Idea: rather than lock skills, focus on efficiency.
Let a character obtain all abilities, but make it hard to become efficient at all but a few.
Break skills out into categories (mostly as they are today): combat, defense, trading, crafting, building, exploration, gathering, etc. Within each category…
…it takes a short amount of time (few hours) to unlock all the abilities in that category. Aka a tutorial (much like we have today, but a bit more explicit).
The tutorial process is effectively free (maybe the base skills unlock directly via the tutorial feats?)
For example: crafters end up learning all recipes, builders learn to access all chisel types, fighters can use all weapon abilities, etc.
Once the base abilities are unlocked, players can then focus on the efficiency of that category, by investing time and tradeoffs into the categories they are most interested in…
…each category has an array of efficiency skills that players can choose to unlock.
- For example: crafters can reduce the time it takes to craft, amount of materials required, etc. defense improves damage reduction, atmospheric protection, etc.
…skill points are earned as they are today, and each category’s skills become (exponentially) more expensive the more you invest into that category.
…it should be practically impossible to max out all categories in any reasonable amount of time (years/decades).
…there can still be interesting capstone skills here. For example
- give gathers a skill that causes tools to affect two blocks deep
To encourage variety in the player base, introduce the ability for players to choose a small number of categories to specialize in. Each category specialized in reduces the growth curve for skills in that category.
For example, without specialization, each skill might cost
<base_price> * 1.25 ^ <num_in_category>. But with specialization, the multiplier might drop to 1.1 - which makes skills for that category considerably less expensive.
Aka, you become more efficient at learning the skills for the categories you want to specialize in.
Honestly, this seems to be pretty close to the existing system, with the exception of the specialization idea.
Implied: no skill cap, and skill point earnings must grow at a slower rate than prices increase