◊ Regen

There is a difference between requiring and "making things easier".

Every tank class is still going to have advantages to make certain jobs easier or else you are looking at a pretty damn boring scenario.

edit: As it stands on live the only job I would argue that more or less requires a paladin are the shadow spawns that come out of the void sentinels on M'uru (does't even require you to be prot? so the point is still pretty arguable) which pretty much changes with TC / Swipe going infinite in 3.0............ Yeah. And obviously shear but that is being removed (but still had work arounds did it not (intervene?)).

 

Tacocat

The problem with Thorns is that it is negatively affected as we get more avoidance. The higher dodge you have the less Thorns will hit.

 

• sadris

If one tank is capable of holding aggro through Seed of Corruption spam and others are not, then that tank is required. The original math which I posted that did not include Retribution Aura yielded the conclusion that a Druid tank cannot hold aggro through SoC spam.

 

◊ Regen

In regards to what situation? How soon are seeds started?

edit: Just read over

#1. You're assuming warlocks start seeding immediatedly. Almost never the case
#2. You're also assuming that warlocks immediatedly pull aggro as soon as their threat exceeds the tank

Assume 4gcds, and the druid has a lead of 2800 threat. How long does it take the warlock to catch up? 56 seconds to reach equal threat(3:42, 2:28, 1:14) which is still not riping.

May swipe need a buff? Maybe, but I think your blowing it up to be more than it is, assuming your AP/crit (seems undervalued) and spell damage numbers are all in line at the same tier of gear.

 

Promethius

Are we sure we're not underestimating swipe here? Granted there's a valid argument about needing to use all available global cooldowns to maintain aoe and positioning requirements, but the damage and threat per swipe is quite high if the numbers from http://elitistjerks.com/918674-post2556.html are an accurate representation of what a raid buffed druid will look like after 1st tier 10 man raiding.

Bear (crit% vs. a boss (-4.4% added)): 37.2895%
Bear (AP): 6636
Boss Damage Reduction (after debuff and penetration): 26.1969%
Missed % (Total): 7.41415%

With those stats, swipe spam against targets 3 levels higher yields these numbers:
(base + ap/14)*FI*(1-armor reduced)*(landed hits)*(hit+PI meta gem crit dmg)
(108 + 6636/14)*1.3*(1-.261969)*(1-.0741415)*(1+0.372895*1.26) = 759.9 avg dmg per swipe
759.9 dmg/swipe * 1 swipe/1.5 sec = 506.6 DPS
506.6 DPS * (1.45/0.7) threat/dmg = 1049.4 TPS

I don't know about the TPS of warlock seed of corruption or paladin consecrate at a similar gear level with buffs, but also keep in mind that due to the fact that swipe is physical opposed to magical, that reducing mob level nets more gains to swipe due to lower armor, avoidance, and crit reduction. In addition, when arguing about the threat swipe generates versus consecrate, it's very important to make the distinction between threat per second and threat per global cooldown. Looking at a paladin with 3k ap, 3k stam and say 300 spell dmg from raid buffs and no resists, they'd have a consecrate of 2252*1.13 spell vulnerability = 2544.76 dmg over 8 sec. That's 318.1 DPS or 865.2 TPS. Maybe I'm underestimating the stats of what a paladin would have though. Regardless, a paladin would need close to an additional 1500 combination of ap and stam to bring the TPS of consecrate to the above swipe numbers, and the DPS is still lower. When considering the difference in TPS and how it relates to threat per gcd, a druid would probably get 1 gcd + mauls v. 4 gcds + white attacks/seals against the single ff target while maintaining similar aoe TPS. How this compares I don't know, but if the aoe threat is more to keep them off healers opposed to aoe, a druid does have the option to lower aoe threat to increase single target threat.

Consecrate itself has limitations in that if for any reason the paladin had to move, the threat is lost from any mobs moving out of the affected area. The increased benefit of thorns and retribution aura for the paladin are good points, but looking at how powerful swipe may actually be, if swipe spam and consecrate begin at the same time, the druid will overtake threat and have the mobs on him anyway, denying the reflected threat for the paladin. I think a main distinction of druid aoe from the other tanks is that in terms of snap aggro, it is most forgiving because of no cooldown and only loses to shockwave and perhaps thunderclap for burst threat. Paladins will remain the supreme aoe tanks when mob control has been established and the paladin no longer has to move. Whether it was intended or not, this gives paladins the niche of best handling mobs that come at the group from all directions, gives druids the niche of snapping threat on gauntlet type encounters where quick threat has to be established on new mobs in front of the constantly moving group, and gives warriors and death knights a balance in between the two.

As for the thorns spell damage scaling, I'd lobby for NI becoming a 70% agi to spellpower and 20% stam to int or something, dropping the fel armor (demon armor now?) effect to cat, to fill in that pvp gap and create actual synergy between HotW, iLotP and our itemized gear.

Edit: Forgot Naturalist and potentially Master Shapeshifter. DPS goes to 557.26 and 579.55 and TPS would rise to 1154.34 and 1200.51 with the respective talents. 1 mangle ever 4.5 sec means you could still maintain 800 aoe TPS or so in the latter case.

 

teiglin

I was comparing our best aoe threat to other tanks' best aoe threat. Of course you can mangle on every cd, but you're reducing your aoe threat by a third (a quarter w/o imp mangle), with the result that you're still using fewer GCDs on single-target threat than a warrior and causing significantly less aoe threat than you otherwise would be causing. As has been mentioned by others, there's a difference between having to work harder--such as because of positioning requirements, which are not that big a deal--and being unable to compete--worse scaling, zero free GCDs.

All people are saying is that druids should be able to produce comparable aoe threat as other tanks. Sadris has clearly demonstrated that our threat is significantly worse, to the point that other tanks can produce as much or more threat than aoe'ers while we cannot. This is a problem with the class.

That's a good point, but the same is true of retribution aura and damage shield.

 

◊ Regen

Pretty much what Promethius said. Numbers could be off / blowing things way out of proportion in regards to not having suffiecient aoe threat.

Paladins are going to have better aoe threat assuming he remains static.

Working harder was in regards to holding aoe threat + much higher single target threat for the main assist (for instance if you choose to aoe M'uru adds all in one spot, so seeds are wearing the entire group down + M'uru while melee are simultaneously single targeting mobs down ---- this is also the most dps and time effective way to kill M'uru phase 1)

 

• sadris

Good analysis, not really sure you can get 6600 AP in bear form with tanking gear, though.

Also, the AP scalar for swipe is 0.057, not 0.07 (1/14), as done in your model. This would bring the final number down to 880 tps.

 

kalbear

Perhaps someone can help me with the modeling here. I keep coming up with the notion that it's better to get to a faster Savage Roar so that all other attacks can be under its influence. Which means going for faster combo points, which appears to be more mangles.

This is based on the assumption that the best cycle will be to keep 5SR up at all times and then do 2 finishing moves in the time SR is active. This allows at least one tiger's fury up during the 34 seconds of SR activity, which means a total of 400 energy is available. Of that 400 energy 70 will be used for finishers, leaving 330 to use on abilities.

With shred, this means you will get just under 8 shreds in (330/42), which is fine; at least two of the attacks must be mangles to keep the debuff up. So you're looking at 6 shreds, 2 mangles, and two finishers (and 10 extra energy)With a crit rate of 40%, this gives 11.2 combo points on average. That's not enough to do two finishers and a 5-point savage roar in that time. Even inserting a couple of rakes does not allow for this, as there is not enough energy saved.

With a mangle cycle, however, we can do 9.5 attacks given the energy we have. Let's assume 9. This gives us an average of 12.6 combo points in a given cycle, which still isn't quite enough to do 2 finishers + a 5 point SR all the time - but it is very, very close to it. With a 50% crit rate we're looking at 13.5 combo points in the same cycle, which should give us a 5 point SR either at the end of SR or after the first attack immediately after.

Now, let's assume one free attack from OoC during that 34 seconds. With that, the shred cycle gets a 9th attack and gets an average of 12.6 combo points at 40% crit rate. The mangle cycle gets a 10th attack and gets 14 combo points on average. This means it is almost 100% certain to get a new SR either immediately following its end or after one more attack.

What kind of damage loss are we talking about here? For the shred cycle, let's go with 2 rakes, 4 shreds and two mangles. For the mangle cycle we'll go with 7 mangles and 2 rakes. Assume any OoC is a shred, since it's free and the best DPS you can do. The difference at this point would be 4 shreds vs 5 mangles.

At 10k AP and a 40% crit rate, each shred does 7971 damage on average (including crit) with savage roar up (14k AP) Each mangle does 5400 damage with savage roar up. 5 mangles is 27000 damage. 4 shreds is 31884 damage. Thus the difference in damage is 4884 damage for that part of the cycle.

How much do we lose for the part where we don't have SR up? Each normal attack at 10k AP does 1268 damage on average. At 14k AP it does 1733 damage. Assume that for the shred cycle we have to do two attacks without SR up, and they are shreds. This is 84 energy + 25 for the SR, or 109 energy. We have leftover energy from the prior cycle - 26 energy. So we need a total of 83 energy to do this, or 8.3 seconds of time. In that time we will do two unboosted shreds and 8 normal attacks. Unboosted, Shred does 6338 damage on average. The total damage for this period is 23206 damage. The 'lost' damage we would have had had these attacks be boosted was 27066 damage, so we've lost around 4k damage.

Now, let's go with the non-shred cycle. With this cycle, we can assume at best one combo-generating attack that is unboosted, and we'll go with mangle. This is 34 energy. We end up with 22 energy at the end of the cycle, meaning we have to wait 3.7 seconds for the energy to do SR + mangle. This is a total of 8977 damage. The 'lost damage here is 11815 damage, so we are down 2837 damage. That's not quite enough to make up for the loss during the main cycle, so in this case shredding wins.

Does this seem about correct? The problem I'm having is coming up with a reasonable model to illustrate when mangle spam will catch shred due to the attacks/specials not having SR - or if it ever will.

 

Promethius

Numbers were based off Nightcrowler's previous analysis, so the validity of that raid buffed is on the accuracy of his simulation.

Are you sure about the swipe scaling?

is the code in Rawr. and The Druid Wiki » Druid_Game_Mechanics seems to imply around a 7% ap scaling coefficient as well. I can't remember where I originally found the ap/14 as swipe's scaling coefficient, but these documented numbers seem to imply it's closer to that than 0.057.

 

Tappin

Nobody else really loves high armor on necks, or trinkets. They would rather have more sta, defense, dodge/parry or even str. How often do you see other tanks wearing badge of tenacity?

The more important factor is that our weapon is the one major thing we can change in combat. Paladins, warriors and druids can swap from tank shield/weapons to dps. We WANT a tank weapon that is different from our dps weapon. This allows us to better fill our role swapper role: offtank with your bear staff, 'go cat' and swap to your dps weapon.

It makes a lot more sense to have our version of 'sword and board' swappable via armor weapons and remove ridiculously inflated 4k+ armor trinkets or high armor necks. We're already limited in our selection of weapons due to the FAP mechanic. Why not make our selection in the other slots more open.

Remove armor from all necklaces, tone down the dual insane-armor trinkets, and put the armor bonus back. Now we won't cap as early and share more trinkets and necklaces.

 

seminarca

Honestly, I see what you're saying .. but I don't agree with the "being chained to armor jewellery with no upgrade path" bit. The entire basis of the argument is that there isn't going to be an upgrade path for armor jewellery, whereas there is absolutely no evidence to suggest this (and no, historical failures such as Wildfury Greatstaff and Badge of Tenacity in Sunwell don't count). On the contrary, in WotLK there's another class that can make decent use (if not as much as Druids) of bonus armor, not to mention GC has stated/insinuated there will be progression for armor items this time. There is exactly 1 epic trinket (that we know of) in the ilvl 200-226 range, but as long as they exist in each tier (better than the ones before), progression exists (also for other slots).

AP based pseduo-block mechanics have been suggested (I mentioned this previously), and then summarily ignored by Blizzard. It's now less than 2 weeks away from patch 3.0 (if it coincides with the arena thing) and less than a month and a half away from WotLK release. Do you realistically think that at this stage Blizzard is going to go back, redo a crapton of itemization, fiddle with numbers on Druid and DK base skills & talents and hope there aren't any further cascading effects that aren't immediately apparent? Vs simply turning a knob that says "give Druids less armor"?

It's clear you are fairly passionate about this topic, and that's fine, I'm not trying to change your mind on this or anything. Just saying that given the current circumstances, the solution Blizzard is apparently aiming for is probably going to have to be good enough.

 

Mijae

Your OoC estimates are a bit low. Depending on total haste, you should expect at least 2 procs in 30 seconds (more like 2.5 in 34 sec). Your crit rates are also a bit low, ~45% in blues or ~53% in epics with full buffs.

With 50% crit and 2 OoC procs (low estimates), the Shred model will average 15 CP per cycle. Replacing Shreds with Rake and the number goes up still. I would argue not to base actually cycles on 5Rip/5FB, but on time left on SR. Each finisher should take 10-12 seconds. If you have a bad non-crit streak, reduce one or both down to 4 CP. It just requires more dynamic cycles. If your CP generation is really that low or inconsistent, dropping FB is a better option than Mangle spam. The DPE loss (without 2t6 & imp mangle) just isn't worth it.

Feral DruidStats spreadsheet
v3.2.0.0.xls - updated 8/4/2009

 

nightcrowler

wait. You only considered missed attack but you will have:

Missed % (Total): 7.41415%
Dodged %: 0%
Not normalized attacks parried by the boss: 7.25%

so the avoided attacks are at least 14.66%, also I haven't yet simulated block (I dunno how block works regarding threat).

On a side note:
I don't how people come out with numbers but usually I see for Paladins and Warriors too low numbers, but the same people push out low numbers also for druids so It's not strange if a paladin have higher stats than the one reported in the previous post.

 

kalbear

Thanks, Mijae. I was trying to go for a more pessimistic case, but 2 OoC procs does change the scenario a bit. In that case I wonder if the manglespam could reasonably go for 3 finishers in the 34 second uptime. Probably not, as this would start bumping into the GCD limits fairly quickly, but it's possible with a decent enough crit rate and haste rate. Something on the order of 18 CPs during that time would be the requirement. Especially true if you throw in the 2pT6 bonus.

 

Promethius

Under that case, including the parry and the forgotten naturalist multiplier, we'd have:
(base + ap/14)*FI*(1-armor reduced)*(landed hits)*(hit+PI meta gem crit mod)*naturalist
(108 + 6636/14)*1.3*(1-.261969)*(1-.146645)*(1+0.372895*1.26)*1.1 = 770.4 avg dmg per swipe
770.4 dmg/swipe * 1 swipe/1.5 sec = 513.6 DPS
513.6 DPS * (1.45/0.7) threat/dmg = 1063.9 TPS

This means that for non boss level targets, swipe is even stronger than originally perceived, since parry and miss rates are considerably lower. If avoided attacks were lowered to 3.4% (do even level mobs only have 5% parry?), the swipe numbers go up to 581.3 DPS and 1204.2 TPS against even levels before even factoring in effective increased crit and lower mob armor.

Max rank consecrate is (.32*(ap+spellpower)+904)/8 for base DPS, times 1.1 one handed spec multiplier, with a 1.13 raid buffed spell dmg modifier and 2.72 threat modifier. A paladin with 4k ap and 1400 spellpower ends up with 408.9 DPS and 1112.3 TPS. Those numbers seriously look more like what a ret paladin would have opposed to a prot paladin, but even if prot had those kind of stats, they're still be lower on DPS with a slight lead on TPS against boss level mobs and still slightly behind overall against even levels.

Main point is, unless paladins are suffering with aoe threat (in which case either dps' aoe or all tanks' aoe threat needs to be looked at), swipe should be sufficient for aoe purposes. Yeah, it'd be nice if it were tripled in damage and on a 6 sec cd so you don't have to spam it, but at that point, all you're really asking for is a mangle that hits everything 8 yards in front of you.

 

Mijae

I'm assuming those numbers were based on a level 83 boss with full raid buffs/debuffs. In an AoE pack of level 82s, the damage reduction will actually be higher (around 36-40%) due to lack of sunders and FF. Similarly, if 3% of that crit came from Heart of the Crusader it would not be present for an AoE pack. The AP probably assumes BoM or Battle Shout, SoE totem, and Unleashed Rage. These are probably safer to assume present, but should be noted. Most bear sets should include a decent amount of expertise and at least some hit rating, so 2-3% avoided is about right.

Feral DruidStats spreadsheet
v3.2.0.0.xls - updated 8/4/2009

 

Grubsnik

the AoE whine
Oh noes, swipe spam can't out tps the prot paladin... we need fixing. Or wait. The lock is not standing inside the crowd, so needs to out-tps us by 30% to pull aggro (he doesn't). He doesn't have do his damage instantly, so we will still be ahead of him, and finally, he has a threatdump (soulshatter) on instant cast.

3500 AP in bear form? Wearing nothing but the Staff of the Plague Beast, from naxx10 you get 3000 AP. This is naked stats i'm talking about.

Absolutely worst case, the lock will have to go for a slightly slower on the seed spam, does that really break our viability?

Armor and jewelry
I think there are a few things we want really.
1. We don't want to be forced into a single upgrade path, the fact that i can filter all non-leather items who doesn't have +armor, without ever considering what they do, is a problem.
2. We don't want to hit the armor cap. Especially not in T7.
3. We would still like to be the "high-armor" tank.

My suggestion:
1. Remove the bear armor modifier from all jewelry slots (neck, rings, trinkets). Let it stay on weapons and cloaks.
2. Adjust bear armor modifier to compensate. Yes, this means it should go up, not down. Goal should be to leave us with ~10% more armor than other tanks, without armor from jewelry.

My suggestion is simple and relatively straightforward to implement (bearform armor modifier already ignores armor bonuses from agility). This means it's viable to become implemented. It does not require changing itemization across the board, and does not affect other classes.

If anyone wants to repost the part about armor and jewelry in the us-beta boards, i'd appreciate it.

 

summlan

I totally agree with you grubsnik, i dont feel like i need to be on par with protpaladins reguarding AOE-threat, as far as i know blizzard said there would still be some difference between the tanks but not to the degree it would be impossible to do encounters without a certain tank (like illidan).

If it requires the dps to focus one target first before unloading their AOE im perfectly fine with it. Unless feraltanks will be less wanted because of aoe-threat being to low i couldnt care less if our swipe isnt on par with other tanks abilities for aoe-threat. If we can get the job done im more than happy with it.
One could argue that swipe is alot better if mobs are suppose to be sheeped since i can just move the group of mobs a couple of yards and continue swipeing, with a paladin you have to wait until consecrate is gone to sheep the mobs.
Or that we can easily run around with our pack of mobs continuing building the same threat with swipe where the paladin have to keep them in one spot for consecrate to work.

 

nightcrowler

As asked above I've made dps/tps reports for different configurations.


The following reports has the following assumptions:

T7.10 level gear (only level 200 and 203 items). A full stacked raid with 10% ap buff working. All the possible consumable, the best in slot enchants, epic gems (matching socket bonuses), all the "minor" buffs like heroism, devotion aura, drums and so on.
The bear simulations doesn't take into account on use items and barkskin.

The simulations relative errors are about 0.25%

Specs:

FULL CAT
FULL BEAR
HYBRID

P.S.
Average swipe damage (from simulated values) in tank gear with naturalist and taking into account misses, dodge, etc. (no blocked value) is 576 so the tps is 795.

 

Mihir

Atm even level mobs are at 5%, and for every lvl of difference 0.5% gets added.

 

Sadirin

I am very grateful, thank you very much. It seems, that the "Hybrid"-Spec with specialised equipment seems to be the way to go, delivering both high DPS and "close to top tanking" abilities. :-)

So, whats your opinion on mangle vs shred spam?

SYL

 

Malazaar

I think the issue has been misunderstood by many people. The question never really was if mangle could be a feasable substitute for shred at raid situations. My table some pages back clearly states that with RnT shred is still a good deal ahead of mangle.

Mangle spam is more interesting for leveling and at entry level content, especially if you are still wearing 2t6 for the additional bonus.

 

Sadirin

Ah, thats sad. Well, back to the drawing board then and practicing the new rake-rotation. My fingers are already hurting. :-)

 

Helban

That was a very productive post nightcrowler, thank you. I have just read the data into an analysis tool, and over all it shows your hybrid spec with specific gear to be a complete winner, which has always been a personal preference of mine so the news is good to these ears. Any idea how removing RandT + IW and bringing in NS + MS to the hybrid build would affect those numbers? Something like this build

Proposed Hybrid Build