I'm actually not using the CD assignment as the first step. I'm simply looking for a rough formula which can be used to determine what a pure damage ability should deal in a vague sense as a guideline for setting damage values while designing the intended attack strings in a holistic manner. The closest I've gotten is that all pure damage attacks start off with a baseline 1.0 modifier (equal to the base damage dealt by your basic attack without factoring in procs) and a percentage of a further 1.0 modifier based upon the percentage of the CD compared to a 15 second "generic" CD (6 second CD, which is 40% of 15 seconds, would equate to *roughly* 1.4 "basic attacks" worth of damage). This is a *very* vague guideline since it assumes that the damage is dealt roughly equally across most attacks (so it works for classes like Consulars and Knights but not as well for Troopers which tend to deal a *lot* of damage within a small number of attacks and rely upon their base attack more), but it works to some degree.
This is the whole reason why I suggest that you need to pull more time into the calculation to determine the "value". Without including a long period of time into the calculation, you don't really see what the overall effect of the ability is. If it's a 15 second CD - it's a 4 cast per minute. And while it might do signifigant damage per use, it doesn't influence the dps much over the course of a typical battle. DPS abilities with long CDs are intersting to use, but don't necessarily effect balance very much.
I'm actually not paying attention to casting time since it's more of an abstract applied to lengthen the GCD and moderate damage. When looking at abilities, I'm actually looking at damage per GCD and only within actual practical application (so I'm not looking at TkT for Shadows except for Kinetic Shadows and, only then, when you have 2 stacks of Harnessed Shadows; since it's a 3 sec cast, I'm functionally dividing the damage by 2 to represent the fact that it consumes 2 GCDs).
I also recognize that it's a holistic assignment in the first place, so it's not like CD is deterministic for damage or vice versa. You have to consider the whole of the attack rotation/priority rather than the individual abilities themselves. I'm simply looking at a guideline for a baseline,
Maybe the basline you're looking for is just an average dps number from a parse ~ uninfluenced by AC, gear, or spec choices. Alternatively you could just average together a large number of "naked" parses from different ACs. You could probably just assume that TOR actually works in clock time units of 0.1 seconds, and ignore the hundreths. The engine may not actually be using true hundreths anyway, considering that would be above the resolution of the human eye. So if you just convert everything to dpTs ~ everything fits nicely. Divide longer casts by 10X the cast duration, and instants by 15. At least for damage you get a nice number, for quick comparison sake. But I still think you need to consider how frequently an ability can be used in a practical sense.
I'm not really seeing the similarity between HiB and Project: Project costs a crapton to do mediocre damage and a little stun. HiB deals a lot of damage without the stun and costs comparatively less. TkT and Full Auto are roughly similar however.
The differences in damage & cost - are just balance changes. The reason why I compared them, is that they both benefit from increased dps output as a result of spec choices in the dps trees.
There are some "standard" archetypes for attacks that all classes use: Ion Pulse, Double Strike, Slash, and the like that are baseline consumers that deal more damage than the basic attack and have no CD; Stockstrike, Blaster Whip, and Blade Storm that are on CDs but deal substantially more damage for their cost to make up for it.
Charged Bolts actually has more in common with Disturbance for Sages, Double Strike for Shadows, Ion Pulse for VGs, Quick Shot for Scoundrels, and Charged Burst for Gunslingers than it does with Mind Crush. Mind Crush is a "special" attack (generally reserved for proc'd effects making it effective) whereas Charged Bolts is a baseline consumer (since it tends to proc other attacks rather than getting improved by various in class procs).
Exactly - in truth, at the engine level almost all of the instant cast abilities are actually the same ability ~ it's just copy and paste. The only way they differ is in their ranges,damages, and animations. Cetain abilities are given casting times, which makes players choose between using them, or moving their toon. CDs and cost just limit what portion of a rotation they can fill. Sometimes the abilities were given associated buffs, just to add flavor to the game play. Charged Bolts & Disturbance probably originated as "the purple button".
I've actually done all of this already except for the last. Since the mitigation effects of certain attacks only really take place in tank specs, such benefits are factored in as part of the survivability calculation based upon the use rate of the given attack rather than as an abstract attached to the damage/threat balance functionality of the attacks. As such, rather than trying to find some way to balance the 5% acc debuff on Force Breach for Force Sweep for Shadow and Guard tanks, it's simply considered to be a static benefit attached to the ability in a mechanically arbitrary manner: it wouldn't matter which ability it was tied to as long as it was tied to *some* ability such that it could maintain the same uptime. The quantification of non-direct mitigation benefits (such as KBs and conditional stuns) are less significant than you might imagine, only really being present for flavor (no one really cares that Project, Force Sweep, or Full Auto ends up stunning the weak and standard targets since those targets are largely unimportant and can be beaten by just slamming them with your basic attack ad nauseum; it's just a tangential secondary benefit that makes some degree of sense to have on certain attacks without really making a substantial balance difference; the "real" stuns are less concerned with their damage and actually follow pretty standard formulas for duration, range, damage, and cost based upon the PvP balance equations).
Threat doesn't really need to be considered when it comes to balance, so long as the dps of the tank specs is balanced ~ threat will be by default. I agree with you about the stat boosting passive abilities ~ you used accuracy as an example. The tech trees are full of these, with a crit increase/damage increase /cost reduction for all sorts of active abilities. You're also right about the mitigation effects that only occur against weak mobs in PvE - those effects that only work in PvE vs mobs ~ they don't have to balance ~ no one would notice if they didn't. But not all of those abilities are fringe benefits, the snare effects from Tk Throw or Weaken Mind for example. Both of which do have an effect on PvP balance.
The way I look at it is this.
(Mob damage output - player mitigation - player healing done) - (player damage output - Mob mitigation - Mob healing) = X
If X = 0 the encounter is balanced. Otherwise a victory condition is defined before the event occurs.
Passive mitigation of any type can be a huge part of this. It could be a tanking spec choice, a class ability, a stun/ root/ snare, a buff from an ability used, or just using range against a melee oponent. And in well balanced content, X is going to be close to 0. We consider helaing and mitigation to be sepperate, because we're used to the "trinity" structure. But there never really was a trinity, its always been just "doing damage" and "offseting damage" all along. What makes mitigation different is that it's largely passive, while healing is active. Force armor is considered a heal, because it's spammable on large group of friendly targets, but it's got more in common with a defensive CD.
The variation in damage type and attack type is actually relatively simple to quantify: using the standard ~35% K/E DR that ops bosses get allows you to "increase" all I/E damage by a factor of ~1.5 to reach the damage equivalence; attack types follow a similar formula insofar as you can roughly quantify the chance to miss (or itemization lost to make up for said losses) and apply a roughly 5% higher damage to those attacks to reach vague parity.
So striike that variable... the work has been done.
And I highly doubt that any MMO class was designed purely in the abstract without ever first being built upon a foundation of paper. The pure theory represents a baseline of performance and provides a structure for the remainder of the class to grow through the organic process of playtesting and simulation. The quality of said pure theory development has pretty substantial long term effects: just look at Guardian tanks compared to VG tanks or Shadow tanks; it's readily apparently that Guardian tanks didn't really have much in the way of on paper development because most of their functionality is just thrown together from ability archetypes in previous games and then modified later on to achieve numerical balance without much thought towards how it actually plays. Well thought out classes are readily apparent based upon their elegance of design. The intent would be to design a class with abilities that work together in a manner such that the end result would work together seamlessly without the rough hammering that would be required otherwise.
Design on paper - is abstract... but we're on the same page here. Knight was something that was hashed out from the same basic template of abilities, and given a unique resource mechanic just specifically to make it different. The class could possibly feel a lot better if the focus mechanic were changed to a large pool instead of rough blocks. Just to open up the ability choices a little bit more, and allow for a little better tuning of the focus generating abilities. I don't think it was based explicitly on content from another game intentionaly. But I agree that a clear vision didn't exist when they created the class. And I'm sure the more "elegant" classes served as the baseline that knights were balanced to fit with. (it felt like that when I leveled my knight) That's the real obstacle to creating any new class to fit with an existing design for another. And that obstacle gets larger with every class and active ability that is added to the game.
As far as avoiding the "rough hammering" ~ I don't think that you can predict whether or not it's going to be necessary without just shooting for blandly average. An interesting class is going to feel different because it IS different. And TOR has failed to produce a different feel for all of their possible specs. I don't consider it to be something to dislike the game over, but when you have 4 classes, 8 ACs, 24 specs ~ 3 ea heals/ tanks, HOW are you supposed to make 18 (pure) dps specs feel unique? The shared trees are going to feel different ~ but if you take those away that still leaves you with 14 dps specs. The PvE content is balanced around players in a 1:1:2 ratio ~ but the specs are in 1:1:4.6, and the problem gets worse when you consider the viable hybrid dps specs. This game really needs more tanking & healing specs IMHO. You have to be willing to drop the idea and start over if you're having to hammer on it too much, or you risk reinventing the knight. But each level increase give you a little more room to "hammer" out the dents in the base classes. RotHC is a chance for BW to fix some of the knight's problems, and make the other classes feel even more distinct.