Remove Advertisements

On Progression MT Gearing

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

On Progression MT Gearing

Postby Meloree » Sat Aug 29, 2009 11:53 pm

I'm starting this topic as a discussion thread for an interesting tangent that cropped up in Theck's excellent expertise mathcraft thread. The start of the tangent can be found here: viewtopic.php?f=6&t=25611&start=30

I suggested, only partially facetiously, that WotLK had reduced protection paladin gearing to exactly 2 stats for progression maintanking, Stamina and Armor. My position on this is built primarily on personal experience, and discussions with my healers, and the fact that EH gear tends to have threat and avoidance built in, in sufficient (in my opinion and experience) quanties to reduce the rewards of further stacking compared to good old fashioned health.

I'm not sure that any really in depth mathcraft has been performed on maintank optimum survival gearing theory recently, or at least I haven't run across any in my lurking. Given, then, that I'm speaking mostly from experience and personal perspective in formulating my current position and thinking on the subject, I'll share some background.

I'm the MT and RL for Edge of US-Garona. We're 9/9, and ranked US-47 according to wowprogress.com, although I expect that ranking to fall somewhat as a couple of guilds that beat us to everything else kill Yogg0. I have progression tanked every hardmode kill for Edge, although for both Thorim and Vezax it was not in an MT role, as they were extremely unfriendly to 3.1 era protection paladins.

I have found in my experience that my survival over the course of any given fight is best served by basically pretending that every socket is blue, and that health is the best option for any enchant once over the defence minimum of 536. Yes, I consider the resilience/stam shoulder enchant to be strictly superior to dodge/defence. I'm not totally comfortable with this opinion, and I'd love any of this boards brilliant theory and mathcrafters to make some argument that allows me to appeal to my sense of efficiency and start picking up useful socket bonuses again, it feels like I'm leaving itemization on the table, but it's also, in my opinion, lower-value itemization than the 6 or 9 stam I pick up by ignoring said bonus.

Now, to pick up where the previous discussion left off:

Knaughty wrote:For hard content in Ulduar, the first two things you do are have the minimum number of tanks (little flexibility) and then minimum number of healers. Every hard-mode I can think of has an in-play enrage timer - you have to stack DPS to win. Other than GV, they also tend to have a ton of raid damage.

Raid healers switching to the tank just leads to dead raid members.


But this isn't true at all. Looking at the tank-killing encounters: The tank is in the most danger at Steelbreaker just after a fusion punch, but the raid is in the most danger right after Static Charge lands. The two dangers are at least partially decoupled. At Algalon, the raid damage is predictable and paced, and the raid healers can spare a GCD for the tank at any time they choose without danger to the raid. At Yogg0 the raid is taken care of by splash healing. At Mimiron the danger to the tank is in P1, and the danger to the raid is in P2 and P4. At Thorim you use a feral druid. Either way, with a 24% avoidance penalty during the danger time, avoidance loses a lot of it's value.

cordelia wrote:I have seen many tank deaths 4-6 seconds after a spike begins. Isn't 4 seconds long enough for a healer to react? Their shortest spells are all <2 seconds. Isn't 6 seconds long enough for them to react? By your statement, we should never see tank deaths where the spike string lasts longer than 3 seconds. Maybe my guild just has poor healers, but we do fairly well in progression, 5/9 hardmodes, and 3rd on the server, so the evidence doesn't support that.


cordelia wrote:
Knaughty wrote:And tank health spikes up and down hugely in Ulduar. How are the raid healers supposed to tell the difference between "normal" wildly swing tank health and "tank will die" wildly swinging tank health? It seems a bit "hindsight is 20:20". Tank dies, raid leader asks why, tanks and healers don't know, because everything seemed to be going fine. Healers were healing, tank was facing boss, etc. Then tank reads his log and discovers he collected 4-5 of the last 6 hits, and didn't get topped off for 6 seconds or so. If you had to consult your log to work out why you died, it's irrational to assume a raid healer can notice and switch targets.


This is exactly what I see in practice. 2-4 seconds in, the tank is healed up to 70-80% health, but that still means the tank is susceptible to being 2-shot, which is what inevitably happens.


It's certainly not inevitable. I'm also not sure that this doesn't argue into my point pretty effectively. What if a tank isn't 2-shottable at 70% health? What is 70% health? How much does the boss actually hit for? If it's anywhere under 50% of your healthbar, there's no reason not to expect enough healing to land in the time between hits to survive said 2-shot, no matter what the time between hits is. Again, we're speaking far too generically here, but I'm not certain of the correct scenario(s) to build to provide a better discussion platform. I'll share a bunch more information, though, for the mathcrafters to pick apart, because I really am fascinated by this discussion, and if I can improve my performance and survivability, I'm all about that.

My armory: http://www.wowarmory.com/character-shee ... &n=Meloree
The short version: 41879 hp, 25720 armor, 552 defence, 25.35% dodge, 20.56% parry.
I tuned in a little more armor for physical heavy fights, that setup has algalon legs over saronite, and Titanskin Cloak over Shadowed Sun.

I've stuck stamina in every socket I could find, with the exception of the token red, and one def/stam gem in boots for a 9 stam bonus, because I need the defence in some gear setups. I've enchanted stamina and health everywhere I can, including 18stam to shield, because in my experience 1 point of stamina is worth much more than 1 point of any rating, over defense minimum.

I'd love for some of the great theory and mathcrafters here to weigh in, Maintankadin is where I learned to tank, and I have an immense amount of respect for the site and the body of knowledge and experience represented here.

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby Serv » Sun Aug 30, 2009 2:02 am

I can't imagine to gear for anything but max Stamina/Armor. I started out as a twink when 3.1 went live, switched realms and my long loved fun to play pally became the onyl toon to play. I didn't switch to a guild hard on progress. We are like a family and a lot of "older (35+)" players in it. There I learned the harway that avoidance wasn't the way to go - and blocking was neither.

It was the hardest time for me. Trying to get accepted as a tank with my twink gear and all that crappy stuff. In the end it all worked out and every ID i could drop some of my block gear and since a few weeks my healers hug me every time I join the raid. I can just confirm what Meloree was stating: my healers rather like to heal a manasponge than an avoidance tank.

'dunno if that was what you would like to hear on that threat.
Image
Thanks to LAVE for the sig

www.rent-a-tank.net ...:::... www.nil-gilde.net
Serv
 
Posts: 228
Joined: Fri Jun 05, 2009 8:50 am

Re: On Progression MT Gearing

Postby Nadir » Sun Aug 30, 2009 4:28 am

I believe where one comes down on hybrid gemming versus pure stamina rests largely along an experience fault line. Most bleeding edge progression tanks focus on the latter, while more average level tanks focus on maximizing things like "tank points" or whatever metric that justifies the use of hybrid gems to meet socket bonuses.

I concur with your opinion, and unsurprisingly I'm part of that latter group. I'm somewhat interested though, is there a top end Protection Paladin that raids with a high avoidance setup?

Let's use your character as any example.

Your original character has 41879 HP and 54.52% total avoidance.

Let's switch out your STA enchants to chest/shield/shoulders for the 'standard' avoidance alternatives. Every red is gemmed with a Regal Dreadstone and every green is gemmed with an Enduring Eye of Zul.

"Hybrid" Meloree has 39884 HP and 57.44% total avoidance, or a 1995 HP loss for a 2.92% avoidance gain.

What's the better progression tank? I don't know how to objectively answer that question. Anub'arak hard mode is going to be a repeat of Yogg 0 in terms of tank spike damage (8+ adds that have a proximity damage increase buff), so I'd lean towards squeezing as much stamina as theoretically possible out of your gear.
Image
Nadir
 
Posts: 266
Joined: Mon May 04, 2009 8:37 am

Re: On Progression MT Gearing

Postby Meloree » Sun Aug 30, 2009 9:15 am

Nadir wrote:I believe where one comes down on hybrid gemming versus pure stamina rests largely along an experience fault line. Most bleeding edge progression tanks focus on the latter, while more average level tanks focus on maximizing things like "tank points" or whatever metric that justifies the use of hybrid gems to meet socket bonuses.

I concur with your opinion, and unsurprisingly I'm part of that latter group. I'm somewhat interested though, is there a top end Protection Paladin that raids with a high avoidance setup?


Your definitions are a bit soft. I'm not sure, for example, if I qualify. Either way, Nutron and Markoh both match sockets (at least on armory right now) and they're certainly more progressed than I am.

Nadir wrote:What's the better progression tank? I don't know how to objectively answer that question. Anub'arak hard mode is going to be a repeat of Yogg 0 in terms of tank spike damage (8+ adds that have a proximity damage increase buff), so I'd lean towards squeezing as much stamina as theoretically possible out of your gear.


Being able to objectively answer the question would make me a happier tank. Even if the answer comes down in the way we both seem to already lean.

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby Panzerdin » Sun Aug 30, 2009 2:10 pm

It seems to me that the best answer to this question would be to simulate likely progression scenarios hundreds of times over in something like MATLAB and see which gearing option comes out top.
User avatar
Panzerdin
 
Posts: 5504
Joined: Sat May 24, 2008 2:51 pm
Location: On a picket line, protesting against the changes to Maintankadin

Re: On Progression MT Gearing

Postby Meloree » Sun Aug 30, 2009 3:03 pm

Panzerdin wrote:It seems to me that the best answer to this question would be to simulate likely progression scenarios hundreds of times over in something like MATLAB and see which gearing option comes out top.


I don't envy that task at all. Even getting started on building theoretical models seems pretty daunting, I wouldn't even know how to get started. What simplifying assumptions can you make without risking significant changes in the results? The whole system would seem to be hugely dependant on how you build the healing, and what kind of ability to react you give the healers.

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby Panzerdin » Sun Aug 30, 2009 3:07 pm

The way I see it, the safe assumptions are as follows:
1. The boss will deal you regular autoattack damage
2. The boss will deal semi-regular or random special damage
3. You will have healers who react to damage in a certain manner
4. These healers will have reactions roughly similar to those of an average human

Based off of those, you could probably make a decent model.
User avatar
Panzerdin
 
Posts: 5504
Joined: Sat May 24, 2008 2:51 pm
Location: On a picket line, protesting against the changes to Maintankadin

Re: On Progression MT Gearing

Postby Meloree » Sun Aug 30, 2009 3:58 pm

Panzerdin wrote:The way I see it, the safe assumptions are as follows:
1. The boss will deal you regular autoattack damage
2. The boss will deal semi-regular or random special damage
3. You will have healers who react to damage in a certain manner
4. These healers will have reactions roughly similar to those of an average human

Based off of those, you could probably make a decent model.


Well, 1 and 2 you pretty much pick your boss to model. There's more than enough information to build a fairly accurate Algalon, for example. 4 leaves a lot of room open for interpretation. If my healer plays at 50ms, and yours plays at 250ms, and you add an average human decision making cycle, mine may get the critical heal off, and yours may not. But the big point for debate, then, is how do those healers react? From the other thread, it appears that Knaughty and Cordelia have different expectations for healer reaction than I do.

Further, there's also a lot of "small" heals going around that really add up over time, but aren't particularily easy to model. How often does DS hit the tank on a particular fight, as opposed to just melee. Can you rely on a CH bouncing to the tank out of melee in the middle of a spike due to it's smart behaviour? What hots do the raid healers leave ticking on tanks. Is there liable to be a wild growth in there on occassion? All that stuff adds up, and it all makes a difference. It all closes the healing gap required to survive spikes, so it's all important, at least intuitively.

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby theckhd » Mon Aug 31, 2009 7:48 am

Nadir wrote:"Hybrid" Meloree has 39884 HP and 57.44% total avoidance, or a 1995 HP loss for a 2.92% avoidance gain.

What's the better progression tank? I don't know how to objectively answer that question. Anub'arak hard mode is going to be a repeat of Yogg 0 in terms of tank spike damage (8+ adds that have a proximity damage increase buff), so I'd lean towards squeezing as much stamina as theoretically possible out of your gear.

I suppose that all depends. If he's gone from 54.52% avoidance (57.44-2.92) to 57.44% avoidance, then he's reduced his damage intake to 93.58% of his previous value (1-0.5744)/(1-0.5452) = 0.9358. In other words, that 2.92% avoidance reduced his incoming DTPS by 6.42% (averaged over a fight). If he's never dropping below 2k life in his current set-up, then that's almost certainly worth the trade-off. But I'm also one of those tanks that tends to match socket bonuses (at least when they're worthwhile - I usually skip those with <9 stam or <6 avoidance in favor of pure EH).

I guess I should really get to writing that "Gemming Philosophy" theorycraft article that's been on the back burner for a week or two. Though it would really fit the context of this thread quite well.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7753
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: On Progression MT Gearing

Postby Meloree » Mon Aug 31, 2009 9:46 am

theckhd wrote:In other words, that 2.92% avoidance reduced his incoming DTPS by 6.42% (averaged over a fight). If he's never dropping below 2k life in his current set-up, then that's almost certainly worth the trade-off.


It's actually even more than that. Those aren't raidbuffed avoidance numbers, add the 6% or so from raidbuffs and the 2.92% avoidance is even stronger. It's even the really cheap kind of avoidance, the kind where I'd be getting socket bonuses back. But that's also relatively immaterial. My DTPS taken over the course of any given fight (excepting a Vezax-ish one) is fairly immaterial. Certainly in farm Ulduar I already have more stam, more armor, and more avoidance than I need, but the question isn't really about optimizing farming, and 6.42% lower average dps (or 8%) isn't going to save me a healer.

So, onto the second clause. I almost never drop below 2k life in any content I currently have access to. I certainly agree that my tank setup is not optimal for farming. I don't know if I'm going to drop below 2k life in heroic Colisseum. I don't know if any of the fights are going to qualify as "tank checks" at all. So, frankly, I don't actually know what gearing setup is better for that yet. I do know where I'm leaning for progression. And I even know why. A lot of burst is unavoidable. One assumes Icehowl keeps his stun in heroic. One assumes further that with latency, you're going to take stun+melee at least, even if you're tank-swapping. I dunno how hard he hits for, but I'd kick my own ass if the total was 52.9k and I went and regemmed as above.

theckhd wrote:But I'm also one of those tanks that tends to match socket bonuses (at least when they're worthwhile - I usually skip those with <9 stam or <6 avoidance in favor of pure EH).


Okay, and you're the resident math god and that holds a lot of weight (with me, anyway). Given this, though, you're ranking the value of 1 stam as less than 1.667 dodge/parry defence rating (6 stam for 10 something else from a 9 stam bonus), but more than 1.11 (you take 9 stam over 10 of something else). But given that, it's inconsistent to pick up a 6 avoidance bonus at the cost of 15 stam.

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby Salamandra » Mon Aug 31, 2009 10:18 am

It'd be getting 16 avoidance at the cost of 15 stamina. Still inconsistant, but only by a very small margin.
Image
User avatar
Salamandra
 
Posts: 373
Joined: Mon May 12, 2008 5:45 pm

Re: On Progression MT Gearing

Postby Meloree » Mon Aug 31, 2009 10:39 am

Salamandra wrote:It'd be getting 16 avoidance at the cost of 15 stamina. Still inconsistant, but only by a very small margin.


That's what I said. 6 avoidance bonus. That said, for small margins, in the same price range is PVP shoulder enchant (1.16 avoidance per stam), shield enchant (1.11 avoidance per stam), chest enchant (275 health, roughly 20 stam... 1.1 avoidance per stam). If you're NOT picking up 6 stam options, those are all in the ballpark of skipping a 6 stam bonus.

EDIT: I'm not trying to be difficult. But the question on the table is, really, how does one value these stats to increase your chances of success in progression fights. If even Theck doesn't have a consistant policy, what hope do the rest of us have?

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby theckhd » Mon Aug 31, 2009 10:51 am

Meloree wrote:
theckhd wrote:But I'm also one of those tanks that tends to match socket bonuses (at least when they're worthwhile - I usually skip those with <9 stam or <6 avoidance in favor of pure EH).


Okay, and you're the resident math god and that holds a lot of weight (with me, anyway). Given this, though, you're ranking the value of 1 stam as less than 1.667 dodge/parry defence rating (6 stam for 10 something else from a 9 stam bonus), but more than 1.11 (you take 9 stam over 10 of something else). But given that, it's inconsistent to pick up a 6 avoidance bonus at the cost of 15 stam.

That's because you're looking at it in terms of Stamina, and not itemization points (or ilvls).

Each point of stam is worth 2/3 of an item level (or 3 stam = 2 ilvls)
Each point of rating is worth 1 ilvl.

My statement could then be reworded as follows:
"I tend to skip socket bonuses that provide less than 6 ilvls of benefit at the cost of EH."

But I'm a little over 2/3 of the way done with my Gem Theory post, so that should clear up all of this very shortly.

<edit> Just to check for consistency using ilvls:
A 9 stam bonus costs you 6 stam (4 points) and nets you 10 of something else (10 points), for a bonus of 6 points. I take the bonus (>=6 bonus of avoidance at the cost of EH).
A 6-stam bonus costs you 9 stam (6 points) and nets you 10 of something else, for a bonus of 4 points. Hence why I skip the bonus in this case (<6 bonus of avoidance at the cost of EH).
A 6-avoidance bonus costs you 15 stam (10 points) and nets you 16 of something else (16 points), again for a net bonus of 6. I take the bonus.
Last edited by theckhd on Mon Aug 31, 2009 11:15 am, edited 1 time in total.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7753
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: On Progression MT Gearing

Postby Meloree » Mon Aug 31, 2009 11:09 am

theckhd wrote:That's because you're looking at it in terms of Stamina, and not itemization points (or ilvls).


I'm well aware of all that. I think I failed to communicate my point.

The value of Stamina in terms of avoidance is not fixed at 1.5 stamina = 1 dodge rating = 1 parry rating = 1 defence rating for survival in progression encounters. If it were, you wouldn't have spent all that time on doing the math for when one avoidance stat is better than the other. Obviously you value stamina higher than it's ipoints equivalent of any avoidance stat, as you skip 4ipoint bonuses for pure eh. But given that, it doesn't make sense to look at said bonuses in terms of ipoints. Hence, translating it all to the "stamina cost" gives a more stable reference than ipoints. And given that, your valuation of stamina isn't consistant. Does that make any sense?

--
Meloree, <Edge>, Garona
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: On Progression MT Gearing

Postby theckhd » Mon Aug 31, 2009 11:56 am

Meloree wrote:The value of Stamina in terms of avoidance is not fixed at 1.5 stamina = 1 dodge rating = 1 parry rating = 1 defence rating for survival in progression encounters. If it were, you wouldn't have spent all that time on doing the math for when one avoidance stat is better than the other. Obviously you value stamina higher than it's ipoints equivalent of any avoidance stat, as you skip 4ipoint bonuses for pure eh. But given that, it doesn't make sense to look at said bonuses in terms of ipoints. Hence, translating it all to the "stamina cost" gives a more stable reference than ipoints. And given that, your valuation of stamina isn't consistant. Does that make any sense?

I've edited the previous post already to address the consistency question. My choices are completely consistent with the statement I made about choosing items based on ilvl benefit. You could argue the statement is illogical or that it's stupid to value things in terms of ipoints, but not that I was being inconsistent. :P

As far as survivability goes, it really doesn't matter whether 1 dodge = 1 parry = 1 defense rating for this exercise; in general you'd pick the type of avoidance that matches the socket, and in the case of dodge and parry you'd use whichever gets you closer to the ideal 1.88 ratio, because that gives you the biggest benefit.

I think the issue is that the rule I use doesn't have a fixed cost (like 1.5 stamina = 1 avoidance rating) for evaluating upgrades. Because of the way I throw out small socket bonuses, the value of stamina compared to avoidance rating will vary, leading to the inconsistency you perceive. If I stuck to a pure 1.5 stam = 1 avoidance rating rule, I'd grab every socket bonus I could, because that would give me the max number of points.

Really, what I'm doing is saying, "We want avoidance, and we want EH. In fact, we want lots of both. But we do like EH slightly more, and so we're going to avoid making trades that sacrifice EH unless we're getting more than our money's worth of avoidance." Alternatively, you could think of it like this: "We want to go into this fight with the highest-ilvl gear we have, with a good balance of EH and avoidance, but with a slight edge towards EH." If you look at it on the average, across all of your gems, you'll see that this strategy does give a slight edge to EH, despite making one or two trades like the 16 avoidance for 15 stam one that fall below a strict numerical X stam = Y avoidance rule.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7753
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Next

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: Bing [Bot] and 1 guest


Remove Advertisements

Who is online

In total there are 2 users online :: 1 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 380 on Tue Oct 14, 2008 6:28 pm

Users browsing this forum: Bing [Bot] and 1 guest
?php } else { ?