This topic is realy fun to read, and the idea behind tanking as a rogue is realy sweet
My guild is currently end T5 (only kael missing) and I've been running some numbers to see if perfect physical avoidance is reachable for me - initial numbers put me close at 98.2%, which got me looking up more items as perfect avoidance is a must (any kind of hit can kill you with rogue armor :/ ; 99% avoidance is just not enough).
I started to use the scales posted earlier in this thread, then moved to a set of my own (which add up to much nicer numbers)
basically I'm using the following scale:
Agility: 6325
Defense Rating: 5073
Dodge Rating: 5284
Parry Rating: 4227
with these numbers the scores directly result in avoidance * 1000 (so just divide by 1000 and you get the avoidance of the item) Loot Rank page with these scales
I still have to realy fiddle with the whole bit to see if I can reach the cap at my guilds progress level (and without using any epic crafted gems), but I thought I'd post the above here to get some confirmation of my math
Last edited by zeth_russ : 12/21/07 at 6:22 AM.
Reason: fixed the math
If you get it together, it might be worth testing it out vs. Attumen (he doesnt do any magical damage from what I remember), and seeing how far you can get.
I was going to say that Nalorakk would be another good target, but I forgot that he does that annoying shout, which will unfortunately cause you to die fairly quickly - bring one healer though, and it should be doable. healer stands behind the boss and heals you and he through the shout, you tank Nalorakk (through both phases), laughing all the way home.
Edit: In fact, there are probably quite a few bosses, that whilst not 'soloable', will be duoable by avoidance-rogue + healer, as long as the healer can heal through the various periodic magic damage. Malchezaar would fall into this category too, if the healer can regen enough to always heal (and/or dispel) the shadow word pains and shadow novas, and you can keep up concentration to dodge infernals for long enough, he'd be doable.
Getting to the point where you can actually solo/duo stuff is actually a fair amount harder than the calculation we've been doing; if you look at most of the gear sets posted so far, they tend to factor in Scorpid Sting, Insect Swarm, Mark of the Wild, Grace of Air and Blessing of Kings, meaning you need a hunter, druid and shaman with you for the actual attempt, and a Paladin at the very least to prebuff... but if the fight goes over half an hour (which it easily could with an undersized group) you'd need the Paladin inside with you as well.
In order to actually solo stuff, you'd need to hit the numbers in question without those buffs; but those buffs are giving 7% miss chance and a couple hundred agility, meaning that you're gaining something like 15-20% avoidance from them all told - which is just way too much to make up in gear.
Still, even the ability to 5 man some of these bosses with Ret Pally/Enh Shaman/Hunter/Balance Druid/Rogue Tank would be pretty funny.
Still, even the ability to 5 man some of these bosses with Ret Pally/Enh Shaman/Hunter/Balance Druid/Rogue Tank would be pretty funny.
The funnier part is that 3 of those 4 supporting classes are threat machines with no innate ability to deal with overaggro.
If such a thing could be done, I'd lobby my guild for it, just so we can have a ret pally and boomkin in our raids.
Considering Aldriana's setup above, you could replace [Drake Fang Talisman] with [Ancient Aqir Artifact] which would bring you that much closer to 102.4% without needing a much desired trinket (one of our tanks has it but opts for Moroes' watch instead). There is of course the loss of dps stats to be considered.
I subtracted out the avoidance penalties from a level 73 attacker when coming up with the numbers for that gearset; hence, the total number it needs to reach is 100%, not 102.4% so it's already at perfect avoidance. Just in case that wasn't clear.
I've been getting into this thread more and more as it progresses, and I think the perfect fight to attempt a rogue tank on would be Kael'thas. Most guilds will have 1-2 advisors still up when he comes, so it gives the rogue a decent headstart on agro, he doesn't really require positioning outside of avoiding a flamestrike or two.
The perfect part of the fight however is the 1 minute pyroblast cycle.
Pyro1: CloS
Pyro2: Cheat Death
Pyro3: Kick
If you don't take physical damage and have a few interrupters you can trust to handle fireballs, you can run a phase 3->4 transition without having to take any healers off of the raid.
With 4 priests for a Fear Ward rotation, Archimonde would be fun too, though it wouldn't have quite the same synergy with your skills/talents.
I calculated the same basic scale as you, except I've got Agi: 6.325 or .06325% Avoidance.
1 Agi = .05% Dodge, but each point of Agi gives 1.265 from Kings and Sinister Calling (1*1.1*1.15).
1.265 Agi = .05*1.265 = .06325.
yeah, and thats where my problem lies, I used 1+0.1 +0.15 for all my agi calculations instead of the multiplicative one.
Will redo my math, thanks for catching that
Originally Posted by TownFool
I also then took that scale and setting Agi as the most effective stat we can see the other's % effectiveness in comparision:
Which can be turned around to see how much of the other stats you would need for them to be equal to 1 point of agility:
Agi: 1
Dodge: 1.216
Defense: 1.246
Parry: 1.496
yeah, I see you point there, I like to have the weights like I have them tho since it allows me to directly compare items by avoidance which makes it easier to find those last 0.5% avoidance that one needs
I agree with Zeth_Russ there, the weighting of the stats based on 1% avoidence is definitly superior to any other weighting. It's funny, but after being directed to Loot Rank from this thread i came up with exactly the same Idea resulting in this weighting: Loot Rank
edit: And one of the biggest problems is... 2 daggers in the "Best of"-list, which will hurt threat-generation a lot.
My best-of list -> Free file hosting by Savefile.com
edit: And one of the biggest problems is... 2 daggers in the "Best of"-list, which will hurt threat-generation a lot.
that only happens if you plan to socket with [Delicate Crimson Spinel]. if you go for [Delicate Living Ruby] then [The Sun Eater] (which is a sword) becomes better again. As I cant afford to buy the expensive gems off ah I'll go for the cheaper way
Hmm, turns out I'm spending far too much time thinking and fantasizing about this whole topic :P
I do believe that It is very important if mainhand (and offhand enchants) could be switched for mongoose - besides the increased threat the additional agility of a single proc gives you another 7.6 % -unpredictable- dodge. while you cant count on it It can save your hide if somebody lets IS or Scorpid drop. For the same reason Ghostly Strike and evasion (together with prep ?) provide an additional safety margin - if one notices the debuffs falling off (or you see cheat death kicking in) you can shout on vt and click ghostly to get into the safe region again.
I do believe that the whole above would be made much easier by an addon that keeps track of your personal avoidance + enemy debuffs and warns you if you have to hit evasion or gs. anybody want to code one ? :P
So, I put together a reasonable option in terms of a plausible gear set for making this happen as a rogue, by a couple of criteria:
1) Hit 100% avoidance
2) Don't use any high-end items that you wouldn't reasonably have anyway (i.e., no Shadowmoon Insignia)
3) Generate as much aggro as possible.
Everything on this list is reasonable rogue DPS gear, reasonably socketed for DPS, with the exception of 2 BoE items (Necklace of the Deep and Delicate Eternium Ring) and one Karazhan drop (Moroes' Lucky Pocket Watch). Some of the enchants are screwy, but that just costs money; you don't actually have to request anything totally unreasonable for a rogue to have from your guild (although they might balk at the number of Crimson Spinels).
Now, if I'm adding this up correctly:
730 agility, 38 defense rating, and 98 dodge rating from gear; Adding an NE rogue's base 163 agility and 181 agility from buffs (see post 50 for details of what all these buffs are), plus 15% for Sinister Calling and 10% for Blessing of Kings, gives a total of 1358 agility. 38 defense rating works out to +16 defense, and 98 dodge rating is 5.18% dodge. Hence our total dodge rate is 5% (talents) + 5.18% (dodge rating) + .64% (defense) + 1% (NE racial) + (1358-11.5)/20 = 79.14% total dodge, minus .6% from being 3 levels below, for a total effective dodge of 78.54%. Parry is 5% base plus 5% from talents, +.64% from defense, minus .6% for being 3 levels below, for a total of 10.04%. And enemy miss chance is 5% + 7% (debuffs) + .64% (defense) - .6%, for a total of 12.04%. Hence, your total avoidance is
78.54% + 12.04% + 10.04% = 100.62%
against a level 73.
Note that actually getting a Shadowmoon Insignia in place of DFT allows you to do useful things like put Exec/Mongoose on your weapons which will significantly increase your DPS (and thus aggro); however, even without it you have reasonable DPS stats; by my count your unbuffed stats are 2168 AP, 31.38% crit, 105 hit rating, 25 expertise rating, and 350 armor penetration. Oddly balanced? Certainly. Weaker than a full T6 dps-geared rogue? For sure. But with a 0/20/40+1 spec, this could plausibly generate enough aggro to not drive your DPS nuts.
Hmm, I have most of the gear listed (2nd chest/pants/boots that I'm using for battlegrounds currently). I'd have to pick up the next Insignia (which isn't a problem probably since all our tanks have one.) and a tanking cloak (since I only have Drape and Dory's) ranged item is probably the trickiest since no one is doing Doomwalker, Zerak's or Alar gun are viable alternatives I guess, if one uses a pure tanking cloak perhaps?
I also have a Belt of Deep Shadows with 2 delicate's, that would be dodge wise a better option I guess?
Fans glory to the Gladiators,
Gods glory to the Heroes.
Blade of Infamy MH is probably a worthwhile tradeoff over Umbral Shiv for threat/damage reasons, while only losing 4 agi.
Melador> Incidentally, these last few pages are why people hate lawyers.
Viator> I really don't want to go all Kalman here.
Bury> Just imagine what the world would be like if you used your powers for good.
I've been getting into this thread more and more as it progresses, and I think the perfect fight to attempt a rogue tank on would be Kael'thas. Most guilds will have 1-2 advisors still up when he comes, so it gives the rogue a decent headstart on agro, he doesn't really require positioning outside of avoiding a flamestrike or two.
The perfect part of the fight however is the 1 minute pyroblast cycle.
Pyro1: CloS
Pyro2: Cheat Death
Pyro3: Kick
If you don't take physical damage and have a few interrupters you can trust to handle fireballs, you can run a phase 3->4 transition without having to take any healers off of the raid.
With 4 priests for a Fear Ward rotation, Archimonde would be fun too, though it wouldn't have quite the same synergy with your skills/talents.
I think you chose 2 of the more difficult bosses to try rogue tanking
Kael'Thas with his MC, Flamestrike and Phoenix spawns has more than enough unpredictable components that kicking all his fireballs is not a trivial task. 2 of them getting through in quick succession would almost certainly kill the rogue tank.
And dont forget the clouds after an antigravitational phase.
And on Archimonde I certainly would not bet that all the right raidmembers are always in range to debuff (scorpid, insect swarm) and buff (grace of air, fearward) in time.
There are a lot more convenient bosses where you could try rogue tanking in BT/Hyjal.
e.g. Gorefiend, Najentus in BT.
Does the scream in phase3 of RoS convert energy into damage?
Does the scream in phase3 of RoS convert energy into damage?
I'm nearly positive it only does mana and rage. I've never actually seen a rogue stand in front of EoA to test it, but according to the Wowhead spell description it only does mana and rage.
I'm not 100 % sure how the mechanics work but surely normal probability theory should apply?
That would mean 20 % dodge gives you a 80 % chance to be hit and if you also have 10 % parry ontop of that you shall have a 72% chance to be hit not 70 % as has been suggested since you must multiply and not add. (note also that since it multiples the order is irrelevent)
HIT 100%
/ \
Dodge 20% HIT 80 %
/ \
Parry 8% % HIT 72% NOTE 10% parry = 10% of the hits that failed to dodge
etc...
This would mean that you would never be able to have 100 % invulnerability unless any SINGLE stat gives you 100 % chance.
Except that attacks generally use a one-roll system, not a two-roll system; that is, 80% dodge, 10% parry, 10% miss is converted into the attack table:
0-80 dodge
81-90 parry
91-100 miss
Hence, they do, in fact, strictly add; the game does not check once for dodge, once for parry, once for miss, etc.
I'm not 100 % sure how the mechanics work but surely normal probability theory should apply?
That would mean 20 % dodge gives you a 80 % chance to be hit and if you also have 10 % parry ontop of that you shall have a 72% chance to be hit not 70 % as has been suggested since you must multiply and not add. (note also that since it multiples the order is irrelevent)
HIT 100%
/ \
Dodge 20% HIT 80 %
/ \
Parry 8% % HIT 72% NOTE 10% parry = 10% of the hits that failed to dodge
etc...
This would mean that you would never be able to have 100 % invulnerability unless any SINGLE stat gives you 100 % chance.
This isn't how the melee attack table works - it works off of a 1 roll system where 10% parry = 10% of all attacks are parried, even if you have 20% dodge, 35% to block, etc. More info here: Attack table - WoWWiki - Your guide to the World of Warcraft
Would this mean that +hit from gear and Shadow Focus effects (Increases chance to hit with shadow spells) would counter Unbreakable Will and other increases chance to resist fear stun and silence ? After gear has worked off inherent resistance?
Example
what is X, Y & Z?
Priest 1 with unbreakable will
has X % chance to resist Priest 2 Silence
has Y % chance to resist Priest 3 Silence
has Z % chance to resist Priest 4 Silence
Priest 2 with 10% hit from gear no Focus
Priest 3 with 5% hit from gear Focus (10%)
Priest 4 with 0% hit from gear Focus (10 %)
I assume you refer to PvP, which almost certainly is two-roll system (for example, you can block a crit). Furthermore, spell resist mechanics are a two-roll system ANYWAY.
Regardless. Look at "Working Theories of Theorycraft" stickied at the top.
Edit: I'll indulge you anyway. You have a 4% chance to "miss" a level 70 target with a spell base, 1% chance to "miss" regardless of +spell hit.
First roll: 0-1 miss. 1-(95+spellhit%): hit. Else: miss.
Additional Roll #1: Chance to resist based on school-of-magic.
Additional Roll #2: If successful so far, roll to check "resist X mechanic"
Edit 2: Post #409 of "Surviving as a Priest" indicates that the "resist stun, etc" mechanics add to the base miss chance, and can thus be negated by Shadow Focus. I haven't seen evidence either way, so perhaps I should apologize for presenting the above as fact.
I propose a 3 stage test:
1. Test total avoidance is actually (not just theoretically) achievable.
2. Test calculated avoidance is actually correctly calculated.
3. Reproduce test 1 + 2 vs. level 73 mob.
Test 1
Since it is difficult to reach total avoidance vs. level 73 mobs, it might be a good idea to test against lower level mobs to see if total avoidance is actually (not just theoretically) possible. The idea is to verify that it is indeed possible to negate a mobs ability to land a hit.
The reason why this test is interesting is to stop the speculation about why rogue tanks who should have evasion tanking in the 110%-130% range sometimes get oneshot, and to establish a baseline for more demanding testing.
The test could be easily done against mobs of much lower level to reduce gear requirements. A sample of a few thousand hits should be sufficient to give a strong indication of whether achieving total avoidance is possible.
Test 2
This thread has a lot of gear combinations posted, with calculations to work out what avoidance percentage it should all amount to. It is possible to test if a calculation is correct by testing against a mob that has exactly the highest level for the given avoidance percentage, verifying that the mob is unable to land any hits, and then repeat the test against a mob precisely one level higher, and veryfying that this mob can indeed land hits (probably about 1 in 100).
Note that both test 1 and 2 can be done by any class with virtually any gear. Simply calculate your theoretical avoidance percentage and chose appropriate level mobs to test against.
Test 3
Reproduce test 1 and 2 against a boss mob with a combination of gear and abilities that gives total avoidance against level 73.
got some issues with your proposed tests - the problems being that I do not believe the same game mechanics apply when talking about a) lower level player vs higher level mob versus b) lower level mob vs higher level player.
now, this is only a gut feeling from having played this game so long, but I cant remember mobs in deadmines not hitting my rogue at lvl 70 (obviously, I cant guarantee here that all of em where always in front of me, so take it with a grain of salt).
a simple and easy thing to check the viabilities of the test (normal geared players vs low level mobs) might simply be to grab a low level mob and let it hit your char from the front for a while.
Another anectotal thing here: my lvl64 warrior was once gibbed by a level 1 frog in Zangarmarsh.... (hitting for 1s between parry and dodge).
Might add a bit more to this once I get back from an appointment
Originally Posted by Kyri
I propose a 3 stage test:
1. Test total avoidance is actually (not just theoretically) achievable.
2. Test calculated avoidance is actually correctly calculated.
3. Reproduce test 1 + 2 vs. level 73 mob.
Test 1
Since it is difficult to reach total avoidance vs. level 73 mobs, it might be a good idea to test against lower level mobs to see if total avoidance is actually (not just theoretically) possible. The idea is to verify that it is indeed possible to negate a mobs ability to land a hit.
The reason why this test is interesting is to stop the speculation about why rogue tanks who should have evasion tanking in the 110%-130% range sometimes get oneshot, and to establish a baseline for more demanding testing.
The test could be easily done against mobs of much lower level to reduce gear requirements. A sample of a few thousand hits should be sufficient to give a strong indication of whether achieving total avoidance is possible.
Test 2
This thread has a lot of gear combinations posted, with calculations to work out what avoidance percentage it should all amount to. It is possible to test if a calculation is correct by testing against a mob that has exactly the highest level for the given avoidance percentage, verifying that the mob is unable to land any hits, and then repeat the test against a mob precisely one level higher, and veryfying that this mob can indeed land hits (probably about 1 in 100).
Note that both test 1 and 2 can be done by any class with virtually any gear. Simply calculate your theoretical avoidance percentage and chose appropriate level mobs to test against.
Test 3
Reproduce test 1 and 2 against a boss mob with a combination of gear and abilities that gives total avoidance against level 73.
Oh, as a small p.s.: Under the assumption that normal mechanics apply for low level mob vs high level player, what are the exact formulas we're talking about ?
Oh, as a small p.s.: Under the assumption that normal mechanics apply for low level mob vs high level player, what are the exact formulas we're talking about ?
For each level difference add 0.2% to dodge/parry/miss to your Character Sheet dodge/parry rates.
Assume 5% miss rate for non dual wielding standard mobs of same level.
I was bored so I ran a first test to show total avoidance is possible (not that I had doubts on that).
My Character Sheet (i.d vs Lvl 70 mobs) dodge rate was 37.12%.
Parry Rate was 23.14%.
The miss rate was increased by 9% due to my 575 defense.
I let an Lvl 20 elite mob in deadmines attack me for some time.
My effective avoidance rate against this mob was expected to be
47.12% dodge, 33.14% parry and 24% mob miss (assuming 5% standard miss rate here), due to 50 lvl difference enhancing avoidance by 10% in all categories.
Note that the one hit by the mob was on purpose since my recap tool does not show a statistical breakdown (miss/dodge/parry) if I do not take damage at all.
Interesting observation:
The observed dodge and miss rate almost match the theoretical ones.
The parry rate was obviously truncated.
And this I find somewhat odd, because parry for me is more valuable than (mob)miss and dodge.
If I had to choose, I'd go for the following preferance: parry>miss>dodge
With these stats I should be untouchable to lvl 27 mobs, and hittable by lvl 28 mobs.
Testing this right now.
EDIT: I realize, my german UI might be of no real use to some so here come some quick explanations.
I was hit one Time (on purpose, by turning my back to the mob). All other 2866 attacks were avoided.
46.5% of these were dodged, 24.2% missed and 29.2% were parried.
12/20/07, 10:11 AM
Similar Threads
