Alright, we have a mission statement for Four Realms, but that and ten dollars will barely buy you a coffee.  How about a skeleton for the system.

One element that is both a strength and a critical weakness of D&D and many other gaming systems is “character classes.”  On the one hand, it can be convenient to have a handful of archetypal roles to fill; it allows quick character creation and an easy guide for behavior during play.

On the other, there are so many archetypes, and so many variations possible on the theme of each one, that the system runs into trouble as soon as someone comes along who thinks it would be interesting to play something other than a standard version of one of the provided archetypes.  Do you refuse to allow it, thus losing their interest over a relatively trivial issue?  Or do you provide a wealth of possible classes and/or customization options, resulting in system sprawl and losing any benefit of having imposed defined archetypes?

So no character classes for Four Realms, then.  No “levels,” either; they can be a handy mechanic, but only by sacrificing a lot of verisimilitude and flexibility.  Why would learning how to swordfight make you better at knitting, juggling, and price-haggling all at the same time?  Each character should be defined mechanically by nothing other than their innate abilities and their acquired skills.

On that note, I’m not against a system of “advantages” (e.g. ambidexterity) and “disadvantages” (nearsightedness).  However, first – that kind of system would be one of the “optional layers of rules” I mentioned last time, and second – it’s a system ripe for abuse if you’re not careful.  I’d have to find ways to put limits on it, for example tying each (dis-)advantage to a specific set of possible trade-offs.  We’ll see.

I dislike the term “core mechanic,” but the “d20” system of recent D&D iterations does have the benefit of simplicity.  One part that strikes me as odd about depending on a single d20, though, is how entirely random it is.  A staggering failure or overwhelming success is just as likely as a run-of-the-mill result.  Fortunately, there’s a solution to that: roll more dice.

If you roll more than one die, you get a bell curve of results.  The more dice you roll at a time, the higher the center of the curve is and the more quickly the outskirts attenuate.  I like the span of a d20, so (for the time being; it may change) let’s say 4R will use 2d10.  Most check results are going to be average-ish.  This means there’s a relative emphasis on ability over luck; the trick then is to keep static modifiers, such as situational bonuses, in check so that they don’t become overwhelmingly large and take out any uncertainty or excitement.

One other thing: the vast majority of checks, as I see it, are going to be skill checks.  If you want to perform a (non-automatic) action, you roll 2d10 and add your skill level and any other relevant modifiers.  But what if there’s an action that depends on nothing but strength, such as shoving a heavy object out of the way?

It would be easy to use the same 2d10 mechanic, with lower target numbers, but for some reason I like the idea of needing to roll low: if you have a Strength rating of 12, you roll a die and 12 or lower succeeds.  It seems to me, though, that a bell curve would work against you here.  Players with characters of below-average rating could get discouraged and stop trying.  So I’ll leave that as a straight d20 roll.

So there you have it.

1. Each character has Abilities and Skills that define everything about what they do and how (and how well) they do it on a mechanical level.
2. Abilities are checked by rolling a d20; a result equal to or lower than the ability rating is a success, and higher is a failure.
3. Skills are checked by rolling 2d10 and comparing the result to an opposing roll or a target difficulty.

Next time: what Abilities should the system track?  And here we’ll really see me trying to balance between versatility and simplicity.