Umbral

For a while, I've felt that Stone Armor has a few problems. The biggest is that the set is weaker than it should be outside of Granite Armor and stronger than it should be while in Granite Armor. It creates a binary set composition wherein the difference between pre-32 and post-32 are enormous. The other major issues of the set are that the entire set costs way too much endurance and that the set pays for what capability it gets while in Granite Armor with an inordinate amount of secondary functionality costs (mainly reduction in mobility) in all states.

The basis of my survivability argument, operating outside of Granite Armor, is based primarily upon Arcanaville's spreadsheet and secondarily upon experience in game. In Arcanaville's spreadsheet (link here), assuming more than a single target in melee range (I would generally assume an average of 5 or 7 targets as "average"), Stone Armor is weaker than Invuln I13 (the version we currently have) in all but the infinitely sustainable category (which inordinately favors any kind of damage recovery mechanism and isn't particularly useful as a comparative balance tool outside a very small number of situations). Similarly, in most players' experience, non-Granite Stone Armor characters are viewed to be weaker than other armor set options. I doubt you could find many people that would disagree with this general sentiment, though I'm reasonably sure some will crop up.

The issue with Granite Armor should be obvious to anyone that looks at the set: without it the set is too weak so, in order to be effective, you're forced to use the permanent tier 9 at all times and forced to work around the mechanisms that are there to discourage the use of it as such. Rather than being discouragement, the negative effects of Granite Armor are seen as being part-and-parcel with the set as a whole: like toggle IH of old, the high "costs" are simply built around rather than being used in the actual intended manner (turn it on only when you really have to).

The basis of the endurance cost argument is rather simple: compare the unenhanced endurance costs of the various armor sets. Excluding damage toggles, Dark Armor costs 1.48 end/sec (removing Cloak of Fear reduces that to .96), Electric Armor costs .73 end/sec, Energy Aura costs 1.04 end/sec, Fiery Aura costs .52 end/sec, Frozen Armor costs 1.04 end/sec, Invuln costs .73 end/sec, Shield costs .78 end/sec, Super Reflexes costs .78 end/sec, and Willpower costs .83 end/sec. Stone Armor costs 1.25 end/sec. Discounting Dark Armor (which ignores standard design for armor sets pretty obviously), Stone Armor is more expensive than any other set, made even worse by the fact that, unlike Energy Armor and Energy Aura, Stone Armor doesn't have any native endurance recovery mechanism.

For damage auras, Stone Armor is similarly expensive. The standard cost for every other damage aura in the game is .52 end/sec. Blazing Aura, Death Shroud, Icicles, and Lightning Field all cost the standard amount, even ignoring secondary effects. Mud Pots costs .78 end/sec. Nominally, the reason is supposed to be because Mud Pots has the benefit of applying both a slow and an immob (though too low mag to affect anything other than a minion) even though it deals less damage (10% less) than the standard damage for damage auras. Within context, it's paying a lot for what amounts to barely any benefit.

The final consideration, mobility restrictions, can pretty easily be summed up by looking at Rooted: it prevents you from jumping or flying and applies a 90% reduction in movement speed. Thematically, it's appropriate, but, balance wise, it's a lot of work for what amounts to no more than what any other mez protection toggle provides (Integration provides 50% more +regen, not enhanceable, than Rooted while only costing a pittance more in endurance). There are better mechanisms to use to apply the thematic "penalty" without functionally requiring the mez toggle be turned off in order to be reasonably mobile.

My suggestions are essentially four tiered: the improvements to survivability, the improvements to endurance cost, the improvements to mobility, and the tweaks to Granite Armor to remove the overt reliance on the set's tier 9 power while allowing it to maintain an acceptable degree of usefulness (numbers are assumed to be Tanker values, rather than Brute).

The improvements to survivability are pretty simple, and I chose them specifically to avoid incurring the wrath of the Cottage Rule.

The first half is to modify Stone Skin so that it grants 11.25% +res(all) rather than 10% +res(s/l). There's already precedent for a low tier baseline functionality passive power to be this powerful. Consider High Pain Tolerance and True Grit: HPT provides 20% +hp (30% enhanced) and 7.5% +res(all), which amounts to ~25% reduction in incoming damage while TG provides 10% +hp and 15% +res(f/c/e/n/t), which amounts to roughly ~15% reduction in incoming damage.

The second half is to modify Rock Armor and Crystal Armor (the defense toggles) to provide an additional 2% +def(all), essentially 18% +def(s/l)/(e/n) and 2% +def(f/c/e/n)/(s/l/f/c). The precedent for this is remarkably simple: the SR toggles. Focused Fighting, Focused Senses, and Evasion all provide 18.5% +def to their respective type, adjusted for Tanker numbers. Conversely, if the increase to all defense types were viewed to be too complicated or far reaching, an increase to 18.5% to the respective defense types of the power (along with Stone Skin being upped a bit to providing 12.5% +res(all)) would achieve the same level of resistance that I achieved with my numbers.

The endurance cost improvements are a bit more far reaching. Because the set is so dependent on toggles compared to every other defense set in the game, in order to prevent the set from being too expensive, the toggles would need to be cheaper than "standard". I would reduce the endurance cost of Rooted to .21 end/sec, Mud Pots to the "standard" .52 end/sec, and Rock Armor, Brimstone Armor, Crystal Armor, and Minerals to .16 end/sec. Since there isn't a hard and fast formula for the endurance costs of toggles, the only real precedent possible is to compare total endurance costs. The total unenhanced endurance cost excluding the damage aura would be .85 end/sec, slightly more than "standard", but substantially lower than the currently exorbitant price.

For the mobility issue, the fix is rather simple: give Rooted the partial Grounded treatment. Specifically, I would suggest allowing the mez protection and resistance portions of the power to be active at all times while the +regen portion is only active while you are in contact with the ground (or in close proximity to the ground). Doing so would allow the power to be used effectively at all times while maintaining the thematic "penalty" of the power. The applicable precedent is pretty obvious: Unyielding. In its original incarnation, it forced you to stand still while providing no additional benefits beyond mez prot. Its current incarnation has no movement penalty and provides the additional 5% +res(all) that we currently enjoy. This change isn't even as extreme as that.

The changes that I suggest that will likely draw the greatest ire from the populous are the changes to Granite Armor. Some like how the single power renders half of the set completely irrelevant. I consider that to be simply bad design and designed these fixes to address that. On the same note, I think tier 9 powers should be powers that highly augment your survivability for a short period of time. These changes address that point too.

The changes are pretty overarching. In order to address Granite Armor rendering the rest of the set obsolete, I would reduce the +res component to 17.5% +res(all but psi), the +def component to 6.25% +def(all but psi), and remove the mutual exclusivity (so that it can be used on top of the rest of the set and turned off and on without requiring you to take time to turn everything else back on). These changes would allow Granite Armor to maintain roughly the same level of survivability it now enjoys (on par with Invuln + Unstoppable) while preventing it from completely rendering the rest of the set null. On top of that, I would have the power automatically turn off after a set period of time (120 seconds) in the same manner as Phase Shift, while increasing the recharge to 180 seconds with the recharge unaffected by recharge enhancements (like OwtS and SoW). To make up for what most people would probably view as substantive nerfs to the power (even though the survivability remains roughly the same and the only real change is forcing the power to be used with a known downtime), I would remove the -dam and -rech penalties while adding 35% +recov (to grant .5 more end/sec than the cost of the toggle). There would similarly be no crash, though the mobility penalty would still apply (cuz you're still a giant rock). The power would just turn off when you chose to turn it off or when it had been on for 120 seconds, whichever came first.

Overall, the change would do nothing to your survivability while Granite Armor is up, allow you to deal full damage and generate threat while it is up, and force players to not spend their entire lives post-32 with it on (such as was fixed with other perma-tier 9s way back when).

I would tweak AI in a couple of ways.

The first big change would be to provide an overarching group AI that loosely governs the specific actions of individuals enemies within the group. The presence of higher rank enemies (bosses, lts., EBs, AVs), as well as the ratio of enemies to allies, would increase the number of group AI options that there would be, such as encouraging enemies to disperse to minimize the effects of AI or to magnify the aggro effects of debuffs. Without higher rank enemies, group AI would largely govern the likelihood of a spawn to scatter and attempt to join other groups (i.e. transfer their ownership to other spawns).

The other change I'd make would be to make the AI of individual enemies a bit more random. Depending on how far you'd be willing to go with it, I'd either make it so that enemy actions are governed by a random action table (wherein previous actions, effects, and group AI reduce or increase the likelihood of certain actions), or give most enemy categories specific AIs that encourage or discourage specific actions (such as most robots refusing to run away or smarter enemies specifically aiming for allies that are buffing and/or debuffing).