Extended Roll Nightmare

*: That raises a whole other issue, that of Extended Rolls.

Some RPGs are too vague. There's no answer offered to questions like "how fast can I run?"

Others are too specific, offering precise mechanical formula to derive speeds (and other minutia) in a way that results in there never being a contest because the character who's just one-point better always wins. These tend to suffer from one of two issues:

1. Disconnect from Drama / Disconnect from Reality: Difficulties are being set by formula, in a vacuum where the PCs stats aren't taken into account. Sometimes this results in a good plotline or obstacle idea being rendered insurmountable because the magic formulas don't perfectly match reality. Often this is a result of not enough (or too much) gradation in the stat range for a normal human. RPGs are at best an imperfect simulation, after all. As a result, it becomes very easy to improvise your campaign into a corner: If I say "the wall is made from bricks, and stands 50 feet high and 5 feet thick" it has a quantified difficulty to climb or topple, whose interaction with the PCs stats will be outside my control if I didn't do my homework first. Such a system encourages the GM to play "by the books" cause the players will notice if he veers to far afield.
2. Infinite Extended Rolls: The way around that is often presented as extended rolls. When I first encountered Extended Rolls, I thought they were nifty. It was a cute twist on dice mechanics. But it's very easy to go overboard. Case in point: Scion.

You knew I was about to criticize Scion, didn't you. While the above was a general rant about RPG mechanics and philosophy, the following is honed in on a single game, and in particular one system within it. I feel a tiny bit bad writing this. Conceptually, Scion rocks, but with every session I grow more dissatisfied with the mechanics. (With the exception of 2 strangely tense sessions, I really enjoy the setting and characters, so I intend to carry it through on my planned story-arc all the way to Legend 12 even if it hurts sometimes.)

There's a PC in my Scion game who's a bit of a weaponsmith. The game doesn't really offer any guidelines on how to handle that, especially pre-Godhood. The first weapon she made, I let it be a little too good - she got 11 successes, and since the best mortal weaponsmith couldn't score more than 10, I figured I'd let the dagger she made be damn good. Too good. It was quickly overshadowing the 3-dot Relic that was her primary weapon. I had to bump up full-fledged Relics a tad at Demigod in order to compensate.

Then I worked on that Hitori Hanzo system. This was an improvement in that it gave structure to what PCs could build, but it unleashed the ugliness that is Extended Rolls. The PC in question could make a 1-dot Sword in 8 to 16 hours and 1 to 2 rolls - that sounds just fine. She could also make a 5-dot Sword in about 72-88 hours of work, via 9 to 11 rolls. She literally can't fail, so there's no reason not to set her sights high. However, do we really want to just sit there rolling a two-handed dice pool 9 times when there's no drawback to failure? How boring. At that point, we might as well just say "you make a 5-dot relic every 80 hours, automatically". But if we do that, the PCs have a never-ending supply of beefy weaponry, and balance (between characters as well as PC vs NPC) becomes shaky.

And just in case anyone reading this doesn't know: an Extended Roll is really multiple sequential die rolls with their results added together. Typically it's used (along with a high difficulty) to represent lengthy processes, such as research, smithing, or warfare, abstracted out to a roll ever X days instead of fiddling constantly with daily minutia. The idea is that it occupies a middle-ground between "one roll summarizes your life's work" and "rolling once per combat round or discrete individual action". Often, though, these systems aren't aggressive enough. They either mitigate overall failure into non-existence, or they result in dozens of tedious rolls, or both.

So perhaps I could use the system in Scion: God. Maybe not. That system conveniently makes all Relics just involve 2 die rolls - one Intelligence, the other Dexterity. Planning the Relic, then following through. On the surface, it sounds great. But it has two problems.

1. The difficulties are 35 and 50, with no concise statement that it is or isn't and extended roll.
   If this is not an extended roll, it means that making even the simplest magic item requires 7 dots of Epic Intelligence and 9 dots of Epic Dexterity. She'll likely never be able to make a Relic on that system, as she lacks Arete and currently has no plans to ever buy the 9th dot of Dex. This doesn't fit well with the commonality of Relics in the setting - clearly it can't be that major of an ordeal to make a Relic, or else a PC shouldn't be eligible for 15+ of them in the course of attaining Godhood.
   If it is an extended roll, then, per the Hitori Hanzo paragraph, she can currently make a Relic if she just rolls about 7 times. Not only would that have the aforementioned balance problems, it contradicts the notion of Hephasteus and the Cyclopes making all the Greek items for the other gods.
2. No power-level differentiation. A 1-dot Relic has the same difficulties as a 5-dot Relic. The material components are the same unless the later provides access to multiple purviews. In fact, Purviews seem to be the only thing the author of page 79 of Scion: God deemed worthy of specifically addressing or restricting, despite there being things (speed reductions and unique powers) that are rather more directly potent than simply granting Purview access.

So, it's time to look at that drawing board again. I need a Scion Relicsmithing system that:

• Involves no more than 2 or 3 rolls per Relic.
  
• Differentiates between Relic powers.
• Has a chance of failure, especially if you ambitiously overstep your talent.
• Isn't restricted to being used by min-maxed forge gods. (Heph & company can make really good Relics, but he, Weyland, and the 'Clops aren't the only Relic makers in the universe. Or at least, they shouldn't be.) Such Gods should merely be better at it, being faster, or somewhat less prone to failure, or perhaps capable of making overall more impressive items.