Thursday, April 24, 2008

A Scrabble-based Magic System

A month or two ago, I was browsing around the various Fudge pages and ran across mention of a magic system that uses Scrabble tiles. I never found a fleshed-out version; just a few discussions of possibilities. But it got me thinking...

Most of the discussion focused on spelling out English words with a limited number of tiles (usually 7 or so) to cast spells. I thought this would be too limiting and time-consuming. But the idea of having 7 tiles prepared at a time, and drawing from a larger pool, was interesting.

In Scrabble, there are 26 unique letters and a blank. Each letter tile (but not the blank) has a score marked on it. Because the score is in the lower-right corner, even symmetrical letters (like O and I) have clear orientations. So, if each orientation of a letter had a different meaning, it would be possible to have up to 4 * 26 + 1 (the blank) = 105 unique representations. Some letters are much more common, and there is a rough correlation between score value and frequency. This suggested to me that it'd be nice to use the score somehow as a casting cost. Therefore, there'd be lots of have cheap, low-level spells, all readily accessible due to their frequency. Expensive, high-level spells would be vary rare (1 in a 100 tiles).

With over 100 representations to work with, I thought first of a full but simple language, complete with verbs, nouns, and prepositions. I looked through the first two or three hundred of the most common English words for ideas. Then I realized someone had probably already done this language-constructing work. As an example, I found http://tokipona.org/, which is a constructed language with a little over 100 words. Examining this made me realize this was too intense--you'd have a to learn a whole new language just to play a spellcaster! That said, if you're embracing the idea that RPGs are a life-style game, and wanted to reveal the magic system slowly over months of play, this could still be a fun avenue to pursue.


Back to the drawing board, I decided to just focus on the 27 unique letters (including blank). This suggested some rune-like combo system of subjects and verbs. GURPS improv/rune magic has 25 runes. Fudge's Gramarye has 22 colleges and realms. And then there's Fudges 4x5, which as 9.

It turned out to be pretty easy to fill in 27 runes. Soon, I was creating 4 variants for each rune, each of increasing complexity. For example, the Earth rune might have 4 levels: 1) sand/clay 2) stone/solids 3) metals 4) crystals. Similarly, the Sense rune (separate from Knowledge) could involve: 1) heightened senses 2) see through material 3) true-seeing (see through magic) 4) scry (see through distance). Perhaps rotating each tile would add one to its basic mana requirement (which is initially its Scrabble score).

By this point I realized I was back to the original problem: 105 representations is too much! And that's before you even start combining them to make spells.

The other annoying thing I was running into was lack of orthogonality. That is, there was frequently more than one way to do something. For example, if Mind is a noun, then should controlling that mind require the Move or Transform rune? Or should there be a separate Control rune for mental actions? Is steam Air or Water? Do you need both Earth and Water runes to affect mud? Even if it's pretty thick mud? I decided that, if I was going to do this whole combo approach, I would just stick with the 4x5 system. (And, actually, I'd consider combining Mind and Spirit into one category, making it 4x4.)


At this point I was getting ready to give up the project as just too complex. But then I realized I didn't really have an idea of what I was trying to do here, other than use Scrabble tiles in some way. I had no overarching vision. So I went back to examine the real Magic System Basics in order to consider my options.

In the last couple months, S. and I have been watching and reading a lot of Dresden Files, which is basically an fun urban fantasy setting. It gels nicely with other similar material, such as from Neverwhere, American Gods, Nightwatch, etc. This gave me a flavor for my magic: I wanted some sort of subtle magic that could pass in an urban setting, but that is still capable of producing the world-shaking spells of myth and legend, and can include any kind of historical magic system.

We've also been playing a lot of Huffy/Fluffy lately, so I decided I wanted a simple magic system--which is almost in conflict with the complexity naturally suggested by a Scrabble system's potential.

I also realized I don't need to use the full Scrabble tile set. In fact, if I did, it'd mean every magic-using player would need their own full Scrabble game! So I cut the set in half. This means a few characters--such as K, the only 5-point letter--might get thrown out... but only if every caster needs the full half-set.


So, the past few days, I've been doodling ideas and finally inspiration hit last night that brought all my various notes together. I will use a single Magic skill to determine casting ability. Magic is an exertion of will on the world, and so it must overcome the inertial disbelief of those around you. (Though I've never played, or even read much of it, I guess this is basically the Mage: The Ascension idea here.) This makes it easy to adlib spell effects--it's basically dependent on how overt and unlikely the spell effect is. This is the core, simple thing used to adjudicate spell effects, and the system could probably just run on this alone.

The complexity--and flavor--is then in what kinds of magic a character knows, since not every character is good at every kind of magic. This is represented by the Scrabble tiles. I'd like to have these gained through gameplay, rather than simply bought with skill points. Each tile corresponds to a magical verb. There are no nouns; those are implicit/provided by the casting ritual of the spell. Each verb has a mana cost corresponding to the tile's Scrabble score. I'm not messing with different tile orientations, but multiple tiles of the same letter can provide more intricate effects (with appropriately higher mana costs). Most spells would involve a single letter, but some higher spells could involve more.

The challenge for characters in this system is to come up with enough mana at once to cast spells. A higher Magic skill gives a caster a couple more points, but the rest comes from special locations, rituals, items, etc. I'm using the magical laws of Similarity and Contact in all of this too.

Anyway, I think the basic ideas are together, but I still have to flesh out all the details. If it works out, this will likely become a possible Zludge/Fluffy magic system. Sadly, I'm no longer using the full information-encoding potential of the tiles, or the mechanic of tiles being drawn randomly from a pool, or 7 tiles being held in the mind at a time. But I think I made the right decision to start with the flavor of the magic and the nature of enclosing gaming system I wanted instead. I think the tiles will be a nice touch just as a physical representation of accumulated knowledge, as well as handy representations of spell costs.

No comments: