I don't like the idea of resilience. Or, to be more specific, I don't like it on your only choice for gear for that particular progression path.

Right now PvE gear is viable in PvP, and vice versa.

A PvP geared player will survive a bit more damage.
A PvE geared player will do a bit more damage.
In a fight these two things should balance out.

PvE gear is definitely viable in PvP right now. It's arguable whether or not PvP gear is totally viable in PvE. For hunters the epic PvP set is about equal to Striker's Garb damage-wise (slightly better, imo)(and yes, that's accounting for one being 6 piece and the other a 5 piece). The offspec sets are rather good starters for feral druids and shadow priests (mostly due to the lack of other gear). Great gear for DPS warriors. I do occasionally see a rogue wearing 2 pieces, for the 20 stamina bonus I assume. For DPS casters I'm not sure, certainly the 2 piece bonus can improve survivability, and the dagger has very nice spell damage.

PvP gear in TBC is overloaded with stamina (a little bit too much, imo), along with a bit of resilience. Both armor and weapons have resilience. If I want to gear my character up from PvP I'm going to end up with plenty of resilience. I don't care for it much as a PvP stat, I'd rather have stamina that helps me against everything (warlocks/shadow priests included). I really don't care for it as a PvE stat. As a hunter I'd get a much better return on avoiding that boss 4 shotting me with normal hits rather than trying to avoid a crit but getting 3 shot with normal hits. PvE mobs just don't crit that often.

Really, I'd much rather have a decent amount of agility, AP, and int (PvE+PvP stats) than defensive stats, but resilience is definitely a much larger offender than stamina. "But Lozzle, PvP gear isn't supposed to be viable for PvE!" Well, why is PvE gear viable for PvP then? Aren't the two paths supposed to be equally viable? I'm not saying that I should be able to grind out some honor and go slay Illidan, but I'd like to see PvP gear have a use outside of arenas.

The best solution I've thought of would be to balance PvP gear so its defensive stats (stamina) are slightly higher than offensive stats, with 0 resilience on it. If a player wants to add resilience it'll have to be through socketing. If a player wants to add offensive stats to make it slightly more PvE viable (although not 100% as viable as equal level PvE gear) then they should have the option. The strength of the socketing system lies in the choices it gives players, but you really don't have much choice when the base gear is crap for you.

Of course, you could spice it up a bit by throwing in some different items. Throw in some great purely offensive items that every player will want (Don Julio's band, anyone? I don't think I could replace that until late AQ40) Throw in a few defensive options (perhaps with a slightly more defensive model design of the default arena gear), maybe defensive shoulders for every class, along with a few random defensive slots.

In closing, uh, I have other stuff to do so I'm going to make this short. The only really exciting honor/arena rewards that I've seen are the crossbow and, to a lesser extent, the axe. The crossbow is slow, good for PvE and PvP. It does have a few resilience on it, but oh well, they can't really take the offensive edge off of a weapon. The axe is pretty good for PvE and PvP, too. It has a little bit of resilience (ick), but I really dig the decent amount of stamina, and the spread out hit/crit/AP taking advantage of the item budget.

P.S. I really hate hunter/shaman items that have 0 agility and a bunch of mp/5 on them, too.

Getting crit for less is more noticeable, but going by the numbers wouldn't you have diminishing returns?

Say you're fighting a rogue with 20% chance to crit you that hits for 100. Going with very rough numbers here, not hit figured in, etc blah, IIRC talented Rogue crit damage bonus is 30%, so that rogue with 20% crit would have 80+(20*2.3)=126% of base white damage. If you have X resilience that gives -5% crit, -10% crit dmg, then that rogue is now hitting you for 85+(15*(2.3*.9))=116.05 dmg. If you have twice as much resilience, call it Y, for -10% crit, -20% crit dmg, then that rogue is now hitting you for 90+(10*(2.3*.8))=108.4 dmg.

So going from "unreduced" to X you avoid 9.95 damage.
Going from "unreduced" to Y you avoid 17.6 damage.

Since you doubled the amount of resilience you have, you should have double the amount of damage reduced, right? But as you reduce your chance to be crit, you also reduce the chance that your reduced damage taken from crit part will apply. Saying that sentence would probably have been easier than doing all that math, but I needed to prove it to myself as well as give you all a chance to tell me I'm horrible at math.

I'm not sure if 10% crit damage reduction would figured as 2.3-(2.3*.1) or 2.3-.1, I don't know if it would affect the results either, I don't think so but I'm going to bed either way.

P.S. I just realized I spent the better part of an hour typing out 950 words (surely more by the time I finish this postscript). I'm not sure if I should be sad that I wasted all that time, or happy that I improved my skills as a writer while enriching this great community. On that same note, shout out to the mods, you guys are awesome. Ok, I'm really going to sleep.

They're supposed to be equally viable, and that's why they're being separated so much. PvE gear isn't going to be as good as PvP gear for PvP and vice versa. I imagine some people will go a more glass-cannon route and show up in their PvE gear, and while it might work for some classes, for most it means they're immedietly going to become the first target.

The tier 5 PvE sets generally have less than half the stamina as the first tier of gladiator sets. Presumably the current gladiators sets are balanced vs. tier 4 and they might update by the time we're wearing tier 5 too. There's definitely some pretty clear separation going on.

There is more to PvE than simply defense versus resilience.

What about Defense versus Resilience+Defense?


The following are equal in terms of stat budget:

A. 120 Defense Rating (2% to miss, 2% fewer melee crit, 2% dodge, 2% parry, 2% block) [8% mitigation total + the block]

B. 120 Resilience (3% fewer crits, 6% less damage from a crit) [3% mitigation total + tiny bit if not crit immune]

C. 40 Resilience + 104 Defense Rating (~1.73% to miss, 2.73% fewer melee crit, 1.73% dodge, 1.73% parry, 1.73% block, 2% less damage when crit) [7.93% mitigation + block]

D. 20 Resilience + 114 Defense Rating (1.9% to miss, 2.4% fewer melee crit, 1.9% dodge, 1.9% parry, 1.9% block, 1% less damage when crit) [8.1% total mitigation, with some block]

A 5:1 ratio of defense to resilience beats out pure defense for a class that can parry. An item with such a ratio is superior than one with pure defense. In addition, such an item gets you crit immune more rapidly, at the expense of miss, parry, and dodge. As one approaches the threshold of crit immunity, resilience becomes worthless and dodge / parry / agility become more valuable than defense.

From an earlier post:
----------
25 defense = 60 defense rating

1% parry = 31.5 parry rating
1% dodge = 18.9 dodge rating
1% block = 7.9 block rating
1% resilience = 39.4 resilience rating
-----------
Since dodge is the most valuable of the above per rating point, an item with about 2:1 defense to dodge ratio is extremely efficient. Resilience in small amounts on the same item with defense is as well, but dodge much more so since it is cheap.



using round numbers:
1% mitigation comes from:
15 defense rating (ignoring block)
40 resilience (ignoring reduced size of crits)
20 dodge rating

Mitigation per item budget
Item Budget, C = ( (D)^3/2 + (R)^3/2 + (G)^3/2 ) ^ 2/3
C is a constant.
Mitigation, M = D/15 + R/40 + G/20
M varies for a constant C as D, R, and G satisfy the first equation.
D = defense rating, R = Resilience, G = Dodge rating
Where the second derivative of this equation = 0 is the optimal balance of stats for max mitigation, assuming one is not crit immune and a class that can parry. Adjustments could be made for druids and when crit immune. I went part way through deriving the 'optimal' balance of two stats A and B given weights X and Y where total value is = X*A + Y*B and the itemization formula is constant C = (A^3/2 + B^3/2)^2/3. But I'm sure someone has done this before....

Anyhow, items with small amounts of a less valuable stat combined with a valuable one are in total better than the "good" stat by itself.

This is true of AP and STR for a cat druid, where STR is better per point but a mix of the two best, or Resilience + Defense for tanks. There are countless pairs and triplets of such stats in the game, and the "worse" stat is not bad, if in small quantities on the same item as the better stat. Blizzard doesn't typically itemize with two of such synergistic stats very often.

For example, every +heal item in the game could have a mix of +heal and +spell damage that would total to the same ammount of +heal, but would also give some spell damage. But Blizzard doesn't do this.
Additionally, it is possible to slightly increase the total +heal on an item by sacrificing a bit of +heal for more +spell damage.



-------------
One other thing: there is a some talk about gems and sockets in here. I noticed that two gems of the same Ilevel were not following the multi-stat power law rule.
The PvP reward ones are something like:
gem A: 5 stam, 10 AP
gem B: 20 AP
gem C: 10 stam

The above aren't the real values, but the point is that item budgets for gems seem to simply follow a linear equation.

Gem item budget = (Value of stat 1) + (value of stat 2)

no 3/2 or 2/3 powers.

I would assume that every gem slot counts as either a fixed item budget cost, or a fixed percentage of the total Ilvl of the item. Likewise for meta-gem slots.
Since anyone can mix/match gems how they wish, you can't use the power law effectively to itemize them. They are linear.

Actually, a socket seems to cost about 8.7 of a pure stat. More than a blue gem (+8 to stat), less than an epic (+10 to stat).

Its chief advantage lies in the fact that, if you add a gem with an already existing stat, the added points thru gem cost less in the budget than if they had been added directly. However, if you add a stat that didn't exist on the item, you lose out in potential.

Assuming you value the stats like Blizzard does (1INT=1SPI=1STR? who are you kidding?)

Another interesting facet to this argument, at least from a priest standpoint (and to be honest, a priest is a class I see this resilience thing applying to quite a bit since they usually are PVP tanking) is the Blessed Resilience talent, which has a 60% chance of being activated on being crit. Reducing your chance to be crit through stacking resilience will marginalize talents such as these. That marginalization most likely isn't to a massive extent, but it is still there. You can either rely on Blessed Resilience to do the work for you and stack other things, or move the points out of that talent and put them elsewhere. The issue then becomes having enough stamina, and if you still have to wear the PVP gear to get a respectable amount of stamina, then you don't have a choice of having the resilience.

I'm going to go ahead and ask why people are busy calculating the value of resilience for warriors who are getting heals. If a warrior is getting heals, he really shouldn't be taking damage anyway, his healers should be taking damage. Hence they're the ones who are trading off resilience for other stats. How much +heal would a priest trade off for resilience? Saying that resilience is more useful than stam because it gives you greater benefit when you're getting heals is imho rather irrelevant since if you're getting healed, your healer is the one who's going to be taking the next 12k of damage, not you.

One thing I'll remind people is that to date, Stamina's value varies depending on who coded the item, and when they coded it. Though I'm sure it will be standardized before release, there are still some items with Sta being worth 50%, and others with it worth 66.67% of other stats.

Not necessarily. Anyone in the arena is at risk of dying to burst damage, not just healers. A team who loses all its DPS is in as much trouble as a team who loses all of its healers - possibly more.

Absolutely right.

After watching the new wound poison in action, people will quickly realize that "focus their healers first" is not the mandatory step it seems to be, especially when you add MS to the equation.

Even in Live, once the damage started ramping up to insane levels people started realizing that if you focus that warrior down first, the priest and paladin behind him with no dps and +1000 to healing aren't going to be able to kill you. In Live 2.0, this is even more prominent with the huge bias towards damage classes.

Speaking of which, does the new Wound Poison stack with the MS debuff? If this were the case, assist trains will be unstoppable with a Rogue/Warrior. A well organized assist can only be dispersed by a massive disturbance or heal through.

They stack to 75%, which seems pretty ridiculous to me. It becomes next to pointless to heal them at that point.

I think the only class who can cure it somewhat effectively is a druid since abolish poison ticks pretty fast. Poison cleansing is too slow, and cleanse isn't practical.

Because in any circumstance other than receiving significant heals, resilience is terrible. I was doing my best to favour it. At this point the math is simply reduced to, if you don't expect to take >20k damage in a pvp fight, you're better off going with stamina.

Well the argument between stam and resil is pretty much moot because they come hand in hand. The only general situation where you get to choose one or the other is in gem selection. I think the bigger debate that will be coming up is how much dmg or int (aka sustainability for casters) you should sacrifice for that resil.

Abolish won't do much good against a rogue running around with Vile on, either, since it's a 40% chance to resist dispel, and wound poison needs to be dispelled one layer at a time.

Dwarves! Have to really say stoneskin grows in power the more the game goes on it seems.

If two priests are spam healing, you CS/Spell Lock/Silence/Silencing Shot them, then DPS the Warrior down because he's not wearing any stamina/resilience gear. See how PvP theorycraft works? ;)

To add to the combined stats building discussion, I have an item on test with "increases hit avoidance rating by 14". I don't see anything on WoW Wiki about it. I need to level once more on Blade's Edge before I can equip it to test it out, although it probably increases the enemy's chance to miss.

If it does increase chance to miss, then building an item with resilience, dodge, and hit avoidance would be an interesting experiment for itemization, although it seems unlikely that it will provide significant itemization savings over defense, as was previously calculated.