charms are basically identical to equipment (e.g. permanent and/or strictly beneficial), then they're probably better off simply being melded into equipment somehow, rather than being spells with all these weird (and difficult to balance) special-case rules.

@Archaeo
As I said, I'd be totally open to charms being removed if we can't think of enough of them, but Crawl already has some examples of the kinds of things I proposed. Infusion costs mana with each strike, Song of Slaying causes noise, Swiftness slows they players once it runs out, Ozocubu's Armour slows you during its effect and Spectral Weapon shares damage with the player. While these may not all be doing their job as currently implemented, I think the general idea is sound and the best way of balancing Charms in an elegant way. I'm afraid I don't understand what you mean by "hex-y" though; what does this entail and why is it bad?

[eliotn]

I just had an idea to fix regeneration: Regeneration costs a percentage amount of health to cast, and will only regenerate external hp damage after the cast. This prevents the optimal strategy from being just to cast regeneration while resting every time.

[syringe]

HP costs are good for active abilities. HP reservation only serves to nerf charms, so if shroud of golubria is worth 10 hp on average, then with HP reservation it will be worth 8 hp on average. If the idea is to have players do things like turn RMsl off when dealing with monsters who don't fire projectiles, that sounds like even more micromanagement. Scrutinizing HP or MP reservation distracts from the point that charms can be always on right now -- with an ugly twist (the interface). But yeah, the fact that you generally get more MP from XL than from skill (and get free spell levels) doesn't help the MP reservation cause.

[Siegurt]

So a cheap fix to "charms can be always on" is to make it so that they literally can't be on all the time.

The more expensive (in terms of dev time and overall redesigning things) and possibly more interesting fix is to make it so that it is literally not useful to have any charm on all the time.

Weird "discouraging" tactics like this one always have holes, caveats or awful micromanagement-y impacts that don't work and or aren't fun.

[sooheon]

Personally, I consider charms defaulting to always-on as the best solution. Putting xp into several other skills gives your character passive benefits, so it seems perfectly workable to do it with one more skill. Balance can be adjusted as a separate problem. Failing that, however, total removal is preferable to any of the other solutions I've heard proposed.

We've got always-on charms, too. It's equipment and brands. The charms that can't be made into hexes can just be made into new brands or new items. The fact that you have to sink XP into using charms would have more merit if there were any charms that were even half as good as equipment now (that a player could reasonably be expected to learn in the 3-rune game, anyway).

I agree with you that always-on charms are very similar to brands and equipment. The crucial difference is in how you gain access to them, and how much agency you feel. How good does it feel really, to see unrand centaur barding or the plutonium sword as a spriggan? Or to find a great artefact heavy armour and you haven't leveled strength enough? Isn't it nice that there are some elements, such as the Necronomicon, or the Book of Annihilations, which the player can *choose* to strive for? Better availability of high level spells is part of the reason people go Kiku or Sif or Vehumet. Moving charms into equipment just takes agency away from the player, and makes crawl even more of an RNG fest. It also buffs Oka and Trog, and nerfs Sif and other magical gods indirectly.

Charms (and tmut) should be always on with a max-mp cost, with a chance to dispel based on spellpower. Heavy armour wearers who cheat by taking their plate off just to cast will find their charms dispelling once they put in on and take actions or get hit. Those with appropriate spellpower and failure rate will get to enjoy their fun buffed characters, make decisions based on tradeoffs of reserving mana, change forms/charms for different encounters, and generally enjoy their games.

[tasonir]

Regarding archaeo's claim there aren't any charms half as good as equipment: the regeneration spell gives 1hp/turn, compared to the ego which gives .4 regen a turn for amulets, and I think randarts are the same? It may have changed but it's around there, and less than half the regen of the spell. I think that's a good balance, too - you can get some regeneration for "free" on lucky randarts, or for your amulet slot, but if you want stronger, high regen, you have to invest experience. The experience cost is the real cost of using regeneration, not the 3 mana or even the turn to cast it. In other words, the fix I'd like is to have regeneration be permanently on; it could have a chance to dispel on hit, or just be truly permanent. There's also some strategic choice - characters with very low int are basically locked out of meaningful regeneration since it requires decent spell power, although that's fairly rare as even with medium int you can use it well.

[Robotron]

Merge Charms and Evocations.

[tabstorm]

If we didn't have the current semi-smooth spell failure system, what you could do is have all charms be permanent taking up a portion of MP, to be determined by spell level. Assume you have a flat breakpoint of "You need X skill to cast this spell with your current stats and encumbrance". Then, if this ceased to be the case, the extra encumbrance could "disrupt" the charm, turning it off. This lets charms be like magical equipment, that you don't need to recast constantly, while avoiding degenerate play like armor swapping.