Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Magic

44 (19%)

Magitek

38 (16.5%)

Steampunk

65 (28.1%)

WE DON'T NEED NO WEAK CONTRAPTIONS! (Normal)

28 (12.1%)

All of the Above (Moddable)

56 (24.2%)

Total Members Voted: 231

[NW_Kohaku]

Magic is a tricky problem. Too ordered and you make an industry of it. Make it too chaotic and it has little use. Too powerful and all else fails. Too weak and it becomes insignificant. So we want to avoid mass produced potions, have those without magic hold some sway and remain strong enough to at least move the story somewhat. I'd say having only a few powerful magic users works well, but magic use should come with it's own risks.

Actually, I like it ordered, and I'd like to have a mass produced potion assembly line.  Rather than flinging fireballs from the outstretched hands of dwarven wizards, I'd rather my marksdwarves carry out an order to use their incindiary-tipped ammunition.  (Or the bolts that have "warheads" of jars of poison gas.)

I suppose magic shouldn't be so much chaotic, then, as dangerous.  You should have to consciously choose to play with powers beyond the mortal ken, and anything potentially devastating not caused by your own hubris should carry harbingers allowing you to defend against or even prevent the incoming mayhem.

Well, I think the best way to consider it is seeing it through a cost/benefit analysis:  Although I, personally, enjoy setting up clothing factories with everything I make dyed in either blue or green (or black if I have it) dyes, a great many players won't use dyes, since it only increases value, and has no practical benefits, wheras many other players will just let their dwarves go naked entirely - there's no real benefit to the expenditure of labor (or rather, the lag that having a few more laborers creates) for your clothing factory.

So apply this line of thinking to magic - the big problem I had with the "random magic" idea was not so much the "potentially kill your entire fortress, or at least permanently transform a legendary armorsmith you worked so hard to train into a useless toad" as it was the "why the Hell would anyone want to put up with that in the first place?"

Or, rather, whenever you want to introduce a mechanic that you want players to use, you have to ask yourself "what problems does the player have that this mechanic will let players solve?"

If you introduce a tool players can use, but they can solve all the problems that tool can solve already, and those other solutions don't have anywhere near the potential drawbacks of a wizard that is basically a walking timebomb waiting to go off, why would anyone want to subject themselves to the headache when they don't have to?

(This is exacerbated by the fact that this is a game that is all about taking precautions against every possible risk, not one about managing a crisis once it arrives.  In a fort, randomness is the enemy, and predictability is your friend.)

So, basically, this means that either you have to invent the problem that only magic can solve, or else you have to give massive benefits that only come from magic that would outweigh the headaches or outright existential threats to your fort that magic creates.  (The only thing I can really think of would be the exact ability to give the player a chance to respond to a crisis on the fly, and not merely ones that wizards themselves create, but an ability to stop a cave-in or a flooding or to generate a wall that stops an invader, or to just vaporize that stupid dead rat that nobody is going to bother to pick up, and is spreading miasma through the central stairway, so that EVERYONE gets negative thoughts from it.)

[NW_Kohaku]

I made an edit onto the response to Vattic, but it ballooned enough that I think it should be its own post.

I want dwarven alchemist/wizards to have their ammo for what they do become based upon having a safe, stable enviroment that can capture and extract components from creatures, or can cultivate difficult-to-grow ingredients (made possible with the Improved Farming changes).  Maybe there's some limit to how many magical items they can throw without resting as a sort of "magic point" mechanic, but the need to set up an industry and base of operations for the production of goods related to one aspect of dwarven industry is essentially one of the basic tennants of DF. 

This is what relates in with the systems we already have in Dwarf Fortress.  In the arguments I've had about Improved Farming, I remember this one (snipping out only the most relevant parts):

I've said (a long time ago) before that magic, if implemented, must capitalize on the systems already present in DF - temperature, growth times, item properties, etc... To introduce ideas like 'runes' and 'mana' and 'elements' willy-nilly is a just plain terrible idea. It introduces detail, homework, without introducing satisfying complexity.

This is a game that isn't finished by a longshot.  That means that some things get added.  I'd honestly like a magic system that works more like an Alchemy system, especially from a Gust game, where I could either mine for or grow ingredients needed to perform single-shot magic on the farms in this game.   If that farming takes advantage of the new "arbitrary" system, but then makes use of the same farming system instead of "mana", does it "un-arbitrary" the magic system?

Actually, yes. There's less 'homework' involved - anyone who knows their way around the farming system now has a significant leg up on magic. The system becomes more complex and satisfying, without introducing unnecessary detail.

Also, I just brought up the magic system as a clear example of what I meant by an 'arbitrary' system versus a 'non-arbitrary' system. Have you seen the suggestions people have made for magic? Many of them eschew almost all of the game's preexisting mechanics and places a completely outside system into DF. If you want to talk about lore-breaking, talk about suggestions to add the four classical elements as a distinct concept to DF. That's far more perception-breaking than, say, adding an ability to modify the game's weather and temperature systems. Once again, these are just examples to illustrate a concept - not necessarily tied to any one suggestion or person.

Basically, I think Normandy has a point on this regard - if you are going to introduce a new system because "making it an industry" is a bad thing... you have to do more than give a compelling reason for why it is a bad thing, you also have to say why something different would be a good thing, and would be good enough to warrant something entirely different.  (See the "what problem is this the solution to?" part of the last post.) Or in other words, if "making it an industry" is a bad thing, then burden of proof is on you to prove that assertion, because industries are what dwarves have now, and industry is such a basic aspect of DF that "A short, sturdy creature fond of drink and industry" is their description!  (And really, is the notion of a "wizard's tower" having a magic component processing plant and storeroom in the basement all that destructive of the notion of a "dwarven wizard", anyway?)

Simply saying "because magic is special" is not compelling - everyone has a different idea of what magic is or should be, which is exactly why I just plain gave up on the last argument over magic, since there's nothing you can do to argue over something that essentially amounts to "I like vanilla ice cream more than chocolate" versus "I like chocolate ice cream more than vanilla".  It has to relate to actual game mechanics, and what a player can use it to do, and how it impacts their gameplay.  (Or, as with the "what problem is this the solution to?" part, you have to start with what you expect the player to do, and then work backwards to setting up the mechanics that make the player want to do that.)

[Quatch]

Don't bash too hard on advanced civ for being disaster recovery rather than managment or prevention. Most of the point of technology is prevention and reduction of disasters, and if you arn't one step ahead of the game in setting up your empire to be resilient against the disasters that are likely to come in the next few turns, you are unlikely to win.

I see steampunk as being more like the conventional magic: X+Y=Z, dwarven calculators, etc; while 'magic' will have more of the unpredictable outcomes, changing in between applications and games.

[The_Fool76]

In my mind there are really only two styles of 'magic' that fit with DF.

One is using magic to make stuff.  Fire-proof plate, wound recovery potions, animated skeletal cats.  Urst McMage goes off to the mage workshop and combines 3 units of elf ear powder with a glass flask and some rum to make a potion that re-grows noses.

The other is putting it in OUR hands Dungeon Keeper style.  Urst McPriest sacrifices a caged goblin on the alter in a long ritual, giving us the ability to teleport one dwarf wherever we wish.

[NW_Kohaku]

Don't bash too hard on advanced civ for being disaster recovery rather than managment or prevention. Most of the point of technology is prevention and reduction of disasters, and if you arn't one step ahead of the game in setting up your empire to be resilient against the disasters that are likely to come in the next few turns, you are unlikely to win.

I see steampunk as being more like the conventional magic: X+Y=Z, dwarven calculators, etc; while 'magic' will have more of the unpredictable outcomes, changing in between applications and games.

I'm not bashing Advanced Civilization for being disaster recovery rather than prevention, I'm praising it.  I said it was a weakness of DF that you have to spend all your time preventing disasters because it's extremely difficult to recover if they actually strike.  Advanced Civilization, by contrast, makes disaster recovery very easy (and limits your overall expansion, so that it's actually kind of a waste not to get hit with a disaster), but basically ensures you get hit with at least one per round.

I've actually written fairly lengthy rants on the subject of randomness in games, and how they apply to the player's decisions... There's "card game" randomness and then "dice game" randomness (also, I distinguished "solved games", like chess, where nothing is random but the players, and "Candyland", where functionally everything is random), where in most card games, the randomness is in what you draw (the situation you find yourself in before you make your decisions), but once you have a hand of cards (situation) to play, you know exactly what the results of the actions you take will be.  Dice games, (such as most tabletop RPGs) on the other hand, have randomness after you make your decision - you choose your action, and then roll to see whether or not what you want to happen will actually happen.

"Solved games" demand long-term planning - the only real way to win in a game where both sides are equally strong is to have the ability to better forsee the consequences of your actions than your opponent does. 

"Dice Games," on the other hand, tend to prevent long-term planning.  You can try to make plans, but because you have no real way of understanding what the situation will be like in a turn or two, you can't really make long-term plans that are more than vague tactical advantages, like moving your units behind cover, or using spells that give you tactical advantages of bonuses to evasion or hitting.  They probably know that some actions are more likely to work than others, and definitely know when some actions are more risky but have better payoffs.  This style of game design tends to encourage players to solve whatever actions will give them the highest probability of success, and to just keep trying that same set of things over and over again until eventually they hopefully win.  I think the most extreme/absurd case against this sort of playstyle can be seen in this Order of the Stick comic, where the entire "wizard's duel" (well, actually, clerics, but that's beside the point) boiled down to two people standing next to each other, calmly taking turns tagging each other with spells repeatedly at one another until one failed their save.

"Card games" are not as focused upon long-range planning as Solved Games, because you will be stuck with more unforseeable events ahead of you, but because you can be assured of the results of what you are doing, you can be more assured of what will happen than in Dice Games.  When a card is in your hand, you know exactly what its effect will be - and often, you can count the cards, and know that (if we are talking about, say, spades or bridge) your King of Spades will win a trick because the Ace of Spades has already been played.  When you look at the cards in your hand or on the board in a Collectable Card Game, you can generally be assured that they will do what is printed on the cards, and look at the cards your opponent has on the board, and have an idea of what he can do, although you don't know what's in his hand (less so if you know what sort of deck he plays, so you can be assured that his deck doesn't have any counterspells or the like to stop your actions), it's just a matter of not knowing what new cards will be added to your hand next turn.

It goes without saying, but a "Candyland" game will have such utter randomness that the player has functionally no control over what happens, so there is no point in making a strategy at all.  (This is really what that Order of the Stick comic parodies the cleric's duel boiling down to.)

When we are talking about what sort of "randomness" we are introducing onto the player, you have to recognize what ways the player will have to react to that randomness.

In Advanced Civilization, even if you get random disaster cards thrown at you that level cities, you can generally prepare for the next disaster, by gathering people togther at the end of your turns to be ready to rebuild some of your cities as soon as they get knocked down.  The fact that you know what disasters are likely to pop up, and how to head them off and prevent too much damage from occuring when they land on your lap mean that even these "random" events are fairly mitigatable, allowing you some control over something that is theoretically out of your control.

In DF, you can do something fairly like this - you can't control the Forgotten Beast that comes rampaging into your fortress, but you can know that it's probably coming, and can know what sort of powers it has, and can contain the dangers, or prevent yourself from even being exposed to the risk by keeping your fort sealed off.

So then, this comes back to magic: Don't tell me about how "magic should be unpredictable", that's not a useful metric.  Tell me what kind of unpredictability.  Tell me what sort of ability the player should have to handle that unpredictability.  Tell me what sort of playstyle these changes are going to force the player to adopt, because it is the way to mitigate the risks and maximize the benefits.

Because if wizards mean there's a random chance of opening HFS in the middle of your fort, with basically no means of the player stopping the damage, there is no sane response to wizards besides killing them on sight. 

Even if the risks are small, but the benefits of magic give you nothing you can't get from more reliable methods, why bother with something risky and beyond your control - not having control over your game may sound "magicky" on paper, but it sure isn't fun in practice.  People don't like being the pawns of chance - as we grow older, we leave Candyland behind as a kid's game for a reason.

This has nothing to do with "techonology is reliable, magic is unpredictable", because technology is hardly reliable IRL, and I think Clarke's Law should really be more applicable, anyway - it doesn't matter whether it is magic or technology, except for aesthetics; Functionally, if technology can be problematic and unpredictable, and magic can be reliable and predictable, then you're just arguing your personal aesthetics of what magic should "feel like", not what it should be able to do.

So again, I want to bring it back to the question, "What should this make the player do?"  Or, "What problems does the player have that using this game feature will solve?"

Because when you get right down to it, that's what games are: A set of rules, and a problem to solve.  Games reward the behaviors by letting players solve their problems by using the types of actions the games encourage, bounded by the rules.  Whether magic or technology, we're talking about expanding the rules, so we have to ask what sorts of actions are we going to want to encourage the player to take?

[KillerClowns]

Nice rant, NW_Kohaku, and I mean that sincerely.  You're right, magic should definitely have benefits beyond "I need a challenge, so I'm gonna try not killing every mage I see."  Magic should be Fun, but it should also be fun.  Players shouldn't have much control over exactly what magic does, aside from perhaps making general requests (something like "focus on enchantment/summoning/alchemy/protection"), otherwise it becomes just another dull industry.  But they should be able to control when, where, and to what extent it operates.  I reiterate that magic should, unless intentionally allowed by the player to grow to a grand scale, be either a potentially beneficial nuisance ("hey, the alchemist just managed to call forth three barrels full of ambrosia from the realm of the gods... I think we can forgive him for summoning those obnoxious ethereal rats last winter") or treated as a natural disaster with opportunities to prepare beforehand ("the omens tell of an unnatural blight that will make our crops wither in two years... time to grab the jobless peasants and get them plowing").

Now, why would a player allow magic to grow to a grand scale?  Well, one possibility is a player intentionally seeking a high risk play style for their own amusement.  Another would be fighting fire with fire.  "I'll put up with a bunch of magi occasionally contacting fell spirits that seek only terrorize the living if they can stop the my crops from withering from blight."  A third might be a semi-controlled industry; give your magi a bunch of steel swords and instructions to the tune of "here, play with these," then chuck the ones they botch, sell the ones that are a mixed blessing, and keep the ones that could become the stuff of legends. Such would be a gamble; you're taking a sure thing and hoping you don't end up with a bunch of cursed swords that insult their wielder and openly mock their hygiene, because you might get a blade that annihilates demons as soon as it pierces their flesh.

[Africa]

I like DF as it is now: some technology, pretty much in the form of mechanical power and magma-generated heat, and the only "magic" is the presence of dragons and such, plus shit like demons and salt wraiths. Nobody's casting spells or anything. It's a very low-fantasy kind of feel and I like it how it is. I say keep it this way, but of course there is a planned magic arc.

[NW_Kohaku]

KillerClowns: Mmmm... I'm not quite sure how to say this, but, I think your paradigm only shifted about half as far as I was hoping. 

Although my using metaphors seems pretty hit-or-miss generally, I'll try another one (noting the irony that I'm using a risky manuever to argue against risk):  When people try to build a lottery system, they have repeatedly found that if you let the player of the lottery pick their own numbers, you will get three or four times as many players willing to play the lottery.  The reason is that people inherently fear or dislike a lack of control over their life.  (It is that same Candyland randomness I talked about earlier - the player is irrelevant, they have no control over the outcome.)  When you give them any amount of control, even purely cosmetic, illusory control, like picking a "lucky" number, they are much more happy to play. 

This is why I don't really like the kinds of randomness I keep seeing suggested in much of these magic threads - it's that people want magic to be some kind of Harry Potter "LOL RANDOM!" magic... If the hallways magically shift around and close off access, it may sound cool when you're reading it in a book, but when you actually have to run a system where the paths are always changing, your just going to make dwarves burn more processor power trying to pathfind through hoops that now move for absolutely no benefit.  Who the Hell would want that?  To again turn to a webcomic parody, I think Sluggy Freelance's Torg Potter best summed it up when Torg watches the toilet paper magically animate itself, jump off the roll, and leave him stranded in the toilet with nothing to wipe with.  In the next panel, he's screaming "THIS PLACE IS TOO MAGICAL!"

So then, getting back to what this whole thread is actually about...

I wouldn't like magic the way you're outlining it, generally.  Not even the "gives you ambrosia" effect, as it's random, and I didn't want it or ask for it.  Why should someone put up with random effects to get something they could get out of a process where they know the outcome?  Even if ambrosia is a food that gives more happy thoughts, that's no real bonus, as happiness has to be kept up at all times, and can be kept up at all times through mundane means.  You aren't solving any problem that couldn't be solved without taking any risk whatsoever, and as such, you're just introducing risk and taking away control of the game from the player for the benefit of making some "lol random!" effect whose novelty will shortly wear off, but whose nuisance will live on.

That "third option" of magic is the only one that I actually would agree with seeing in the game.  If you start creating magic swords with random effects, but where you could generally prevent the effects from taking place once those swords are created by just plain not using them (or dumping them into the nearest atom smasher or pawning them off on some sucker merchant), and the effects of that sword are known, and not completely random, then it's fine. (If it's a sword that deals extra cold damage 10% of the time, that's fine, it's just not fine if it's a sword that sometimes shoots lighting, sometimes turns the weilder into a bunny, sometimes teleports the weilder into a volcano, and sometimes turns all the target's clothing into salami, then it's basically a Wand of Wonder, and it goes straight into the nearest magic disposal unit.)

The main problem with having random magic swords (which basically sound like the magical artifacts we're getting later, anyway) is that it doesn't really solve any problems, though.  It's just giving you magic swords that may be better than normal swords, but which, as far as I can tell, don't solve any problem that normal swords won't solve.

This is why I keep asking, "What problem does the player have that this problem solves?" 

This is only a meaningful change that a player would actually want to put up with the side-effects of if magic does something entirely new that they can't do with anything else.  They need to have some sort of problem that they can solve only by using magic if you want to make people want to use magic. 

This is the real point I want to get across here: Stop thinking about this from the perspective of what you "think magic should look like", and start thinking about it from the perspective of "how do you want the player to use magic", and then build the magic system so that players have problems that only get solved by using magic that way.

I also might want to make clear what I mean by Alchemy, since what you said about alchemists producing a random new product kind of raised a warning flag that we aren't thinking the same thing about the same term.  When I say alchemy, then I mean something that is an industry, not terribly far off from the sort of stuff we can do now with modding.  I mean that alchemists are dwarves who take aloe vera leaf resin (grown in your farms or traded for) plus ridge root gum (also farm grown) (or you can substitute in cave worm ichor or fluffy wambler pudge for the ridge root gum), and mix them into a poultice that doctors can use when treating dwarves to give them a bonus to their check to stop the bleeding, as well as increase recuperation rates. 

The point of alchemy, as I talk about it, is to weave the ability to perform some special tasks that go beyond what would be necessarily scientifically possible into a psuedo-chemistry (technically, lye and potash are alchemy already) that can potentially create products that have unusual effects that aren't strictly rational or realistic (although many may be) that can be woven into the industrial base we already have.

Now, you could do some randomness in this regard; Since I'm sure it'd make Andeerz squee, we could have some procedurally generated "alchemy levels" that represent what sorts of alchemical knowledge players have access to, and players can be forced to make do with what plants grow or animals live in the specific biomes they live in, excepting trade with elves.  This would be the sort of "card game randomness" that you are given a different situation each game, but where the actual effects of your actions are always known to you.  A medical poultice always increases healing rates, although I suppose doctor skill could have some dice game randomness effect on just how much it heals the patient dwarf.

Making dwarven alchemists means that the "magic" IS an industry, and not a bunch of "dwarven wizards".

Again, by making a claim that "magic shouldn't be an industry", you are proposing a change to one of the fundamental aspects of DF, where everything is an industry, and the burden of proof's on you to prove that you have explored all the consequences of such a decision, and how that will affect play, and why such a change gives the player more to enjoy than not making that change. 

So... prove the statement "magic shouldn't be an industry". 

Again, I'm certainly not convinced by a proposal that seems to suggest that we take a game where our control is so indirect that we have to take preemptive action to cut off every possible problem, rather than have the ability to react to problems, and instead give the player a mechanic that is a functional crapshoot where the game just keeps rolling the dice, and the punishment for crapping out is a random event that messes with our plans.  Even if it it's ridiculously slim odds, any odds for a demon appearing in your dining hall is something you will have to prepare your fort for if you want your fort to survive, because it's obviously going to happen sooner or later.  Even if it isn't very major, if the game stops every time your beer turns into ambrosia, and you have to divert attention away from whatever it was you were doing in order to babysit the wizards for a little while, it's still just a nuicance for the player.  Frankly, it's just a crapshoot... and I've never liked gambling, as I've never liked games where I wasn't in control of the outcome. 

Where's the fun in being subjected to random magical effects (even if they're GOOD) you can't control?

[NW_Kohaku]

Actually, another example to illustrate what I mean:

In Dungeons and Dragons 1st and 2nd editions, the magic users were all pure Vancian magic users.  Maybe they did this because of a specific desire for the sort of gameplay that encouraged (pre-planning what spells you "pack" as "ammo" for that day), or maybe they just did it out of homage to Jack Vance. 

When they were getting ready to make 3rd edition D&D, however, they decided to change that - they obviously wanted to cater to people who wanted a playstyle with less forethought, more big booms right here, right now, and so they created the sorcerer as an alternate flavor of wizard.  Sorcerers were wizards with no need to pre-plan their spells on a daily basis, because they couldn't, although they did require some thought as to what powers they gained on level up, this was generally not fairly difficult thought, because as they were going to be shackling themselves to those spell choices for every single encouter they got into, it invariably turns into many of the same general choices of direct-damage spells and very broadly applicable utility spells like Fly or Haste, while the niche spells are purely the domain of wizards.

They made this choice completely before they made the choice of how to actually integrate that into the existing storylines and narratives of the game at large - how the game was played was far more important than the narrative, even if the narrative was also very important.  They pretty much just crammed a way of saying "Oh, hey, for some odd reason, people are starting to cast magic in ways we've never done before, and there's no particular reason why it wasn't always this way" into some of the books and storylines.

This is the sort of mindset that I want to encourage - "magic" is something so utterly mutable that we can rewrite it however we want into the storyline, and there's no reason we have to do it the way that anyone else does it.  We just need to start from what most makes sense from the perspective of the options it gives players, and what will be most enjoyable for them to play.  We can backfill the hows and whys of whether it's magic, tech, or magitek later.

[Bauglir]

-snip-

[KillerClowns]

What we have is a difference in perspective.  I'm going to reference player archetypes from this article on Magic the Gathering.  These archetypes actually work well outside of Magic the Gathering, so bear with me.  I'm actually bending them a wee bit from their original definitions, due to a change in medium, but the spirit is the same.

For our resident Johnnies and Spikes, self-imposed challenges are enough.  A Johnny will decide to build a trapless, boozeless fortress with no miners, make it work, and sing a happy song when the thing survives five years.  A Spike will construct a fortress fueled by the churning of a thousand perpetual motion devices, with the achievements of the mighty dwarvern nation carved into the wall, the bones of goblins and elves strewn about the countryside, and a vast economy that renders the single fortress more valuable than the combined products of every elven nation, and several human ones, combined. But me, I'm more of a Timmy.  I like to experience things that haven't been seen before, to try things that have never been tried and see what sort of fun results.  Unfortunately, I'm running out of options here.  I've crawled through the underground and seen pretty much all there is to see, dissected enough Forgotten Beasts to figure out the general algorithm that generates them, wandered around several Demonic Fortresses, and determined that Hell is really not a very exciting place.

In short, here's the problem I'm addressing; there's not enough mysteries left for me, and others like me in Dwarf Fortress.  I like mysteries, which is why I believe making magic chaotic and mysterious would be a good decision.  I oppose industrializing magic because then it would have no mystery, no surprises, nothing of the sort that I so love.  I would enjoy seeing what sort of wacky mischief, good or bad, my magi came up with, and the clean-up if things went wrong would be a price I'd happily pay.  That's just how I play the game.

Of course, you can't build an entire system around a single archetype, so I shall take a moment to address ways the needs of the other two could be addressed.

Even if it's a purely irrational decision, some Johnnies would deal with magic to prove they can.  But let's give them new possibilities to toy with instead.  Rather than "I'm going to build an entire fortress out of steel," they could say, "I'm going to bind every type of demon in the world into servitude and make a Zoo of Hell!"  Why?  Because it's that extra little, "I told you I could do it."

Kohaku, I'm guessing you're primarily a Spike.  I could be wrong, of course.  But your questions are in the vein of "what would dangerous, unpredictable, irrational magic do for me and my fortress that reliable industry would not?"  This is a legitimate question that deserves a legitimate answer: damn near nothing, I confess.  So I suppose simplest solution is to divide magic into alchemy and wizardry; the former has predictable and useful, but not especially impressive, effects, while the latter is a gamble.  Any mage that showed up in a Spike's fort would be given an alchemical laboratory, and would never see a summoning circle in his life.  He'd quietly and efficiently produce potions, poultices, and other such things, and his only experience with the wilder and more dangerous forms of magic would be in defeating them when it comes in externally, ideally via the card-game randomness you spoke of.

Yes, that's right, now that I've genuinely thought about the question from the perspective an archetype different from my own, I realize why you were so appalled by my suggestions.  Now, try and see things from my perspective; a playstyle based on gambling, self-imposed challenges, and exploring possibilities is just as legitimate as building the best fortress you can and amassing wealth through reliable, practical means.

[Jothki]

Magic doesn't necessarily have to be about production. What if instead, it took the form of subtly altering the parameters of the environment? There wouldn't be wizards flinging around spells, but rituals, events, or just random drifts in the flow of magic could result in properties being overlaid on the fortress.

To draw on the existing ideas of divine spheres, if your fortress worships a god of fire, works with lava a lot, or has a lot of things catch on fire, the land itself might become imbued with the essence of fire, altering the weather and lowering the point at which things catch on fire, but increasing the effectiveness of smelters and forges and making dwarves that like fire happier. If your fortress worships a god of death, or a lot of creatures die, the land might become imbued with the essence of death, making dwarves die more easily and undead appear more frequently, but making butchering more effective and dwarves that like death happier.

That kind of thing would be somewhat subtle but would also have the potential to be quite significant. Depending on how significant drift is, it could also serve as a source of the unpredictable disasters some people like. Perhaps altering a world gen parameter could affect how inherently magical the world is.

[Eugenitor]

You know, if you want to have magic with both downsides and upsides, they should be intertwined. Not random at all- you just can't get one without the other.

For example, the cursed sword that regularly calls its wielder a faggot should be the sword that one-hits demons and chops goblins to goblin stew.

If you know what you're doing, it's a powerful tool in your arsenal, and you can mitigate the downsides.

If you don't know what you're doing, you have Fun.

[Tokkius]

I'm in favor of procedurally generated magical effects. Magical abilities for critters, artifacts and locations would be great. Spells, however, I'd prefer kept out of the game.

[StephanReiken]

My heart, long held by the twin vixens of Thief and Arcanum, is invariable led towards the steampunk option.

My thoughts exactly