Cross-school spellcasting is too easy


Although the central place for design discussion is ##crawl-dev on freenode, some may find it helpful to discuss requests and suggestions here first.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Monday, 31st January 2011, 05:35

Cross-school spellcasting is too easy

This is an observation about the different schools in general, and about schools with opposing elements. For most mid to end game casters, it's almost a no-brainer to pick up the same core of utility spells if you find them, including swiftness, repel missiles, sublimation, regeneration, abjuration, mephitic cloud, etc. A high level conjurer can nearly always pick up iron shot for irresistible damage, even if they only have a few spell levels in earth. High int and spellcasting essentially means that you can pick up these spells at no cost, and they're almost always worth the spell slots. In my experience, about 70% of the arsenal of any caster or hybrid is about the same by the end of the game, leading to a general lack of variety in gameplay. Restricting the ability to easily cast across many schools would force casters to make fuller use of the spells in their school, including some that are currently hardly used.

A more severe example of this relates to conflicting spell schools. An earth elementalist can fairly easily pick up flight and deflect missiles, while an air elementalist specializing in conjuration could fairly easily pick up iron shot, or even LCS with a little training (although it won't be terribly powerful). I think these sort of things should be nearly impossible. They make end-game elementalists less distinct and less dependent on the unique strengths and weaknesses of their school. As a recent example of this, I had a DEEE with 20 spells, and only three of those spells had an earth aspect at all (iron shot, LCS, and shatter). I had more air spells than anything else.

There are a few ways to approach this problem. One of them might be to nerf the spells that tend to crop up in every caster's arsenal, so they're no longer worth the unconditional experience investment. This was applied to haste, and I remember reading there were plans to nerf abjuration. However, characters with high spellcasting will still be able to cast these spells with little or no investment, provided they have the spare slots, so they might take them regardless.

I think a better solution might be to limit the spellcasting skill's effect on success rates. Currently, high spellcasting gives a character merit to cast level 4 and under spells at a decent success rate. I think there are already enough reasons to train spellcasting without this boost. The downside of this is that branching out does become harder, and more dependent on finding low level "training" spells to victory dance with. It also worsens a somewhat annoying dynamic that armored casters have, where they'll want to remove their armor to train low success rate spells. This might be fixed by capping the accuracy effect that spellcasting has on spells (so that they may only be brought to "fair" or so, without additional school specific investment).

Another idea which I've seen thrown around before is computing success rate with the minimum relevant spell school, instead of the average across spell schools, or at least changing the weighting so that especially low skills drag down the success rate. I guess another solution would also be to make spell power level matter more.

There's already a system in place to prevent the training of conflicting skills, which I think does a pretty good job. It'd probably be bad design to have a separate system that additionally gimps the spellcasting of conflicting spells, but I dunno. I can't think of anything right now that wouldn't encourage odd meta-gaming of skills to minimize penalties. But I think that somehow it should be nearly impossible for a dedicated earth elementalist to be able to cast deflect missiles.

For this message the author Cybermg has received thanks: 2
dolphin, evktalo
User avatar

Dungeon Master

Posts: 4031

Joined: Thursday, 16th December 2010, 20:37

Location: France

Post Monday, 31st January 2011, 11:25

Re: Cross-school spellcasting is too easy

Thanks for your message and interesting suggestions. I agree with the basic problem that many spells are too good and too easy to cast and that every spellcasters end up with it. Now, about the different solutions:
* Nerfing. I think some spells could be nerfed by simply making them much more dependent on spell power. Repel / Deflect missiles. Mephitic too.
* I disagree with removing spellcasting from spell success. I'd rather remove (or reduce) INT from success, and spellcasting from power.
* I like the idea of weighting the weaker school more heavily. Here is another: for multi-school spells, elemental schools weight twice as much. So deflect would be air/air/charm.
<+Grunt> You dereference an invalid pointer! Ouch! That really hurt! The game dies...

For this message the author galehar has received thanks: 2
evktalo, radzia

Vaults Vanquisher

Posts: 447

Joined: Thursday, 16th December 2010, 22:10

Post Monday, 31st January 2011, 11:41

Re: Cross-school spellcasting is too easy

a radical idea I've had for a while that would require a lot of reworking but would help with this: get rid of the exceedingly dull Charms and Conjurations schools, axing pure spells (I don't know about you, but I wouldn't be too sad to see Haste go) or reflavouring and moving them to other schools (I'd be sad about seeing such a wonderfully nifty spell as IOoD going), and balancing spell power/training accordingly

oh and this would also mean getting rid of Conjurers (they're just duller and less flexible Elementalists so I wouldn't be sad to see them go) and reworking or axing Crusaders.

this would also help with a lot of the ench split gripes

Dungeon Master

Posts: 1611

Joined: Thursday, 16th December 2010, 21:54

Post Monday, 31st January 2011, 12:48

Re: Cross-school spellcasting is too easy

I think the simplest (and most obvious) fix is definitely making spell power actually have an effect on the current no-brainer spells that everyone picks up, as galehar mentioned. Currently spell power only affects the duration of most Charms, instead of the effectiveness. ((Semi)controlled) blinks are completely unaffected by spell power, nor is summon butterflies, and probably a bunch of other spells I forget at the moment.

Snake Sneak

Posts: 115

Joined: Monday, 3rd January 2011, 23:21

Post Monday, 31st January 2011, 15:42

Re: Cross-school spellcasting is too easy

If people feel unhappy about some of these spells, how about raising their level so they're less likely to be "core" spells?

Ziggurat Zagger

Posts: 3037

Joined: Sunday, 2nd January 2011, 02:06

Post Tuesday, 1st February 2011, 01:28

Re: Cross-school spellcasting is too easy

Cybermg wrote:Restricting the ability to easily cast across many schools would force casters to make fuller use of the spells in their school, including some that are currently hardly used.

A more severe example of this relates to conflicting spell schools. An earth elementalist can fairly easily pick up flight and deflect missiles, while an air elementalist specializing in conjuration could fairly easily pick up iron shot, or even LCS with a little training (although it won't be terribly powerful). I think these sort of things should be nearly impossible. They make end-game elementalists less distinct and less dependent on the unique strengths and weaknesses of their school. As a recent example of this, I had a DEEE with 20 spells, and only three of those spells had an earth aspect at all (iron shot, LCS, and shatter). I had more air spells than anything else.


You'd also have to increase the options of all schools so they don't have 90% redundant crap.

Would you really prepare more than those three earth spells, even if you were unable to prepare spells from any other school at all? Damage is pretty much all earth does, and if you want to do something else earth won't do it for you. Once you've got your basic no-hunger popgun, a high-power emergency cannon, and an AoE prepared, there's no point in having another direct damage spell no matter what your alternatives are. If you've got Iron Shot, LCS, and Shatter, you're not going to use Sandblast again even if the alternative is leaving those spell slots unfilled.

And then on the other hand, is it really a good idea to start handing out buffs and debuffs and escape options and whatnot to every school so every caster can have a viable single-school focus? If you can get everything you can possibly need in earth without mixing in some charms or translocations, why not just victory dance that school and never touch anything else? You're actually even better off than in the current system, since you don't need to split your xp even a little bit to make all your support spells castable.

Cybermg wrote:But I think that somehow it should be nearly impossible for a dedicated earth elementalist to be able to cast deflect missiles.


Any particular reason why…? Isn't paying double good enough? Or some other arbitrary other multiplier, anyway? Starting around mid-Vaults you get a whole lot of xp to play with for every build, and you start getting the potential of the whole game opened up for your character. You might end up in Zot with similar abilities each time, but the journey up until that point is always going to be different.

jackalKnight wrote:If people feel unhappy about some of these spells, how about raising their level so they're less likely to be "core" spells?


Part of the trouble is that Haste grandfathers in all the charm spells that are lower in level than it is. Pre-nerf, you'd go for Haste even if there was nothing else in the school, and once you get it castable it doesn't matter much what particular level the lower-level buffs were. Post-nerf, Haste is still awesome and worth spending a lot of xp to get.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Tuesday, 1st February 2011, 02:08

Re: Cross-school spellcasting is too easy

KoboldLord wrote:You'd also have to increase the options of all schools so they don't have 90% redundant crap.

Would you really prepare more than those three earth spells, even if you were unable to prepare spells from any other school at all? Damage is pretty much all earth does, and if you want to do something else earth won't do it for you. Once you've got your basic no-hunger popgun, a high-power emergency cannon, and an AoE prepared, there's no point in having another direct damage spell no matter what your alternatives are. If you've got Iron Shot, LCS, and Shatter, you're not going to use Sandblast again even if the alternative is leaving those spell slots unfilled.

And then on the other hand, is it really a good idea to start handing out buffs and debuffs and escape options and whatnot to every school so every caster can have a viable single-school focus? If you can get everything you can possibly need in earth without mixing in some charms or translocations, why not just victory dance that school and never touch anything else? You're actually even better off than in the current system, since you don't need to split your xp even a little bit to make all your support spells castable.

Cybermg wrote:But I think that somehow it should be nearly impossible for a dedicated earth elementalist to be able to cast deflect missiles.

Any particular reason why…? Isn't paying double good enough? Or some other arbitrary other multiplier, anyway? Starting around mid-Vaults you get a whole lot of xp to play with for every build, and you start getting the potential of the whole game opened up for your character. You might end up in Zot with similar abilities each time, but the journey up until that point is always going to be different.

I think it's more an issue of tweaking the balance, rather than making it impossible to cast across multiple schools (my statement about deflect missiles was sort of an exaggeration), or augmenting each school so that it can stand alone. I don't expect that players can make due only with their primary school, only that they're presented with a little bit more of a choice before they select the standard set of utility spells. I would expect casters to branch out and train multiple schools, but I think it should be a little harder than the free ride it currently is now, once you're past the early game.

Going back to the deflect missiles example, I do think there should be a similar mechanic to what exists with armor and spellcasting right now - it's really hard to have it both ways. It's certainly possible that you could specialize in earth and cast the spell, it'd just take a lot of investment and should never truly be as good as if you were an air specialist. Deflect missiles is one of the capstone air spells (hell, maybe it should even be bumped up a level), and at 25 int, 20 spellcasting, 12 charms, and 3 air, it's at Very Good.

Regarding my point of diversifying within your chosen school, I probably would've looked into keeping passwall, petrification, LRD (if it were post-buff), or even Leda's to replace some of the more commonly used spells. I still probably would've had more than half of my spells spread across other schools though, but that's better than 85% of the spells being in other schools.
User avatar

Swamp Slogger

Posts: 153

Joined: Sunday, 23rd January 2011, 20:04

Post Tuesday, 1st February 2011, 08:47

Re: Cross-school spellcasting is too easy

The spell casting calculations are already complicated enough. So why not add in more to the Confusion!

Maybe after hitting Skill level 10 in a spell school it adds on an automatic Double to spell power/success and Halfs the Power/Success in the other school. It should also Decrease the Players Aptitude for the other schools and maybe even Lower the Level Cap of the opposing school to 9 (to prevent the removal of the Power/success divider).
The effects from this should be felt more in Earth/Air than Fire/Ice, and my numbers are in no way balanced.
(Fire and Ice are very similar so the player has no less of a desire to mix the 2)

But then Earth Magic would need to gain a couple of Utilitarian spells to make it a good substitute for air magic.

If all else Fails then just Go with TROG! RAH RAH RAH!

Snake Sneak

Posts: 110

Joined: Monday, 20th December 2010, 21:11

Post Tuesday, 1st February 2011, 17:13

Re: Cross-school spellcasting is too easy

What about adding other opposing schools besides the elemental schools?

This is Just an example and the actual decisions can be debated:

Air vs Earth
Charms vs. Conjuration
Evocation vs. Invocation (Or we can exempt these two since they aren't really magic schools)
Fire vs. Ice
Hexes vs. Poison
Necromancy vs Summonings (nerf to haunt, i guess)
Translocations vs Transmutations

Now you can't cherry pick the spells you want from every school, since every spell you take might mean another spell you want is going to be in an opposing school.
User avatar

Swamp Slogger

Posts: 153

Joined: Sunday, 23rd January 2011, 20:04

Post Tuesday, 1st February 2011, 17:33

Re: Cross-school spellcasting is too easy

Hexes and Charms?
Where did those come from?
I DEMAND NEW STARTING CLASSES FOR THESE SCHOOLS!

(Please don't Nerf haunt XD)

Ziggurat Zagger

Posts: 3037

Joined: Sunday, 2nd January 2011, 02:06

Post Wednesday, 2nd February 2011, 01:02

Re: Cross-school spellcasting is too easy

omndra wrote:Hexes and Charms?
Where did those come from?
I DEMAND NEW STARTING CLASSES FOR THESE SCHOOLS!

(Please don't Nerf haunt XD)


The devteam anticipated your request, and the new backgrounds for these schools are already being playtested in trunk. They're called the 'Enchanter' and the 'Crusader', respectively.

TGW

Halls Hopper

Posts: 82

Joined: Thursday, 16th December 2010, 22:14

Post Wednesday, 2nd February 2011, 01:52

Re: Cross-school spellcasting is too easy

Under the current split, though, both classes have spells from both schools.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Wednesday, 2nd February 2011, 03:16

Re: Cross-school spellcasting is too easy

omndra wrote:The spell casting calculations are already complicated enough. So why not add in more to the Confusion!

Maybe after hitting Skill level 10 in a spell school it adds on an automatic Double to spell power/success and Halfs the Power/Success in the other school. It should also Decrease the Players Aptitude for the other schools and maybe even Lower the Level Cap of the opposing school to 9 (to prevent the removal of the Power/success divider).
The effects from this should be felt more in Earth/Air than Fire/Ice, and my numbers are in no way balanced.
(Fire and Ice are very similar so the player has no less of a desire to mix the 2)

But then Earth Magic would need to gain a couple of Utilitarian spells to make it a good substitute for air magic.

If all else Fails then just Go with TROG! RAH RAH RAH!


I considered something like skill levels in a school decreasing your ability to cast the opposite school, but this just leads to a number of unintuitive ways to exploit the system. Namely, someone might avoid getting skill levels on one school or only get *just* as many as they need, so that they don't completely trash their abilities in an opposite school. It'd lead to weird long-term optimal behavior, like only casting elemental spells when necessary, to avoid penalties to the opposite aspect.

ryak wrote:What about adding other opposing schools besides the elemental schools?

This is Just an example and the actual decisions can be debated:

Air vs Earth
Charms vs. Conjuration
Evocation vs. Invocation (Or we can exempt these two since they aren't really magic schools)
Fire vs. Ice
Hexes vs. Poison
Necromancy vs Summonings (nerf to haunt, i guess)
Translocations vs Transmutations

Now you can't cherry pick the spells you want from every school, since every spell you take might mean another spell you want is going to be in an opposing school.


I also thought about this, but anything other than earth/air and fire/ice seems unintuitive, and there are a lot of new balance issues that crop up.

Blades Runner

Posts: 546

Joined: Monday, 20th December 2010, 14:25

Post Friday, 4th February 2011, 05:19

Re: Cross-school spellcasting is too easy

Might crosstraining penalties -- or better yet spell power and/or success penalties -- for tertiary, quaternary, etc schools work?

So an FE that has high conjurations and fire skills would have a small penalty on the third school, and larger penalties on the fourth school. It might be tricky to balance. This way, an FE/conjurer could pick up, say, air spells but not regeneration.

Halls Hopper

Posts: 69

Joined: Thursday, 17th March 2011, 23:31

Post Thursday, 7th April 2011, 17:39

Re: Cross-school spellcasting is too easy

Just on a general basis, I fail to see how learning more about one element/type of weapon/ability/skill could possibly conflict with the application of another.

Why would a fire elementalist be stumped with ice magic? Both are forms of magic that deal with the elements, the whole class specification only decided where the first experience points were allocated and the magic users load out of equipment. Why wouldn't a magic user strive to excel in all forms of magic (if said magic does not cross with the principles of the user) ?


A monk can effectively use a shield, sure there are penalties while wielding a weapon, and even more to unarmed, and the monk's initial stats are not developed in shields... But the monk isn't gimped simply because he is primarily trained in unarmed. Class does not define a character in this game, an earth elementalist is may be trained in Earth magic, but why would that not be the first of many steps to becoming a true elementalist capable of control over all four elements?

The whole conception of crosstraining penalties is to gimp characters and steer from a style of play or character development. There is nothing wrong with a list of 10 powerful, varied, useful spells which ever magic user tries to acquire. But there should be no problem introducing new attractive spells, and mechanics to enhance game play and encourage a wide selection of play styles. Not discourage choice.

For this message the author sigfried von murdock has received thanks: 3
roguelikedev, SinsI, The Mantis

Spider Stomper

Posts: 233

Joined: Monday, 20th December 2010, 20:58

Post Thursday, 7th April 2011, 20:18

Re: Cross-school spellcasting is too easy

Well it needs to be said. This is a symptom of the experience system. Kludge all you want the problem will persist. As long as costs for skills grow geometrically and benefits from those skills progress linearly you will get very cookie cutter characters. Of course you can always just continue to code high level over the top spells that trivialize the game to encourage specialization I suppose?

Mines Malingerer

Posts: 44

Joined: Thursday, 7th April 2011, 12:24

Post Thursday, 7th April 2011, 23:34

Re: Cross-school spellcasting is too easy

If you don't want people to grab a spell here and there, and use mainly spells from one or two schools:
1) eliminate schools like conjuration - almost every spell in it is "conjuration/fire", "conjuration/ice", etc. - it is a direct invitation to mix spells from fire and ice.
And something really needs to be done about generic "spellcasting".
2) Give each school a wider variety of spells; eliminate spells that are too similar.
Example: in summoning, make several highly specialized spells - to summon a monster that can teleport you, a monster that can place status effects, a monster that can boost your army, a monster that can dig, a dragon for you to ride over that lava lake....
No need for all those "summon imp/ice beast/mammal" that just give you a different meat shield.
3) Create secondary effects for magic schools that grow much, much stronger with very high appropriate skill - things like Ice Shield and Staff of Fire. Maybe even make them permanent,free and more drastic (so they are like Demonspawn mutations, only they are skill level dependent instead of character level dependent).
User avatar

Ziggurat Zagger

Posts: 5832

Joined: Thursday, 10th February 2011, 18:30

Post Friday, 8th April 2011, 00:44

Re: Cross-school spellcasting is too easy

Make Spellcasting a factor of all spell (plus evocations not invocations) skills.
Value equals all spell skill values / 5 max 27
Make casting from scrolls train evocations to allow non casters build up to read books.

Make ability to memorize a factor of spellcasting level (as determined above) and int.
"Be aware that a lot of people on this forum, such as mageykun and XuaXua, have a habit of making things up." - minmay a.k.a. duvessa
Did I make a lame complaint? Check for Bingo!
Totally gracious CSDC Season 2 Division 4 Champeen!

Mines Malingerer

Posts: 44

Joined: Thursday, 7th April 2011, 12:24

Post Friday, 8th April 2011, 05:34

Re: Cross-school spellcasting is too easy

XuaXua wrote:Make Spellcasting a factor of all spell (plus evocations not invocations) skills.
Value equals all spell skill values / 5 max 27
Make casting from scrolls train evocations to allow non casters build up to read books.

Make ability to memorize a factor of spellcasting level (as determined above) and int.


That would have an opposite effect (you'd want to train as many skills as possible) - although the same is true for current spellcasting.

In my previous list, I forgot the most important reason for grabbing spells all over the spectrum - you mostly get only random spellbooks
(baring god gifts), thus heavily investing in one or two schools means an extremely high risk of staying without level appropriate spells:

4) Gift new spells for training a skill("you research a new spell").

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Friday, 8th April 2011, 09:29

Re: Cross-school spellcasting is too easy

Looking back at this topic after the while since I've posted it, I've seen that some of these issues have been addressed by making spell effectiveness more dependent on spell power, or simply by straight up nerfing the worst offenders. Summon butterflies is probably the best example, and mephitic cloud, swiftness/flight are very close on the chopping block. I think this is the right direction to take things - it makes sense that high spellcasting would allow you to *cast* a spell, but whether or not that spell is any good still depends heavily on the relevant spell schools. It doesn't mean that low power spells have to be useless, just less useful than they are at higher powers.


Another thing I think would help is providing incentive (outside of the few super high level spells) for specializing within a spell school.

SinsI wrote:3) Create secondary effects for magic schools that grow much, much stronger with very high appropriate skill - things like Ice Shield and Staff of Fire. Maybe even make them permanent,free and more drastic (so they are like Demonspawn mutations, only they are skill level dependent instead of character level dependent).
4) Gift new spells for training a skill("you research a new spell").


These are examples of ideas along these lines. Oftentimes, a caster will realize that they'll never be able to cast the next level of spell in their dominant school (or there are no more spells they care about), and be forced to follow the generic route of spreading their skills. We want to provide further incentive for leveling their dominant skill, since the increase in spell power usually isn't worth it at this point. This also addresses the problem of some spell schools becoming worthless at high level.

Secondary effects seem pretty interesting - they'd be bonuses to your character that only start to take effect at the higher skill levels (example growth: power = max(skill-15,0) ). Here are some examples (numbers approximate). I guess one way to think of these is that they're analogous to passive weapon skills, which are in part supposed to provide incentive for training weapons to a higher level.

-Ice - Ice shield: +power to SH.
-Fire - power proportional chance to sticky flame enemies (on melee, or maybe even any ranged contact [chance would have to be reduced])
-Earth - +power/2 to AC, or power proportional chance to petrify.
-Air - power proportional chance to electrocute, and/or chance of missile deflection
-Tmut - Increased resistance to forced mutation.
-Tloc - Reduced tloc glow accumulation
-Necro - Passive chance of converting undead

And so forth. These would have to be well documented to avoid spoiler use (seems easy to add to skill descriptions). To compensate for this buff to specialist casters, you could slightly nerf the super high level spells (which would also make some other people here happy)

For this message the author Cybermg has received thanks:
dolphin

bt

Mines Malingerer

Posts: 47

Joined: Thursday, 7th April 2011, 18:13

Post Friday, 8th April 2011, 13:00

Re: Cross-school spellcasting is too easy

What if fire spells casted by FE were for example 10% more powerfull/succesfull, ice spells 10% less so and air/earth - normal.
That way FE will be encouraged to use fire spells, but won't be restricted from using other elements.

Vestibule Violator

Posts: 1567

Joined: Friday, 21st January 2011, 22:56

Post Friday, 8th April 2011, 13:46

Re: Cross-school spellcasting is too easy

The real problem isn't cross-school casting being too easy, it's that there are lots of low level spells that stay useful even in extended endgame. Sure, being able to pick those up pretty much for free later on makes this problem a lot bigger, but if my EE had to train air to level 5 (compared to the current level 0 requirement) to cast swiftness and rmsl I would still do that. And while getting more ways to cause damage is usually a waste of resources getting more utility spells can often save a char's life. Low skill levels aren't a big investment. The problem with this problem is that it cant be solved. There need to be diverse, strong and interesting low level utility spells to make the early and midgame fun and to differentiate the backgrounds from each other. Making spellpower matter more for buffs is something that should be done regardless, even if it would just make people train air to level 5 or rely on their charms skill. Doing this and removing spellcasting from the power calculation sounds like a great idea though.

In my opinion the best way to reduce this problem is to introduce a cost for generalizing your spellbook (not skills), for example a spell slot cost. This could be done by penalizing knowing many spells (example: normal cost up to your 8th spell, then +2 slots for 9th, +4 slots for 10th, etc.), or by penalizing knowing spells of many schools (example: the first four spell schools in your book don't cost extra, each further school you add costs 5 extra slots, maybe extra +5 cost for having opposing schools in your book). Either way you cant just pick up blink/rmsl/swiftness for free. Doing both would probably be best, with lower costs than in the examples.

Other options to penalize generalization also exist of course. Success rate or spell power could be penalized. Nastier miscasts could be interesting, especially when combined with a spell success penalty for knowing many spells (so you learned rmsl and swiftness, and now your LCS blows up in your face 10% of the time). These systems might make more convenient targets than spell slots because they already are intransparent and complicated compared to the very transparent spell slot system. In any case I think penalizing generalism based on the spells you know and not based on skill levels would be better, because fiddling with skill costs wont actually prevent people from filling their spellbooks with all sorts of things.

As for keeping massively multischool casters viable with these changes, Sif Muna comes to mind. Sif already is the generalist magic god. Just keep things like they currently are for Sif worshippers and penalize everyone else.

More interesting highlevel spells might also help. If more schools had meaningful level 7 and 8 utility spells people might not always use their spell slots on assorted lowlevel stuff from other schools. A problem with this is of course providing interesting options while avoiding power creep (as shown by Tornado being creepily powerful).

Another problem is that four awesome lowlevel buffs that almost every char wants are charms/air (Swiftness, RMsl, Fly, Insulation). Yes, this is strong flavor for the air school. But these spells using the same school combination is part of what makes everyone pick up the same spells eventually. Moving RMsl and DMsl to some other school (fire and earth are pretty one-dimensional) might help a little. Maybe RMsl and DMsl could transmute incoming projectiles into dirt, or translocate little bits of rock into their flight path. Coming up with flavor for fire is quite a bit harder because many projectiles aren't flammable and being sprayed with molten steel would be rather unpleasant. Of course with no extra cost for generalization of the spellbook moving them to different schools wouldn't help much.

As seen above I don't agree that spreading utility spells across most schools would be bad. There might be a little less incentive to branch out into air if earth already gives you RMsl, but there's still Swiftness, Fly and Insulation in air. Having one school focused almost exclusively on causing damage by shooting things with elemental projectiles is fine. Two is pushing it, but currently there are three elements that do this (Fire, Earth, Poison), plus the school of hitting things with projectiles (Conj).

For this message the author Galefury has received thanks: 2
Cybermg, dolphin

Dungeon Master

Posts: 3618

Joined: Thursday, 23rd December 2010, 12:43

Post Friday, 8th April 2011, 14:34

Re: Cross-school spellcasting is too easy

I really like the idea of having school variety cost you in more spell slots. That's simple, easily explained and would create choices. Thanks!

Mines Malingerer

Posts: 44

Joined: Thursday, 7th April 2011, 12:24

Post Friday, 8th April 2011, 16:47

Re: Cross-school spellcasting is too easy

I really like the idea of having school variety cost you in more spell slots. That's simple, easily explained and would create choices

I like it too - very logical, and completely under player's control.
But it should be more gradual. First 4 for free, next one is +1, 6th one is +3, 7th is +5, etc.
The only problem is that players might wish to change their spell selection too often... Other games use things like book spellcasting to remedy this.

4)Gift new spells for training a skill("you research a new spell")

Rods provide an interesting insight how to do that: make each spell have several upgradeable stages. At first stage it is only a Flame Tongue. Once you meet the requirements for second(and pay extra spell levels for research), you can use advanced Throw Flames as well. Improve some more - and you have access to Fireball effect. Three spells in one!
User avatar

Ziggurat Zagger

Posts: 5832

Joined: Thursday, 10th February 2011, 18:30

Post Friday, 8th April 2011, 16:50

Re: Cross-school spellcasting is too easy

SinsI wrote:
XuaXua wrote:Make Spellcasting a factor of all spell (plus evocations not invocations) skills.
Value equals all spell skill values / 5 max 27
Make casting from scrolls train evocations to allow non casters build up to read books.

Make ability to memorize a factor of spellcasting level (as determined above) and int.


That would have an opposite effect (you'd want to train as many skills as possible) - although the same is true for current spellcasting.


Then make higher levels of a skill provide more punch as follows (numbers subject to change for balance purposes)

Remove Spellcasting as a skill that can be trained and make it an ability like Dexterity, Int and Strength.

Every 5 spell school skill levels gives +1 Spellcasting skill point, maximum 27.

skill level 1-10 count as 1 school level (10 levels = 10 school levels = 2 Spellcasting skill points)
skill level 11-20 count as 1.5 school level (10 levels = 15 school levels = 3 Spellcasting skill points)
skill level 21-25 count as 2 school level (5 levels = 10 school levels = 2 Spellcasting skill points)
skill level 26-27 count as 4 school level (2 levels = 8 school levels = 1.6 Spellcasting skill points)

Where spell school skills are all Spell Skills + Evocations Skills, but not counting Invocations Skill.
This allows someone at Evocations: 5 to have Spellcasting 1 and start memorizing spells, but only works if reading Scrolls adds experience to Evocations Skill (in addition to wands, devices, etc.).

Having one spell school skill at 27 gives 43 school levels gives a Spellcasting Level of 8.6

Having four spell schools at 12 gives 56 school levels gives a Spellcasting level of 11.2

This is pretty much par for the course of such a character.

Advantages of this system
1 - It encourages skill diversification
2 - Keeping the 27 level maximum is means it is not penalized by opposing skills.
3 - school level multiplier can be adjusted on a per-race basis to maintain a racial balance.

Disadvantages of this system
1 - tough for a player to track
2 - numbers probably need some balancing

A similar system could be put in place to replace Fighting (which should add to AC, not HP) using the hand-to-hand combat skills and possibly Dodging, using missile and thief skills, if we want to give Dexterity it's own centralized skill.
"Be aware that a lot of people on this forum, such as mageykun and XuaXua, have a habit of making things up." - minmay a.k.a. duvessa
Did I make a lame complaint? Check for Bingo!
Totally gracious CSDC Season 2 Division 4 Champeen!

Halls Hopper

Posts: 86

Joined: Friday, 1st April 2011, 23:44

Post Saturday, 9th April 2011, 06:52

Re: Cross-school spellcasting is too easy

Galefury wrote:The real problem isn't cross-school casting being too easy, it's that there are lots of low level spells that stay useful even in extended endgame. Sure, being able to pick those up pretty much for free later on makes this problem a lot bigger, but if my EE had to train air to level 5 (compared to the current level 0 requirement) to cast swiftness and rmsl I would still do that. And while getting more ways to cause damage is usually a waste of resources getting more utility spells can often save a char's life. Low skill levels aren't a big investment. The problem with this problem is that it cant be solved. There need to be diverse, strong and interesting low level utility spells to make the early and midgame fun and to differentiate the backgrounds from each other. Making spellpower matter more for buffs is something that should be done regardless, even if it would just make people train air to level 5 or rely on their charms skill. Doing this and removing spellcasting from the power calculation sounds like a great idea though.


I think this problem can be solved. Make it impossible to learn spells from a school if you have a difference of (10-Spell Level) between it and its opposing school and remove crosstrain penalties a bit. I imagine this would force people to alternate polarized elements and give you the usual tradeoff of multiclasses, in that they're incompetent in all classes by comparison to pure casters. It also gives you the identical end-game bonus of a multiclasser, in that they have tons of moderate level spells by comparison to your very few end game ones, which are invariably specialized (how often does one cast Meteor Storm or Death Channel).

sigfried von murdock wrote:Just on a general basis, I fail to see how learning more about one element/type of weapon/ability/skill could possibly conflict with the application of another.

Why would a fire elementalist be stumped with ice magic? Both are forms of magic that deal with the elements, the whole class specification only decided where the first experience points were allocated and the magic users load out of equipment. Why wouldn't a magic user strive to excel in all forms of magic (if said magic does not cross with the principles of the user) ?


A monk can effectively use a shield, sure there are penalties while wielding a weapon, and even more to unarmed, and the monk's initial stats are not developed in shields... But the monk isn't gimped simply because he is primarily trained in unarmed. Class does not define a character in this game, an earth elementalist is may be trained in Earth magic, but why would that not be the first of many steps to becoming a true elementalist capable of control over all four elements?

The whole conception of crosstraining penalties is to gimp characters and steer from a style of play or character development. There is nothing wrong with a list of 10 powerful, varied, useful spells which ever magic user tries to acquire. But there should be no problem introducing new attractive spells, and mechanics to enhance game play and encourage a wide selection of play styles. Not discourage choice.


Exactly. Any mage with the ability to study literature should be able to understand the basics of another discipline. Overspecialization limits the mage's abilities; as you get deeper into a discipline you must by necessity learn less of it at some point simply because there is less between you and what is considered expertise. By necessity developing new things requires more time than simply copying, right? If this is the case then it doesn't seem unlikely that a mage would learn articles of casting from other disciplines whilst having a main focus on one or even more skills.

No one ever suggested that biomechanists should be gimped because they studied engineering and biology.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Saturday, 9th April 2011, 07:46

Re: Cross-school spellcasting is too easy

The Mantis wrote:I think this problem can be solved. Make it impossible to learn spells from a school if you have a difference of (10-Spell Level) between it and its opposing school and remove crosstrain penalties a bit. I imagine this would force people to alternate polarized elements and give you the usual tradeoff of multiclasses, in that they're incompetent in all classes by comparison to pure casters. It also gives you the identical end-game bonus of a multiclasser, in that they have tons of moderate level spells by comparison to your very few end game ones, which are invariably specialized (how often does one cast Meteor Storm or Death Channel).


Penalties like this are very bad, because they encourage odd metagaming to avoid them, and hard limits like this are even worse. A player would carefully budget their spells so as not to raise any one school excessively. Likewise, they would victory dance their low schools to avoid the restriction.

The Mantis wrote:Exactly. Any mage with the ability to study literature should be able to understand the basics of another discipline. Overspecialization limits the mage's abilities; as you get deeper into a discipline you must by necessity learn less of it at some point simply because there is less between you and what is considered expertise. By necessity developing new things requires more time than simply copying, right? If this is the case then it doesn't seem unlikely that a mage would learn articles of casting from other disciplines whilst having a main focus on one or even more skills.

No one ever suggested that biomechanists should be gimped because they studied engineering and biology.


I made this topic to address what I think is a gameplay issue - all late game casters look mostly the same. Flavor issues shouldn't get in the way (and the leading suggestion right now about spell slots is flavorful).

Halls Hopper

Posts: 86

Joined: Friday, 1st April 2011, 23:44

Post Saturday, 9th April 2011, 08:22

Re: Cross-school spellcasting is too easy

You're right. These restrictions are designed to encourage specialization, obviously. My interest is flavor. I just want there to be the option to double-specialize, but I feel extreme late game spellcasters should all be almost identical. Masters of their craft shouldn't be ignorant of parts of it. There's no problem with spell slots. It makes the game more complicated. It's also derivative. I don't think that originality is something that should affect someone from implementing something. Unoriginality is a reflection on the maker of the entire product. The product itself isn't affected at all.

To encourage specialization, there has to be bonuses for specializing. It doesn't make sense to hurt someone for learning new things. The better someone is trained, the better they fight. Why don't we add bonuses for specializing, then?

This makes spellcasting more powerful as a whole. We can make spellcasting have a slightly more aggressive failure curve. That would make sense flavor-wise. Having a limited set of spells makes spellcasting weaker as a whole- this is why I am against spell slots. Why do we have to make it harder to be a spellcaster? They already suffer because they do not train Fighting. I wish that we could keep fighters and spellcasters equal. I think that the thought of spell slots not making spellcasters less versatile late game is a grey area.

Halls Hopper

Posts: 64

Joined: Thursday, 10th March 2011, 11:51

Post Saturday, 9th April 2011, 09:20

Re: Cross-school spellcasting is too easy

While considering how to 'encourage specialization', please bear in mind that some background/spellbooks require cross-school spellcasting- like transmuters.

Halls Hopper

Posts: 86

Joined: Friday, 1st April 2011, 23:44

Post Saturday, 9th April 2011, 10:23

Re: Cross-school spellcasting is too easy

Hushed,

Thank you. That's exactly what I was thinking of. Transmutation requires multiple school casting- a transmuter would have to double specialize at least partly, which with CyberMG's proposal would make Transmuters as a class weaker. The proposal is only valid with straight classes - Earth/Air/Fire/Ice elementalists, but not when it comes to things like Translocations (Blink/Passwall)

Lair Larrikin

Posts: 23

Joined: Friday, 1st April 2011, 02:17

Post Saturday, 9th April 2011, 22:57

Re: Cross-school spellcasting is too easy

I completely disagree that the spells named as "worse offenders" are "offenders" of any kind.

Take Deflect, Repel, Swiftness + Flight. Non-Air specialists learn these spells because their schools offer nothing that covers these areas of weakness. What is a spellcaster supposed to do instead of using these for defense? Hope that the RNGs drop boots of speed? Hope that they find a good buckler? Play nothing but Spriggans and Felids?

If you want to encourage people to use other spells, the solution is to make those spells useful, not to make everything equally lousy.

Buff up Cond Shield or Stoneskin so they defend as effectively as Deflect. Make Hexes less ineffective as a school. Then let's see if all the end game casters look the same.

Crypt Cleanser

Posts: 726

Joined: Friday, 11th February 2011, 18:46

Post Sunday, 10th April 2011, 00:14

Re: Cross-school spellcasting is too easy

Ann Hal wrote:I completely disagree that the spells named as "worse offenders" are "offenders" of any kind.

Take Deflect, Repel, Swiftness + Flight. Non-Air specialists learn these spells because their schools offer nothing that covers these areas of weakness. What is a spellcaster supposed to do instead of using these for defense? Hope that the RNGs drop boots of speed? Hope that they find a good buckler? Play nothing but Spriggans and Felids?

If you want to encourage people to use other spells, the solution is to make those spells useful, not to make everything equally lousy.

Buff up Cond Shield or Stoneskin so they defend as effectively as Deflect. Make Hexes less ineffective as a school. Then let's see if all the end game casters look the same.


I kind of agree. When you complain about people using cross-school spells, you're basically talking about air users picking up iron shot or earth users picking up deflect missiles...and that's pretty much it. Fire and ice are way too interchangeable. There might be a spell one has that another doesn't, but what is a player really sacrificing when they choose to go with fire instead of ice, or vice versa?

As for earth and wind, if you were to say that cross-school spells were impossible, how many people would choose to forgo air for earth? Earth has a definite flavor of high single target damage, but most players would simply go with air for the buffs, then use fire or ice for their damage-dealing needs.

One example would be to add a mid-high level earth spell that gives better protection than stoneskin, but only works for melee attacks; basically, making Earth's counterpart to deflect missiles.
User avatar

Halls Hopper

Posts: 61

Joined: Friday, 18th March 2011, 03:16

Post Sunday, 10th April 2011, 02:13

Re: Cross-school spellcasting is too easy

Thumbs up to the "restrict spell slots" idea. I'm not sure if it needs to explicitly recognize school split, either; simply having a lower total slot count would mean being forced to choose between a few high level spells and a smorgasbord of low level ones. If the endgame deep elf who can use every spell he has a book for at excellent+ has 27 spell slots, you are guaranteed to not see Iron Bolt/Deflect Missiles/entire charms school/etc. because there simply isn't room; casters will refrain from "splashing" air for DM when adding Flight, Insulation, etc. to the package is not free. (Splashing earth will obviously still occur, as there are only about four spells in the whole school anyway.)

galehar wrote:* Nerfing. I think some spells could be nerfed by simply making them much more dependent on spell power. Repel / Deflect missiles. Mephitic too.


I like this idea, but dislike the fact that you unequivocally characterized prospective changes with the word 'nerf.' If a low-powered Deflect Missiles is significantly worse than the current version, great, I'll put those experience points into dodging instead-- but if a max-powered DM is not significantly better than the current version, a mistake has been made.

Cybermg wrote: ... unintuitive ways to exploit the system. Namely, someone might avoid getting skill levels on one school or only get *just* as many as they need, so that they don't completely trash their abilities in an opposite school. ...


Imagine that I am a computer scientist who is primarily interested in artificial intelligence. Early in my career, I studied and practiced very hard to become at least Very Good, and if my individual aptitudes allowed, Excellent at graph theory. Learning that "school" to my satisfaction, I took a handful of its best techniques, such as A* and decision trees, which will be useful for the rest of my career as an AI-focused computer scientist. Then I stopped practicing it completely in favor of other skills.

Is this unintuitive? Yes; why would I put time and effort into learning a skill that I'll later neglect? Is it an "exploit?" I don't believe so. I knew what I wanted (the techniques), learned enough of the discipline to get them, and dropped it.

sigfried von murdock wrote: ... There is nothing wrong with a list of 10 powerful, varied, useful spells which ever magic user tries to acquire. But there should be no problem introducing new attractive spells, and mechanics to enhance game play and encourage a wide selection of play styles. Not discourage choice.


I wholeheartedly agree with this, but what I (and probably a lot of the commenters on this thread) want to see is a little more "try" before the "acquire." It's just too easy to grab a double fistful of low level spells and reduce 95% of the game's otherwise interesting tactical play to merely stabbing macro keys.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Monday, 11th April 2011, 09:37

Re: Cross-school spellcasting is too easy

There are a few directions here that I think people have agreed are not great for the game. I don't support these things either:

-Making the spell schools more similar to each other, so that a single school (or two) provides all of the tools that a caster will need.
-Penalizing cross school casting so that most casters can't cast in multiple schools.

However, I completely disagree with the following statement:

The Mantis wrote:I feel extreme late game spellcasters should all be almost identical. Masters of their craft shouldn't be ignorant of parts of it.


This is precisely the problem I brought up in my original post. This causes all of the fantastically varied and wonderful spellcaster playstyles to fold into basically the same character, with a few high level spells unique to their class to differentiate them. I hate to bring up the N-word, but it feels a bit like nethack where casters have an obligatory "ascension kit" of spells that while not *required* to win, certainly each help more than any other in their own way.

It's easy to take specialization or generalization to one extreme or another, and both extremes are bad. Obviously, what I'm trying to strike here is a balance, and I think the balance should be moved a little bit more toward specialization.

To give a high level idea of what I'm looking for (I mentioned this briefly in my original post), I want a system where you can't have it every way at once. This sort of thing exists for spellcasting and armor (you're never realistically going to cast level 6 spells in GDA), and with shields and high base damage weapons. In the end, no (reasonable) amount of skilling is going to get you past these restrictions; you have to make a choice.

Presently with spellcasting, any character with sufficient spellcasting skill can basically cast all of the spells they'd ever want to cast. Yeah, you can't cast both shatter and firestorm with the same character, but you wouldn't want to. Every competent spellcaster can pick up every utility spell in the game, and they all end up looking almost exactly the same.

Now if you're not allowed to cast every utility spell at endgame, it's true that spellcasting in general becomes weaker. Suddenly (for example), you can't put up haste, deflect missiles, phase shift, regen, and ozo's armor at every demon lord encounter, with tele self and abjuration for emergencies. I don't want to greatly change the balance of the game, so I think the change should be minor. It shouldn't be asking too much that a reasonable endgame build can only cast some (probably large) subset of these spells, centered around (but by no means restricted to) their few dominant spell schools. Since there are currently a wide range of playstyles possible, hopefully different characters will have different spell subsets they'd like to select, leading to many differing spellcasting builds for the endgame. Hopefully the selection of this spell subset for a single character is also an interesting choice. I think that the slot restriction suggestion gets pretty close to this (there were some pretty good points made in that post about how adding skill barriers doesn't necessarily solve the problem). It is still just one of many, many things that can be done to change the specialization/generalization balance though. I'm not necessarily trying to give the solution here, just define a problem and give my vision of where I think things should be headed.

For this message the author Cybermg has received thanks:
dolphin

Dungeon Master

Posts: 3618

Joined: Thursday, 23rd December 2010, 12:43

Post Wednesday, 13th April 2011, 23:34

Re: Cross-school spellcasting is too easy

Cybermg: I like the analysis, but I've seen similar discussions before: nothing will happen if we don't actually try to go for a solution. So let's do that. In my opinion, the restriction of spell slots is simplest, and I'll think about this.

I want a system where:
(a) the order in which you learn the spells does not matter (same total slot cost)
(b) sticking to two skills gets no penalties
(c) the more schools you take up beyond two, the bigger the penalty gets

Proposal: A spell you want to learn costs (level + max(0,schools-2)) slots, where "schools" is the number of all schools present in your spell list, including the potential new spell.

This might be too drastic, I didn't play around with examples. It does have the nice side effect that picking up a low level spell from a different school is penalised harder than taking a high level spell (higher relative slot cost).
Obviously, the Memorise screen would need to have a "Slot" column.
Unfortunately, this misses (a) -- assuming you know your preferred final spell set in advance, ideal behaviour is to learn the spells from the dominating schools first.

Blades Runner

Posts: 546

Joined: Monday, 20th December 2010, 14:25

Post Thursday, 14th April 2011, 00:34

Re: Cross-school spellcasting is too easy

I think that part of the problem is that low-level off-school spells are really cheap by midgame. If they were appropriately expensive to dabblers, everything would work well.

dpeg wrote:I want a system where:
(a) the order in which you learn the spells does not matter (same total slot cost)
(b) sticking to two skills gets no penalties
(c) the more schools you take up beyond two, the bigger the penalty gets

An idea for for having additional school cost in spell slots:
Each additional spell school learned beyond the first costs one additional spell slot. Suppose a caster has spells in just two schools. If she wants to add a third, it would costs 2 extra spell slots. Adding a 4th would cost 3 extra spell slots.

This would make it harder for EE and others to pick up low-level air spells. It might wreck transmuters and other builds that use many schools, but the levels of some spells with multiple schools could be decreased to compensate (and the transmuter mid-game is pretty strong already).

Vestibule Violator

Posts: 1567

Joined: Friday, 21st January 2011, 22:56

Post Thursday, 14th April 2011, 15:18

Re: Cross-school spellcasting is too easy

dpeg wrote:I want a system where:
(a) the order in which you learn the spells does not matter (same total slot cost)
(b) sticking to two skills gets no penalties
(c) the more schools you take up beyond two, the bigger the penalty gets

Proposal: A spell you want to learn costs (level + max(0,schools-2)) slots, where "schools" is the number of all schools present in your spell list, including the potential new spell.

This might be too drastic, I didn't play around with examples. It does have the nice side effect that picking up a low level spell from a different school is penalised harder than taking a high level spell (higher relative slot cost).
Obviously, the Memorise screen would need to have a "Slot" column.
Unfortunately, this misses (a) -- assuming you know your preferred final spell set in advance, ideal behaviour is to learn the spells from the dominating schools first.

I think allowing only two schools without a penalty is too drastic in general, Transmuters and especially Wizards (and probably some other backgrounds) would be completely screwed by this. Spell slots are already scarce early on, and screwing with background balance this much would be bad in my opinion (FE has two schools in the starting book, IE has 3-4 depending on whether you want summon ice beast, Tm has 5, Wz has 8). And with the higher relative penalty to low level spells the system you suggest would be especially drastic in the early game, when many low level spells are learned, while the actual problem is people picking up very useful low level spells at almost no cost in the mid and endgame.

This could be adressed by allowing generalism if you stick to low level (or same level) spells only (Wz), or by reducing or waiving the penalty for multischool spells that share a school (Tm). A simple (but unfortunately very bad) proposal that incorporates both is this: every spell costs at least as many slots as the highest level spell in your spellbook that doesn't share any schools with it. This also violates (a), though, and badly. Example: You know Ice Storm (lvl 9), Refrigeration (lvl 5) and Flight (lvl 4) and want to learn Conjure Flame (lvl 3) for some reason. It will cost 5 slots because it doesn't share any schools with Refrigeration which is level 5. It also doesn't share any schools with Flight, but that is only level 4. Ice Storm shares the Conjuration school with Conjure Flame, so it doesn't count. Why is this proposal bad? The extra cost can be completely dodged by learning lower level spells before higher level ones.

This could be fixed by always assuming higher level spells to be learned first. That would make this penalty extremely brutal however. A Wizard with all his starting spells learned would have to pay 9+7+7+7+6=36 slots to learn Fire Storm (actually 33, because 3 extra slots are already being paid for affected spells), which might be a bit much. It would cost even more if he also learned some other non Fire non Conj spells in the meantime. On the other hand, a dude with Firestorm and lots of low level utility spells of various schools is exactly what is supposed to be penalized, so this might actually be okay. For comparison: with this system a Tm learning everything in their starting book would have to pay no extra cost, a Wizard would have to pay 5 extra. Both would face fairly drastic costs for learning higher level spells from other books however (higher level non Tm spells for the Transmuter, pretty much anything for a Wizard).

The cost can be scaled down easily by setting minimum cost to x*level of highest unrelated spell. With x=0.5 for example people could learn Firestorm (lvl 9, 9*0.5=4.5 ==> minimum cost for unrelated spells is 5), Haste and DMsl at no penalty (which is okay because casting these needs skill investment), but they would have to pay 6 slots extra for a set of Firestorm, Blink and RMsl (which require no skill investment). 6 slots doesn't seem like much on paper, but in game even this penalty would be enough to discourage rampant learning of anything lowlevel. If you can replace a level 2 spell with a level 5 one at no extra slot cost you probably will, and if you wont it will most likely have been a meaningful decision.

The main problem with this system is that it is not very obvious to new players due to the delayed effect of being able to freely generalize early on or with spells sharing a school but then being punished for it later. Also it sometimes violates (b), and in many cases (c). It also indirectly penalizes learning highlevel spells, even if they are just a continuation of your focus (a FE with starting book spells who also knows Blink and wants to learn Firestorm would have to pay extra for that because learning Firestorm will raise the level of spells unrelated to Blink to 9). This last one is a problem most systems that fulfill (b) will have if they depend on spell levels though. This system has the advantage of not requiring a complete rebalancing of the early game and backgrounds however, which in my opinion is a pretty big advantage and should be right up there as requirement (d).

Maybe someone can come up with something better. I still think having a per-school penalty instead of a per-spell penalty would be easier and sufficient. Also the penalty doesn't have to be huge. 5 slots total extra cost is already noticeable, 10 will make you think hard if you really need this stuff. 5 slots extra cost per spell is completely crazy. If a high proportion of penalty slots compared to regular slot cost is desired the number of available spell slots could always be raised of course.
User avatar

Dungeon Master

Posts: 4031

Joined: Thursday, 16th December 2010, 20:37

Location: France

Post Thursday, 14th April 2011, 15:41

Re: Cross-school spellcasting is too easy

Wouldn't it be simpler to just reduce the maximum number of spells?
<+Grunt> You dereference an invalid pointer! Ouch! That really hurt! The game dies...

Dungeon Master

Posts: 3618

Joined: Thursday, 23rd December 2010, 12:43

Post Thursday, 14th April 2011, 17:07

Re: Cross-school spellcasting is too easy

Yes, that's good in any case. But I believe this alone will not cut it.

Blades Runner

Posts: 546

Joined: Monday, 20th December 2010, 14:25

Post Thursday, 14th April 2011, 19:03

Re: Cross-school spellcasting is too easy

galehar wrote:Wouldn't it be simpler to simply reduce the maximum number of spells?

That could make it easier to pick up off-school utility spells than to pick up an high-level on-school spell. (Or it might not.) It might also make spellcasting stronger, as spell slots become more valuable. (Or it might not.)

We want to make diversification harder for specialists but not for generalists, right? What about making a training and/or spell power malus that is a function of the square of skills in other schools?

For example, X/sqrt(school1^2+school2^2+...), where X is spell power or skill level, schoolY is a skill, and Z is some scaling factor. Very high skills would make spells less powerful or more expensive to train but low skills would not. Let's compare two casters with 24 skill levels across all spell schools. For a caster whose only only skill is conjurations 24, picking up air magic would be much more expensive: X/sqrt(24^2))=X/24 . For a caster with 8 schools at level 3, picking up air magic would be X/sqrt(8*3^2)=X/8.5. This formula is silly as is, but scaling by a factor of 5 or something might make it work.

I

Dungeon Dilettante

Posts: 2

Joined: Saturday, 16th April 2011, 07:24

Post Saturday, 16th April 2011, 07:53

Re: Cross-school spellcasting is too easy

I'm not so sure that reducing the number of spells memorizable by a generalist versus a specialist is the best option. The problem is that there is no notion of a 'class' in DCSS, simply a background which gives you a specific starting kit. Once you get beyond early/mid game the effects of your choice of expertise have already faded away, unlike choice of race or god.

My attempts at a solution:

1) Make your choice of background/class mean something permanent besides 2 letters after your race, why not have the class add +1 to your race's aptitude for the associated skill. This is easy for the spell casting themed classes (Fire Elementalist = +1 to Fire Magic aptitude) maybe not as easy for the others, though I'm sure there could be an agreement for them. This makes sense just as it does in real life, some people are simply more apt at certain topics than others, not every human learns math as quickly as a mathematical genius for example. I don't believe a change like this would make good race/bg combos way more powerful, but it would probably make less popular race/bg combos a little more palatable.

2) Make higher level spells have a requisite minimum level in its relevant schools to memorize them (this is a perfect way to balance out more powerful spells too.) It doesn't matter how good you are at spellCASTING, what chance would you have at casting Necromutation if you were just a Necromancer newbie and couldn't understand the complex necromantic topics involved in it. I believe this would be perfectly fair, if you want to cast all the amazing spells from all the schools you can still do that, you just have to start from the beginning in the relevant schools like every other rookie. You'll definitely learn faster than a true spell casting rookie since your chance to cast will be much higher, but you still have to learn the basics of the school before you can go off and start slinging the most powerful spells in the school.

I believe both 1) and 2) together would lead to more interesting early/mid/late games with the ability to still have powerful all around casters in the extended-late game that won't have to travel to their stash to forget then memorize spells needed for whatever area they're about to go in.

Abyss Ambulator

Posts: 1221

Joined: Thursday, 10th March 2011, 19:45

Post Monday, 18th April 2011, 10:53

Re: Cross-school spellcasting is too easy

I think a potential problem with restricting spell slots based on the number of schools is that it will hurt specialist spellcasters a lot more than hybrids:

For example, a specialist spellcaster needs to use spells for killing AND defence/buffing, but might find themselves having to choose between their combat spells and 'survival' spells because these spells are likely to use different schools.

But a crusader-type character who relies on melee to kill monsters and uses spells only for buffs doesn't need to train that many schools - they have no use for high levels of Conjurations or Summoning for example. They still have to choose between raising their killing abilities (weapon skills/fighting) and their buffing abilities (magic schools), as they already do, but restrictions based on spell schools probably won't that much of a difference to this choice.

This could have some odd effects like forcing 'pure' spellcasters to train their shields skill to a high level to attempt to replace some of the defence they lose by having fewer spell slots.

If it's decided that pure spellcasters are currently overpowered compared to hybrids, this is probably a good thing - but it would seem a bit odd that a character who spends most of their time whacking things with weapons would find it easier to learn their choice of low level buffing spells compared to one who is a master of magic.

Edit: to me the idea that 'cross-school spellcasting is too easy' doesn't make any more sense than to say 'learning both weapons and invocations is too easy' or 'learning both melee and ranged weapons is too easy'

Vestibule Violator

Posts: 1567

Joined: Friday, 21st January 2011, 22:56

Post Monday, 18th April 2011, 11:50

Re: Cross-school spellcasting is too easy

Jeremiah wrote:Edit: to me the idea that 'cross-school spellcasting is too easy' doesn't make any more sense than to say 'learning both weapons and invocations is too easy' or 'learning both melee and ranged weapons is too easy'

The problem is that by endgame (dedicated casters can do it by midgame) people can pick up lowlevel utility spells for just the spell slot cost (which is low for low level spells) without additional skill investment (either because of dual school spells or because of spellcasting). This is part of the reason why every caster ends up with mostly the same set of utility spells. To use both melee and ranged weapons well you actually need to level up a melee and a ranged skill. I think we can agree that getting something for nothing is bad design, at least in the context of character development. It is currently optimal to learn spells like Blink, RMsl and Swiftness for any caster or hybrid. This is not an interesting decision.

So, why is Spellcasting skill affecting casting success a good thing again?

Ziggurat Zagger

Posts: 3037

Joined: Sunday, 2nd January 2011, 02:06

Post Monday, 18th April 2011, 12:20

Re: Cross-school spellcasting is too easy

Galefury wrote:It is currently optimal to learn spells like Blink, RMsl and Swiftness for any caster or hybrid. This is not an interesting decision.


This is more to do with the fact that these spells have an irreducible positive effect than anything else. Repel Missiles doubles the effect of your evasion against basic ranged attacks, and Blink or Swiftness very nearly shut down most melee threats. These effects are massively powerful, but moving them to a more appropriate cost and level would be rather labor-intensive since the impact would be felt throughout the entire caster game.

And on the other hand, no level of negative incentive for diversifying will overcome the fact that most magic skills simply don't have many spells worth taking. If you're playing a conjuror, you'll have a couple of elemental conjurations. Once you have your staples, all others are zero value. Before I find low-level spellbooks with cross-school spells, I routinely don't bother to fill in 10 to 15 spell slots because I don't need my spell screen cluttered with redundant crap that I'll never actually use.

I'd rather spend 10 spell slots getting basic Blink than 5 getting yet another damaging conjuration to go with the three I actually use.

Dungeon Master

Posts: 3618

Joined: Thursday, 23rd December 2010, 12:43

Post Monday, 18th April 2011, 12:32

Re: Cross-school spellcasting is too easy

I think it is established that success and power are affected by too many factors. No need to ride that horse to death again.

But I agree that Galefury's reasoning indicates why just reducing spell slots (for everyone) will not achieve the goal (although it's probably still a good idea). Punishing skill spread with skill slot loss still seems good to me.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Monday, 18th April 2011, 13:08

Re: Cross-school spellcasting is too easy

dpeg wrote:Cybermg: I like the analysis, but I've seen similar discussions before: nothing will happen if we don't actually try to go for a solution. So let's do that. In my opinion, the restriction of spell slots is simplest, and I'll think about this.

I want a system where:
(a) the order in which you learn the spells does not matter (same total slot cost)
(b) sticking to two skills gets no penalties
(c) the more schools you take up beyond two, the bigger the penalty gets

Proposal: A spell you want to learn costs (level + max(0,schools-2)) slots, where "schools" is the number of all schools present in your spell list, including the potential new spell.

Here's one system I thought of:

Slot penalty = .5 * max( 0, [rank of spell's school in terms of how many spells from that school you know] - 2)
Spell slot cost = spell level + slot penalty

The .5 and 2 are arbitrary constants that can be tweaked. The rank term is best explained with an example. A character knows the following set of spells:

Fire - 5 spells, Tloc - 3 spells, Conj - 2 spells, Charm - 1 spell

These schools have rank 1, 2, 3, and 4 respectively. To simplify my slot penalty computations that follow, I'll change the formula for now to penalty = 1* max(0, rank-2). In this case, this character would have no penalty for learning a new fire or tloc spell, a penalty of 1 slot for a new conj spell, 2 slots for a new charms spell, and 3 slots for a spell of a new school. This example penalty is rather harsh, so you'd probably want to use a fractional multiplier instead of 1 - in this case, you can just keep track of fractional slot penalties. A spell's penalty will go +1 when the accumulated fractions add to over 1.

It's fairly straightforward to apply this to multi-school spells. Each of a spell's schools would add to that school's count (ie, conjure flame would add +1 to fire and +1 to conj). The penalty would have to be the sum of each school's penalty (learning conjure flame in the example given would have a 1 slot penalty). Alternately, multi-school spells could add fractional counts (adding to 1) to the spell schools (conjure flame would add .5 to fire and .5 to conj), and the penalty would be similarly computed (0*.5 + 1*.5 = .5 penalty in this case)

This scheme meets criteria a) since you can always compute the total spell slot penalty accrued based only on a character's current spell set. Just sort the schools by # of spells known, and apply the slot penalty formula. The character above has accrued 5*0 + 3*0 + 2*1 + 1*2 = 3 total slots of penalty. Since the calculation is only based on current spells known, it doesn't matter what path the character took to get there and thus a) holds.

The system makes intuitive sense because it should be easier to learn new spells in schools you already know well, gradually becoming more costly for schools you know less well. The slot penalty changes as a character learns new spells and spell school ranks change. A character with spell counts of Fire-5, Conj-5 would pay 1 slot penalty per spell of a new school, say Charms, until they know 5 charms spells. At 5-5-5, the schools are equally well known and the next spell comes without penalty. If the character continues learning charms, charms would sit at rank 1 and they'd receive no penalty per spell.

This post turned out way longer than I planned, but the system is actually fairly simple.

Vestibule Violator

Posts: 1567

Joined: Friday, 21st January 2011, 22:56

Post Monday, 18th April 2011, 16:59

Re: Cross-school spellcasting is too easy

About the above system: this sounds nice and simple to me, but like other proposed systems screws up the early game for Wizards, Transmuters, and other diverse casters. Depending on how multischool spells are handled Tm, Cr, etc might be fine, but Wizards would need to be looked at. Everything else seems good to me. At first I thought there would be large cost jumps when school ranks change, but this is not the case. In fact, like Cybermg said in his example, upon a rank change you pick up a spell without penalty or at lower penalty cost.

Also rank is not properly defined for schools with equal amount of spells. It's fine though, the order can be arbitrarily assigned in this system without changing anything, because the penalty only depends on the number of known spells of each school. Depending on how multischool spells are handled this might change though, so this is something to keep in mind.

I think multischool spells should count fractionally, having a double penalty for them would feel weird to me. Also you already have to pay double exp for them.

While using the same penalty for high and low level spells is good in the late game it causes problems early on, especially for Wizards. Shaving off a percentage of the character's total spell slots instead of a flat slot cost could help with that. Player level could also be used if pure casters should be treated less harshly than hybrids (casters have more spellcasting than hybrids, so a lower percentage of their spell slots comes from char level). Simply capping the penalty cost at some percentage of total slots (like 30%) is another option.

Some numbers for the version with fractional cost for multischool spells, using the formula penalty = 1* max(0, rank-2):
Wizard: 2 Cj, 1 Tloc, 1 Summon, 1 Hex, 0.83 Air, 0.5 Charm, 0.5 Fire, 0.33 Poison
Penalty = 0 + 0 + 1 + 2 + 2.49 + 2 + 2.5 + 2 ~= 12
Regular slot cost = 16 or something
This is crazy.

Tm: 4 Tm, .5 Necro, .5 Fire, .5 Poison, .5 Ice
Penalty = 0 + 0 + .5 + 1 + 1.5 = 3 = not a problem

Lair Larrikin

Posts: 17

Joined: Monday, 4th April 2011, 02:40

Post Monday, 18th April 2011, 23:04

Re: Cross-school spellcasting is too easy

Radical but straightforward countersuggestion: Make all spells single skill. Rebalance as needed by bumping now-overpowered, formerly multiskill spells (e.g. mephitic cloud) to a higher casting level. Prune away any skills left underpopulated, and the corresponding starting backgrounds. High enough spellcasting and intelligence will still grant up to level 3 or 4 by late game, and that's OK, but things like iron bolt from Conjurations alone are gone.

Mines Malingerer

Posts: 50

Joined: Monday, 31st January 2011, 03:23

Post Monday, 18th April 2011, 23:16

Re: Cross-school spellcasting is too easy

(in response to galefury's post)

Having slot cost depend on a value that changes over time (and is unrelated to spells learned) seems like a bad idea, since spell slots consumed would also change over time. It'd be very bad if someone gives themselves 8 free slots for haste (allowing for penalty), only to discover that this has reduced to 5 free slots because of variable costs. The only thing that should change slot cost is learning and unlearning spells.

I agree that multi school spells should count for fractional schools (summing to 1).

A multiplier of 1 is quite high. I recommended .5 in my initial example, but something like 1/3 or even 1/4 might be more appropriate. Some examples with a 1/3 multiplier:

Wizard starting spell book:

Regular total slot cost: 16
Total penalized slot cost: 20
Slot penalty of spell from new school: 1.71 (easiest way to do this calculation is subtract total slot use afterward from total slot use before)
Penalty of another spell from that same school: 0. It's already one of the top schools.

15 rune Troll Wizard. Small amounts of utility spells in every school.

School counts: Tloc-5, Charm-2, Air-1.83, Conj-1.83, Fire-1, Sum-1, Hex- 1, Ice- .5, Nec- .5, Earth- .5, Tmut- .5, Poi- .33
Regular total slot cost: 48
Total penalized slot cost: 59.94
Penalty for a new, pure poison spell: 1.89. This number is so low because poison quickly climbs the ranks with just the one spell level learned.
Penalty of another pure poison spell: .54

DSNe with 3 runes. More focused spell set.

School counts: Nec- 7.5, Sum- 3, Tloc- 1.5, Charms- 1.5, Air- .83, Conj- .83, Poi- .33
Total slot cost: 50
Total penalized slot cost: 53.99
Penalty for spell in new school: 1.65
Penalty for another spell from new school: .67

Worst possible slot penalty for a spell: 3.33 (learning a pure spell in a 12th place school that remains 12th place afterward).

It turns out the penalties are quite modest, and it doesn't take much to rank a spell school out of the lower levels where the penalty is highest. The focused DSNe pays a very modest cost (about 4), while the Troll's cost is more significant (~12), and would have significantly affected its final spell list. The starting wizard has a definite reduction in spell slots, but it's not hard to overcome once you've started learning a given school.

edit: I guess the C * max(0,rank-2) formula isn't quite accurate once you take into account fractional spell schools. The new formula is more like (slot penalty after - slot penalty before), where the calculation of slots involves sorting by spells learned per school, then applying the penalty above.
Next

Return to Game Design Discussion

Who is online

Users browsing this forum: No registered users and 5 guests

Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group.
Designed by ST Software for PTF.