ToME: the Tales of Maj'Eyal

I love the concept of thaumaturgy; I've always been a sucker for "glass cannon" characters (with massive offensive potential, offset by a lack of defense / health), and the sheer variety and unpredictability of the thaumaturge is wonderful. Unfortunately the current implementation is a bit flawed, since most of the spells the thaumaturge learns are useless. I'm going to spend the first part of this post demonstrating that. Before I run down the list, here's some stats on Manathrust and Fireflash, the traditional mage's primary attack spells:

Level 50:
Manathrust: 53d21 for 25 mana
Fireflash: 430 for 58 mana

Level 70 (i.e. maxed spell school and Spellpower):
Manathrust: 73d29 for 34 mana
Fireflash: 630 for 84 mana

Let's compare that to top-level Thaumaturge spells:
Cost: anywhere from 70 to 130 mana
Ball spell damage: 100
All others: up to 10d100 damage (never more dice than 10, never larger dice than 100)
Area spells may hit upwards of 25 times (?)

In short, everything except for View and Area spells (and force beams, due to the chained knockback effect) deals crap damage for way too much mana. Thaumaturges rely on getting Area spells to be able to kill single targets in an acceptable number of rounds, but if they do get them, then they tend to be overpowered.

Here's my proposed fix, in parts:

1) Replace Area damage spells with Blast spells.
2) Change the damage calculation routine:
* Ball spell damage should be roughly quadrupled.
* Blast spells should deal damage as ball spells do.
* Bolt and beam spells should have lots of small dice instead of a few large dice (target average damage for late-game spells: 20x skill level needed to learn spell)
* View spells are fine as-is.
3) Change mana costs:
* Ball spells should be reduced to 50%
* Bolt spells should be reduced to 30%
* Beam spells should be reduced to 75%
* View spells should be increased to 120%
* Blast spells should be reduced to 50%
4) Regulate the spell generator a bit. No element should show up more than twice on a given page, and every spell type should show up at least once on a given page.
5) Allow Spell Power to improve thaumaturgy spells, recalculating their damage, cost, and failure rate as if they were higher-level.

Finally, I suggest removing Force and Inertia from the list of allowed elements, because they're just miles better than any other elements you can get. Force knocks enemies around predictably (and gets a 30% damage boost if they're knocked into walls), and Inertia slows enemies with no lower limit. Then again, I'd also advocate against Curse being able to slow enemies, so I may be biased here.

The overall goal is to make it so more spells are useful, and the player isn't left between the two extremes of "can kill anything without even trying" and "sunk 60 points into a skill that can't do anything in the late game".

Thoughts?

Generally I am in agreement. I never use bolt, beam, or ball spells (except for Force or Inertia). I use View spells for clearing out large numbers of relatively low level monsters and Area spells for high level monsters. I use Blast Wall Creation, Destruction, and Wall Destruction for terrain modification a lot. But 90% of the spells I get I never use.

The other issue is that lots of the damage types are not going to be useful against high level monsters because they are going to resist them. If you are fighting a GWoP Force and Mana are going to be the only unresisted damage (unless I forgot something). With a current Area spell that doesn't make much difference, but if they get toned down you might not have anything that can take out one.

GWoPs are sufficiently oddball that I don't think it's worth worrying about being able to kill them. You shouldn't need to be able to kill every monster in the game anyway. If you get a random quest for them you can either just skip it or deal partial damage. If monster high-resists work like player high-resists do, then you can be getting anywhere from 50% to 85% of your damage through per-hit anyway, which should be sufficient.

Speaking of the terrain modification spells, I think they shouldn't be limited to blast and area effects. Beam and ball should work just fine, bolt would work okay (embedding the monster hit into a wall), and I guess view would be kinda wonky. There's a further question of if thaumaturges should even have terrain modification spells, given their massive utility; I've had characters choose thaumaturgy from Fumblefingers specifically in the hope that they'd get an early blast wall creation spell. However, I'm inclined to leave that be for now. Heck, the original thaumaturges back in Kamband could get blast trap destruction spells.

I like Thaumaturgists a lot. Just a few musings here.

What if Thaumaturgist spells were upped in power, but they had some kind of counter that temporarily but cumulatively diminished damage and/or range. This might serve to make the class less of a one-trick pony at higher levels (into 3 or 4 trick ponies at least). Imagine the max damage for a spell is 100, counter goes to 1. Next casting of that spell within 5 turns or so would do (5-counter)/5 for 80, counter would go to 2, and the next casting would do 60. Waiting a few turns would restore max damage, but in a firefight there would be an advantage in mixing a few different spells together. Different spells could have different counters or damage reduction effects.

Another possibility to balance out more powerful Th spells would be that consecutive casts of the same spell have higher chances to fail, or to reduce sanity or something like that.

Of course, one could go the other way, and keep more modestly powered spells, but give some kind of cumulative bonus for repeated casts.

A totally different idea:
Crawl has a god that gives semi-random decks of cards with magical effects that come in themes. I could see thaumaturgists working this way too- advancing in skill opens up new realms. Lets say you start out with "fire" and "earth". You see your fire spell is a ball, but when you cast it it has a chance to switch to something else (bolt, area, etc). One thing this would do is limit the number of spell slots so there isn't some rediculous number with most of them obsolete. It would also mean if you get a nice spell like view inertia you need to be more tactical about saving it.

Say, how about this? The player gains spell shapes as they rank up in Thaumaturgy -- at level 1, you can only cast a bolt spell; at level 10 you get ball spells, at 20 beam spells, etc. You only get one spell of each shape, and its element shifts randomly every 100 turns (or whatever). The power of the spell also increases as you gain ranks in Thaumaturgy -- so you'd know that you have a bolt spell that deals 25d14 damage if unresisted; you just don't know what kind of damage it'll do until you go to cast the spell. Maybe at higher ranks you can have two spells of each shape, so if you really want to cast a beam spell at those aether hounds, you're more likely to have something they don't resist.

I have some qualms with randomizing the spell shapes, largely because that would break my "autotarget closest enemy and bolt/ball/etc. it" macros. If every time I wanted to kill a mook I had to go and find a spell of appropriate shape and cost, my gameplay would slow down immensely.

Maybe increases in level could open up both new effect realms and shapes. Realms could be gained at random, but shapes could be more deterministic (starting with bolt, then beam, ball, blast/area, view, or maybe more creative ones we could think of like chains, bouncing/reflecting, etc). Maybe the newer shapes get depleted faster or something, leaving you with 'lower level' shapes like bolts or beams.

I think it would be hard to play without knowing what element was about to come out, but I like the idea of some uncertainty in the shape of the spell. Spamming one spell over and over again gets boring, but there would have to be a balance between variety and requiring too much management.

In situations where thought is warranted, I agree with you that spamming the same spell gets boring -- but there's simply far too many cases where I just want to make an enemy go away as efficiently as possible, and I don't want to have to be reading through my spell list to do that. Some example cases:

* Townspeople preventing me from running (=> autotarget bolt/view, no more townspeople)
* Snagas dribbling into a room as they wake up (=> autotarget bolt them)
* Taking downstairs into a large open room full of mooks (=> view spells, mooks fall down)
* Standing in an antisummoning corridor, churning through mooks while waiting for the big enemies to show up.

et cetera. I simply can't see shifting spell shapes as a good idea, for this reason if no other.

What if we just made it so that you could re-roll your spell if you don't like it?

For instance, you're going for an ice mage and you get a force spell. Well, once per level, you can re-roll a number of spells to see if you can get better ones.

I was just thinking of something like this too- either re-roll a known spell, or a one-time chance to reject a newly learned spell (maybe with some adjustment that your next spell will be better).

Re-rolling doesn't really fix the problem that most spells will be useless; it just means that you're slightly less likely to not have any useful spells in the late game. I'd rather see more spells made useful, personally.

The thing that would be added is that, of course, the spell list would have to be trimmed down a bit