Subject: Re: What Should Magic Items Do? Date: Thu, 22 May 2003 23:23:56 GMT From: bear@sonic.net Group: rec.games.roguelike.development Peter Knutsen wrote: > > bear@sonic.net wrote: > > I'm starting to experiment with magic items. > Magic items, whether for tabletop RPGs or computer games, have a huge > potential, but very few games makes use of it, sticking exclusively to > boring, generic stuff like Swords +1 or Arrows of Fire Brand. Yeah... the problem is difficulty of implementing it. Really good magic items are profoundly creative in a way that's very hard to script. > > Magic Weapons are pretty unambiguous, although I'm > > They don't have to be. Well... It depends, I suppose, on the type of ambiguity you prefer. I like the sharkbow because it does something that's usually good but sometimes dangerous. I also like ideas like weapons that vampirically regenerate the wielder, bladed weapons which are really a kind of animal where the hatchlings can be weilded as daggers and the adults become bastard swords - which are good, but which you've gotta feed and care for... I like the idea of 'spirits' tied to a weapon or item that may or may not do what you want at a given moment. I like the idea that the use of some types of artifact weapons may have side effects that go much further than just the user himself. Picture a paladin fighting with a weapon when he knows that everything he kills strengthens an ancient enemy of his diety, fr instance. But all of that stuff requires actual roleplaying and nuanced response to make it not suck. It's not a simple matter. > In general, I'd advise high pulse costs for everything, because they > you can get in fine distinctions; instead of Boots -1 Movement Pulse > being a huge benefit, they just let you move 10% faster. When you want > huge benefits, like from quest Artifacts, you can just give a bigger > bonus, like a Sword -5 Attack Pulses, letting the character attack > twice as fast as normal (assuming 10 pulses is the normal cost). Actually number of pulses for walking is 1.0 (a floating point) orthogonal and 1.41 diagonal. pulses required for an attack range from 0.75 with daggers to 1.5 with great axes. Dealing with floats, a "high number" doesn't really matter; it's as divisible as I care about. > It is a bit different for roguelikes in that the expectations of loot > are higher, particularly because loot is so random - a Troll might > drop a Wand of Magic Missile, even though it is very thematically wrong. Oy. Haven't got there yet. But yeah, treasure types should suit the monster. > > do they just focus your mana so you can cast blind > > or without books or in other adverse circumstances, > > Or cast faster? I think casting time is a very underused axis of > differentiation. In Angband and Zangband, all spells take 1 Round to cast. Good point. > I like the idea of Foci, which makes one type of magic easier, maybe > by both lowering the fail chance and lowering the mana cost. So you > could have a Focus for Fire Magic or a Focus for Healing or Illusion > Magic. Maybe make it wieldable so that the character has to choose > which Focus he currently wants to benefit from...? Or maybe you have to make a choice whether your primary weapon is a staff or a meelee weapon? > One advantage of systematized systems is that you can give items > general stats, for instance in my tabletop RPG system all items have > an Activation rating, which shows how reliable the item is. Usually > you'd prefer an item that is very reliable, but if you've found this > not-quite-reliable Wand of Fireball... are you gonna use it? Most > players would only use it in emergencies. Good point. Hadn't considered that yet. > But isn't your game going to contain randomly generated items? If it > is, that's the big challenge, to make an algorithm that generates > appropriate items, often according to a theme ("fire item", "frost > weapon"). I'd started to consider this, but I don't know how to do it yet. > Secondly, are you going to let the player create his own items? Probably not. At least not in the first edition. > All the new CRPGs based on Dungeons & Dragons 3rd edition also ought > to let the players create their own items, as the new rules are very > explicit about how one does it. I remember D&D, although I played AD&D. 2nd edition was after my time, although I bought a few 2ed books. But I really didn't care much for the way magic was handled in that system. I thought Hero Games and Iron Crown Enterprises both did better, although in different ways. > > "Really destroy the One Ring?" > > ... > > "You cannot destroy this object! -more-" > > "In your pack: The One Ring {indestructible}" > > ... > > "You drop the One Ring. -more- " > > "The One Ring disappears from the floor! -more-" > > "You feel something in your pack! -more- " > > "In your pack: The One Ring {indestructible, cursed}" > > ......... > > Is that an example of item scripting? I think it sounds neat, but the > game would feel a lot better if most items were just items, and only a > few ones had any "behaviour" or "attitude". It's possible for item scripting. My monsters have a bunch of hooks where you can hang scripts on them, such as "On acquiring line of sight" and "On taking damage" and "on death" and "on healing" and "on closing for meelee", etc... and for each trigger there are a couple of things in registers that the system can check. So if you hit a monster with a fireball, its scripted routine for "on damage", if it has one, can check the "damage type" register to find out that it was fire and the "damage amount" register to find out how many points. If it's a fire elemental, its script will tell it to regenerate instead of taking damage. But if it doesn't have an "on damage" script, it just uses the generic one which subtracts hit points from it. The above would be item scripting, assuming that you have triggers on items for for "on destroy" and "on drop" and "on wield", and that script actions can add to item inscriptions. I've already implemented an "on firing" trigger for the sharkbow, but simple pluses-to-hit don't require a scripted action, so there's not even an "on-strike" trigger yet, let alone some of these other things. > > And somewhere in here, I need to address the question > > of the kind of feel and balance I want in the game. > > And the duration you want. Angband, for instance, takes a really long > time to play. Nethack is, reputedly, much faster. I'm going for user-selectable. A given game is as long or short as the player wants it to be. If you give your game a very short name, the system will make it a five-level dungeon with maybe some kind of 'Svart Queen' or something as the boss monster. If you give your game a very long name, the boss monster could be a couple hundred levels deep and nastier than any of us really understands how to think about. Bear