I guess it is possible to write a system that avoids or prevents it, but my experiences say to me that if you balance a system to say 50 DP to spend on 25 skills, that when it goes to 100 DP to spend on 100 skills the players seem far more inclined to create accounting nightmares, then complain when they can't do something related to a skill they have.
When you increase the number of skills by 4 but only increase DPs by 2, of course there are going to be problems. There needs to be a specific ratio of skill costs to DPs gained, otherwise you will end up with all sorts of problems with balance and other issues.
Actually, one of the best ways to handle a cat/skill would be to standardize the manner in which ALL categories are learned, in which the bonus from the category is figured.
For example -- Have the skill with the most ranks in it, provide a bonus to the category (which is in turn used to determine skill bonus for other skills.
Have a Subterfuge category that includes both pick pocket and trickery. +100 (say 25 ranks + 25 from stats) in Pick Pocket might give a bonus of +25 to all other Subterfuge skills (i.e. 25 points (1 per rank, in skill with most ranks) + 25 from category stats + -25 for no ranks in skill == +25 total). Then allow the +25 to be used if it is higher than the total skill bonus in any other skill in the category.
This same system COULD easily work just as well for lore skills as well. The bonus in the lores where no skill is had representing things that might have been picked up in the course of studying a given topic (i.e. through references to other docs, conversations with others while learning, etc...)
If the bonuses are too high (which would require testing to determine), then the bonus to the cat can be adjusted accordingly -- we want the other skills to be usable, but not remove the desire to learn those other skills, so a bonus of 0.5 per rank in the skill with the most ranks (or 0.1 bonus from all ranks in all skills in the cat) might be more appropriate.
That is just one possibility of how to create a relatively simple, but effective cat/skill system, and it is just something tossed out off the top of my head (meaning it has NOT gone through any sort of testing or vetting process, so somebody who is looking to shoot it down will likely find a way to do so). The point is, if something like that can be created on the spur of the moment, then the possibilities are extremely good that when ICE selects a team and begins working on the revision in a few years (yes, a team, cause I sure as heck ain't doing it by myself -
), that we will be able to solve such issues.
If you want ot avoid this you can limit the number of ranks a character may purchase in a skill from a given tier by the number of ranks in the previous tier...
Actually, when speaking of Tiers, I am referring to something along the following lines:
T1 = few broad skills (rough guesstimate of about 25-30)
T2 = more skill, still pretty broad, but different types separated out (i.e. Animal Husbandry == Animal Handling + Animal Healing)
T3 = Category/skill system, similar to RMSS, but without quite so many skills and cats overall.
Each Tier would be "hot-swappable", so to speak, thus you could use take a Tier 1 character and easily convert it to Tier 3, thus allowing ICE to use a single Tier for its NPCs and such, and then GMs who like differing levels of detail would be able to still use without problems...
At least that is the
current goal of such a system.