
Originally Posted by
lyndwyrm;3366566I'm still in the camp of we shouldn't remove accuracy because then gearing would literally go to purely stacking the better weighted off-stats, in which case we should also just remove the 3rd onward ranked off-stats (*cough* parry *cough* speed *cough*).
I'd be fine with changing it, but that would be a [B
LOT[/B] of work on the dev's part.
I'd really like to see the secondary stats revised a bit in general. Less primary stat weight, more secondary stat weights, (and better weight-balance between secondary stats in most jobs) so we can actually feel like they're having an effect, or like we're actually able to experiment a bit. The "it keeps you from adding stats you'd rather have" doesn't really seem like a benefit to me... I like the idea of adding a little accuracy to deal with certain fights, but on the whole, I don't see why we need 640 of it to attack a giant robot when a small, nimble wasp required 400 to hit, nor do I see stacking it as adding to any raid experience; all it does is ensure that we eventually build two gear sets, our highest ilvl set with acc cap for raiding, and our highest ilvl set with as little accuracy as possible for everything else.
On accuracy, from the tank PoV: My favorite system for armor classes vs. evasion included two extra stats on armor. [Note: here accuracy caused percentile damage.] I forget their exact terms but the first, call it 'Balance', controlled how quickly enemy attacks became glancing blows and how quickly glancing blow damage tapered off, essentially the dice roll range for mitigated attacks and their devaluation curve. The second, let's call it 'mass', was actually a debuff component to a lot of heavier armor sets - being larger, certain attacks that may have missed in a slenderer armor set are now instead minor glancing blows - enlarging the total dice roll range, so to speak. The result was that higher armor classes tended to be more reliable, but were a bit less able to 'thread the needle' via perfect evasion.
Different attack types used different balance points, where piercing highest chance to actually hit at all, but had the least chance of being a glancing blow, while blunt attacks had the least difference between full hits and glancing blow strikes, slashing in between. Weapons had further modifiers for 'impact' (a multiplier of attack power that dealt lingering suppression on the target - slowing, delaying, or canceling its abilities or movements), 'penetration' (which multiplied the attack's impact to attempt to ignore armor completely, and 'balance', which controlled how enemy target's armor mitigation curve - how far one has to penetrate enemy armor for decent effect. Penetration and impact worked against each other slightly; the more penetration (damage priority), the less impact (mitigation priority) dealt. Blunt weapons tended to have the lowest pen, piercing the highest, while both blunt and piercing weapons needed relatively little. Slashing weapons, potentially the most devastating, needed fairly high impact and penetration in order to meet their armor-piercing needs. (Note: penetration literally removed enemy armor, adjusting its balance, essentially chipping away at the part that would be glanced - so it made sense to, say, open with piercing (DRK strike-lead), follow up with blunt to enlarge the opening (MNK in pursuit), and rip through with slashing (DRK reaper). On paper it sounds a overly complicated, but in-game it worked just as one would realistically expect. All that's really needed is to keep things roughly consistent, so that these factors need not be re-theory-crafted with each new gear-set or raid tier, if ever.