I'm working on a game system (currently called "Hack") that began as a D&D "fix" though one thats extensive enough that it would be neigh unrecognizable such (no d20's, no classes/race lists/levels, etc...). The point is to create a system for my group (and potentially for later publication) that meets our needs in running games in "home brew" fantasy worlds. The core mechanic is solid enough and very streamlines/abstract. What I need are suggestions on introducing the kind of "crunch" that the people I play with seem to be addicted too without derailing the universal/abstract (and even though I loathe the term) "cinematic" nature of the rules.

What I'm looking for are all those little powers ("talents"/class features/spells/feats) that give players "hand holds" for characterization. I think one of the problems with more "broad approach" mechanics and character creation/quantification mechanics is that they make it difficult for some players invest themselves in their characters. When you have a named and numbered power, like "Two Weapon Fighting" or "Rage," its easy to immediately see how your character might act in a combat situation and how it will play out mechanically. Strip away the "named and numbered" approach to character abilities (where the system specifically states what the character can do) and players become more unclear about what they should do, even if the system doesn't preclude any of these activities.

So how then, do you put these kinds of "hand holds" in place without restricting options (if you say character X can do Y, then it stands to reason that character Z can't, or at the very least, not as well) and without having them become the forefront of the system in play. Two particular pitfalls I want to avoid are "the shopping list" approach to problem solving, where players look over their character's named and numbered abilities, looking for one  that obviously fits the situation, and if none do, giving up, and the type of ties that a game like D&D puts between story elements and game mechanics. For instance, a character with significant healing abilities must be religiously devout (or I guess, devoted to an alignment). I want to make sure that the system doesn't accidently creep into being an "assumed setting" like D&D and its closest imitators.

First off let me just congratulate you on formulating, in a very theory-esque way, why I like systems with a little crunch in them.

One way of solving your dilemma: Don't have any Merits/Feats/Qualities concerning being able to do anything a real human can. What I mean, is: Don't have a Feat stating that you are able to disarm opponents, or a feat that says that you can use two weapons, or a feat that says you can kick people in the groin. If you want to have feats that deal with these things, have feats that make the characters better at doing those things; Everyone should be able to try to disarm an opponent, but a person with the right Feat might be better at it. Do have feats concerning being able to do things real humans can't.
Also: If your system contains specializations or something similar, you might want to avoid having the same bonuses as Feats (because it just increases bookkeeping). What you can do instead, is have Feats that give skill bonuses, give bonuses to several related skills, but only under specific conditions. You might, for example, have a Feat called "Puppy eyes", which gives bonuses to Persuasion, Socialize, Empathy and perhaps even Intimidation when the character is trying to avoid getting the blame. Another Feat might be called, for example, Prophet of Wikipedia. The character gets a bonus to any knowledge skill roll, as long as the character has no use of said knowledge in his chosen profession. A mercenary could get the bonus if he's trying to impress a girl with his knowledge of poetry, but a writer wouldn't.
Since you described your system as (even though I also dislike the term) cinematic, this take on Feats might work especially well because it enforces a kind of movie/book-style niche protection; it doesn't really give characters bonuses based on how - or maybe why - they accomplish stuff, instead of what they actually try to accomplish.

I feel like you answered your own question.  Create a general mechanic for how each *type* of talent works, mechanically, and than create lists of how you can approach using those mechanics.

For example, you mention two weapon fighting, rage and spells.  Create a mechanic that handles "Delivering an attack" and than make a list of descriptors for your attack. Mechanically, they all work the same, but the character decides their own "style" by calling it "Raging and Greatsword" or "Bravado and Rapier." Then, make a mechanic that handles "Delivering a spell" and make another list of descriptors etc.  Let the players choose any number of descriptors, but give them no mechanical effect.  

I would say, though, that your players are looking for the synergies and systems they can create through mechanics to accomplish a combat strategy.  One player might be looking to, "Make many attacks" while another player might be looking to "Slay enemies in a single blow."  In order to appease that kind of play, you need a complicated system, like D&D, where each "descriptor" has a mechanical effect.

Walruz & David:

Both of the examples you give (Walruz's which offers (if I'm correct) adding a list of "meta-traits" that plug into the existing statistics in various ways) and David C's (descriptor only, no mechanical effect) both encompass to traditional approaches to the issue, and my problem, really, is to find something new. Plug-ins, like Walruz suggested, might be the only real solution to the problem, while the "example description" (David C) fail, IMO, to offer anything at all to creative players who could as easily come up with such examples themselves.

Is there a middle ground? Something between a mechanically unrelated description and a list of plug-in/meta-traits?

Also, to clarify what I'm looking for (and to talk a little theory): the game I'm creating is intended to facilitate Simulationist play, and that has a lot to do with both of the suggested options are unattractive, because it seems to me that they could easily lead to an abashed, unfocused system.

Specific "named and numbered" traits, though sometimes meant to be realistic or representative of genre specific conventions, have a serious problem in that they are hard to balance. Now, balance may not be a focal point in Sim play, but I would also describe my desired play style as "Protagonism" which is to say, I want each and every player character to feel important and central to the story, and mechanical feebleness can quickly lead to a feeling of player helplessness, which then encourages the player to put more and more effort into "mastering" the mechanical aspects of the game, which leads to a since of pride and accomplishment from mechanical success/dominance which lead to a more-or-less Gamist agenda/mind set. I don't wont the players to feel either rewarded or punished for purely "game level" decisions.

(Sorry about the "fear leads to anger, anger leads to hate, hate leads to the Dark Side" set up there, just wanted to explain my thought process)

The other option (purely descriptive) simply lacks the tooth to give the players anything to hold onto.

I hope I grokked what both of you were saying and that my response seems to address your suggestions. I'd appreciate any further thoughts on the subject.

The reason I suggested what I did is, anytime you marry mechanics to fighting styles, you encounter 1) balance issues 2) synergy issues (A = B = C = D, but A + B > C + D) which ultimately leads to the situation you described,

and mechanical feebleness can quickly lead to a feeling of player helplessness, which then encourages the player to put more and more effort into "mastering" the mechanical aspects of the game, which leads to a since of pride and accomplishment from mechanical success/dominance which lead to a more-or-less Gamist agenda/mind set. I don't wont the players to feel either rewarded or punished for purely "game level" decisions.

masqueradeball wrote: Both of the examples you give (Walruz's which offers (if I'm correct) adding a list of "meta-traits" that plug into the existing statistics in various ways) and David C's (descriptor only, no mechanical effect) both encompass to traditional approaches to the issue, and my problem, really, is to find something new. Plug-ins, like Walruz suggested, might be the only real solution to the problem, while the "example description" (David C) fail, IMO, to offer anything at all to creative players who could as easily come up with such examples themselves.

Is there a middle ground? Something between a mechanically unrelated description and a list of plug-in/meta-traits?

Also, to clarify what I'm looking for (and to talk a little theory): the game I'm creating is intended to facilitate Simulationist play, and that has a lot to do with both of the suggested options are unattractive, because it seems to me that they could easily lead to an abashed, unfocused system.

Specific "named and numbered" traits, though sometimes meant to be realistic or representative of genre specific conventions, have a serious problem in that they are hard to balance. Now, balance may not be a focal point in Sim play, but I would also describe my desired play style as "Protagonism" which is to say, I want each and every player character to feel important and central to the story, and mechanical feebleness can quickly lead to a feeling of player helplessness, which then encourages the player to put more and more effort into "mastering" the mechanical aspects of the game, which leads to a since of pride and accomplishment from mechanical success/dominance which lead to a more-or-less Gamist agenda/mind set. I don't wont the players to feel either rewarded or punished for purely "game level" decisions.

Walruz,
Personally, I'm afraid of player defined traits, because it amps up the level of "trust" needed to play. By trust here I'm talking about the ability to abuse the system... which, with player defined traits is normally extremely easy to do and often unintentional on the part of the player. I imagine at this point like it sounds like I'm just shutting down people's ideas, but I think thats maybe because I'm not being clear enough in my desires/expectations, or about how my current rules system works...

As to what you said about Simulationist design being "unfocused" I'm not sure I agree... what I'm talking about with Sim is simulating a certain tradition within fantasy literature... namely medieval fantasy focused on the exploits of a small group who's internal bonds are as key to the story as the quest itself (examples: Lord of the Rings, Dragonlance... and though not medieval or "fantasy" Watership Down, Star Trek, and Stand by Me all follow the same basic structure and content). So I want the game system to be focused in that it doesn't promote (or even allow for) Gamist or Narrativist interpretations and that it supports the specific expectations of the genre its trying to emulate.