Magic Resistance Reform

[Aethrus]

Magic Resistance is a very different resist than any other in the game. Although all of the elemental resistances have differences, rElec has only one level, rPois affects damage over time, etc. Magic Resistance has three main differentiating factors:

1) It is far more granular than any other resistance. Magic resist has a huge continuum for how strong or weak it is, with over 100 discreet ranks

2) Counterintuitvely, the most granular of the resistances is 100% all or nothing, unlike the others. You either completely ignore an effect, or it completely affects you. A granular scale with binary results

3) it is far more inherently random than other resistances. While damage in crawl is inherently swingy, you know with rF+ that you will take so much less damage from fire. Magic Resistance you only have certainty if the chance is 100% or 0% of something affecting you. Because results are binary, if you're unlucky and get banished at 3% chance, your MR didn't help you. If you're unlucky and an orb of fire rolls max damage, your fire resistance is more helpful.

I don't mean to claim that any of these are inherently bad or have to be changed. I think that these details combined with some other elements of the game make magic resistance a less interesting mechanic for the player both as a defense and as something to overcome against monsters.

Magic resistance hits a harder breakpoint for usefulness than any other resistance. Going to Elf? Get 3 pips of MR so you don't get banished. 4 or 5 is useless, they don't grant any additional advantage. The Vaults? Check to make sure you are immune to mark, if you are, you don't need another point of MR.

Similarly, most monsters fall into one of the categories of 'easy to hex if you have been keeping your skill up', 'explicitly immune' or 'has so much MR it may as well be immune' making it a pretty binary choice of whether it's an option in a fight or not.

It seems to me that magic resistance too often creates clear no brainer decisions, in direct conflict with crawl philosophy.

Possible Solutions:
1) non-binary results: Confuse, Slow and Paralyze might have duration based on how poorly you rolled... Or perhaps better, success / failure could be a roll, but more MR straight up reduces duration of effects.

2) more fluctuation of MR / spellpower: if I am in Elf and have 0% chance of banishment, banishers become trivial as their biggest threat is eliminated. What if multiple banish attempts made me more susceptible to further attempts? Suddenly there would be some threat to letting them get multiple attempts, and a group of elves with status effects would be cause for concern rather than a group I can sleep on. There's already a precedent for affecting MR. With scrolls of vulnerability and the Enchantress. I think less drastic changes would create more non trivial choices.

There are probably beater solutions than what I've mentioned too, but I think this is a area of the game worth reexamination.

[nago]

A DE demologist with an hasted exec between him and me isn't properly what I call trivial, unless doing elf at like xl27 with extremely powerful char
And the other kind has also hellfire

[Sandman25]

It's actually hard to get MR immunity to mark, it can take 3+ MR items with some species after 2 runes. Other resistances have a hard breakpoint too, there is no point in having rElec++, rPois++ or rF++++. I am not sure I want to be slowed for 5% duration in every fight instead of having 5% chance to be slowed for standard duration.

[onmach]

Magic resistance is kind of weird. It tends to make abilities that oppose it to either crush or be entirely useless with no in between.

What if you made it so any spell/ability failure caused the target to lose magic resistance temporarily in general at a rate based on:
for player -> the spellpower and amount of sp used and divided by the number of targets affected (in the case of mass confusion or fear)
for monster -> the hit dice of the monster

That way it'd still be hard to use, say tulkima's dance on a elf blademaster, but if you had a few turns to spare and the mp to cast it like five times, you would be more likely to get something worthwhile out of it. Also a low level hex like confuse would be more viable even if your hex / magic ability was low.

On the flip side, a low MR character will be screwed the same as now, but as a high MR character, if they are shrugging one effect after another from multiple units, they would have to think twice about tabbing their way out of it.

[Berder]

I'd like to see hexes have a success rate of (enchantment power) / MR.

That way there's no point at which MR makes you actually immune. It just makes you less likely to be affected. The same goes for enemies, too.

Enchantment power could be adjusted. Let MR be adjusted to be similar to monster HP (with some variations based on the monster, eg spriggans can have more MR), and enchantment power be adjusted to be somewhat greater than conjuration damage. That way, the number of tries to hex something would on average be a bit less than the number of conjuration spells it would take to kill it.

With these changes, hexes would be more balanced relative to conjurations, and a viable way to kill things throughout the game.

[yesno]

Magic resist is like other resists in that more is better, which is not confusing for new players to figure out. The main tweak MR needs to be more intuitive is a listing on the species aptitudes page.

[Hurkyl]

That way, the number of tries to hex something would on average be a bit less than the number of conjuration spells it would take to kill it.

With these changes, hexes would be more balanced relative to conjurations, and a viable way to kill things throughout the game.

That would be absolutely terrible. Most hexes don't have effects resembling death; EH comes closest, and even that only resembles death if you're adjacent to the monster (or nearly so), wielding a short blade, and have a lot of training in Stealth and Short Blades.

[Berder]

That way, the number of tries to hex something would on average be a bit less than the number of conjuration spells it would take to kill it.

With these changes, hexes would be more balanced relative to conjurations, and a viable way to kill things throughout the game.

That would be absolutely terrible. Most hexes don't have effects resembling death; EH comes closest, and even that only resembles death if you're adjacent to the monster (or nearly so), wielding a short blade, and have a lot of training in Stealth and Short Blades.

Well, "a bit less" could be "half as many." And it could depend on the spell, so spells with lesser effect get a higher boost. Something you can EH in 2 shots on average (50% success rate) might take 4 shots of a high power Throw Flame to kill, which is similar to how it works now. It's not intended as a nerf to hexes; the spells can all be tweaked to retain their current early game balance. It's really meant as a buff to late game hexes because something with 200 MR wouldn't be impossible to hex anymore, it would just take twice as many attempts as something with 100 MR.

[duvessa]

Most hexes don't have effects resembling death

EH, confuse, petrify, and cause fear all do, and enslavement has an effect better than death.

"turn hexes into conjurations" is a pretty self-evidently stupid way to balance them but luckily that isn't what berder actually proposed

[Hurkyl]

Hexes being good escape options only resembles killing things in the sense that killing something is an effective way to escape it too.

Enslavement isn't a player spell anymore, and petrify isn't a hex.

"turn hexes into conjurations" is a pretty self-evidently stupid way to balance them but luckily that isn't what berder actually proposed

What he proposed at least sounded like a huge nerf. E.g. late even a spriggan can't reliably kill many things (e.g. giants) in the duration of a single Confuse; if it took nearly as many casts of Confuse to land the effect as it took Stone Arrows to kill something, it would be a completely unworkable method of offense, and a rather ineffective means of defense too.

[Siegurt]

It's been suggested before, but what I personally would like to see is hexes have a spellpower dependant *effect*

Something vaguely like:

Effectpower = (3d(pow)/3)-MR (This is pretty much the existing roll <= 0 is a failure):

EH: Duration = effectpower/10, Turn 1, yawns, flavor only effect, Turn 2, slow (80%), Turn 3: sleep (as per current) wakes up at end of duration, any damage or attacks against the target cancel the effect
Slow: Duration = effectpower/10, Each turn is (elapsed duration)/5+(elapsed duration) reduction in speed ( 50% speed on round 5)
Paralysis: Duration = effectpower/5, Turn 1: movement speed 80% "Your legs stiffen", action speed 100%, each successive round movement and attack rate "Your arms stiffen" decrease by 20% of the original (60% movement, and 80% action on turn 2, 40%/60% etc.) Resulting in no movement after 5 rounds and full paralysis after 6.
Confusion: Duration = effectpower/5 Chance of acting confused (Wandering in random direction) is (duration)/2+(duration)

etc.

(Note these suggestions are mostly off the top of my head and not intended as firm suggestions in any way, more just an outline of one)

[KittenInMyCerealz]

Just have 2 rolls that determine the effect
1st to determine if hit affects you at all, based on hd/spellpower/mr
2nd to determine how bad the effect is. more mr=lower duration.

this way confuse and other spells could remain a threat even in the late game.