I think the easier way is, instead changing how OB is computed (char. lvl + ranks + stat bonus), use the same than in RMFRP (only spell list ranks and the +20 for center target, I allow a RR to evade this +20), and apply a modifier with the difference instead that -5.
This modifier must be fairly accurate because having some more/less OB in a skill with no development (or training) can give us unbalancing results.
Looking at the 'fire ball' table and comparing with the 'fire attack' table, looks like a +50 modifier is appropiate making an aproximated average for all ATs. See that we are changing the original -5 OB to +50 OB.
I am going to post too the
new 'fire bolt' modifier instead the original -10. It looks like a +10 is appropiate.
I am also very curious about the reasoning for having ball attacks to be trainable in the first place. The PP costs for spells like Triad of Flame becomes really strange if fireball basically does the same but give the character a trainable OB against all targets.
It is because for using the
same table then using the same OB system is the more accurate, see that before 'balls' have different tables, with surely more condensed results in the same range values.
Another reason is, as I said before, because the resistance/weakness against that type of attack can achieve unbalanced results against balls, as they have no real OB prepared for a typical DB.
There is difference between 'ball' and 'triad', in 'ball' all targets must be in the closed are, 'triad' is more versatile as the can be at much distance, only inside the angle range. I think they are not much comparable as they have different uses for sure.
