Remove Advertisements

Hit Rating and Expertise

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Hit Rating and Expertise

Postby theckhd » Wed Mar 16, 2011 6:39 am

I'd also like to take a second and point out that this entire argument did not grow out of me saying "there's never any reason to gear for hit/exp." It grew out of an absolute statement from Dem:

theckhd wrote:
Dem wrote:Gearing for heavy survival in farm content is sub-optimal.


Er.... no? Farm content is farm content, but you can still die. Unless that extra DPS somehow makes or breaks a farm kill (it shouldn't), there's no solid argument for shifting to DPS-oriented gear other than "it doesn't matter anyway." You'd be hard-pressed to convince anyone that it's truly sub-optimal.


In other words, I contend that there is no proof that heavy survival gearing is sub-optimal for farm content. You could argue that heavy survival gearing is not needed, so you can afford to boost your DPS. But trying to strictly state that it's sub-optimal is grossly misleading. As I said, we "farm" Nefarian every week, but it still has fairly stringent survivability requirements. We're far more likely to wipe on Nef because of a tank death than low DPS. I'll take one fewer wipe every month over a 5-seconds-faster kill any day of the week.
"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: 7763
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby tlitp » Wed Mar 16, 2011 8:09 am

theckhd wrote:Parse hunting almost universally happens on content that is well beyond "farm" level, so you can do a lot of things that just aren't possible during progression, including (but not limited to) running an extra healer to prop up the tank's survivability. Maybe it's just the experimentalist in me, but I have trouble taking any "experiment" seriously when it changes so many variables that it no longer resembles the control group.

Put yourself in the shoes of a parse-hunter for a moment. On your first attempts of parse-hunting a certain encounter would you directly throw (e.g.) 30k HP and 15% CTC out the window ? I'm fairly certain that you wouldn't; you'd start out with small incremental changes : replace one Sv-oriented trinket with a Dm-oriented one; replace the second trinket; progressively try to replace gems/reforging, etc.
The "important stuff" (changes in absolute thresholds, like assigning one extra healer to the tank) doesn't really happen from the very beginning. As such, I think that it's a bit unfair to state that parse-hunting doesn't have general applicability at all.

Again, let me reiterate that my intention is not to challenge the "conventional wisdom"; it's simply a matter of getting the methodological basis right. As a wannabe scientist ( :P ), I tend to become cautious when I see instances of "always" (or "never") because I know beforehand that exhaustive analysis is particularly troublesome.
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Hit Rating and Expertise

Postby theckhd » Wed Mar 16, 2011 8:52 am

tlitp wrote:Put yourself in the shoes of a parse-hunter for a moment. On your first attempts of parse-hunting a certain encounter would you directly throw (e.g.) 30k HP and 15% CTC out the window ? I'm fairly certain that you wouldn't; you'd start out with small incremental changes : replace one Sv-oriented trinket with a Dm-oriented one; replace the second trinket; progressively try to replace gems/reforging, etc.
The "important stuff" (changes in absolute thresholds, like assigning one extra healer to the tank) doesn't really happen from the very beginning. As such, I think that it's a bit unfair to state that parse-hunting doesn't have general applicability at all.


Fair enough, but realize that this is fundamentally flawed methodology for other reasons as well. As you add on more attempts, people are getting more comfortable with the encounter. Play quality is vastly different between the first few attempts and farm kill #50. In addition, gear is constantly improving while you do this, giving you a moving baseline that's nearly impossible to track. 30k HP might be enough at week #15, but might not be enough at week #10. Worse yet, you have player substitutions to consider - maybe 30k is enough when Alice is healing you, but you need a bigger buffer when Bob heals you because he reacts more slowly, or when Eve heals you because she's geared for longevity instead of throughput.

That's the difficulty with such experiments - it is very rare indeed to find a stable enough test environment to allow for meaningful results. There are just too many variables involved, many of which you have little control over.

tlitp wrote:Again, let me reiterate that my intention is not to challenge the "conventional wisdom"; it's simply a matter of getting the methodological basis right. As a wannabe scientist ( :P ), I tend to become cautious when I see instances of "always" (or "never") because I know beforehand that exhaustive analysis is particularly troublesome.


As you know, I usually share your distaste for universal generalizations like "never" and "always." So I should have been more careful about using them in this context. "Never" is just easier to type than "in almost any situation you will encounter in this tier of raiding." I will try to be more careful about this in the future, since "rarely" doesn't have the same absolutist connotations that "never" and "always" do.



Back to the general topic of this thread: I think it's clear that this tier of raiding is tuned in such a fashion that there's little incentive to gear for hit/exp. Threat is no longer of great concern due to Vengeance, and the encounters are tuned such that approaching them at-gear-level involves a significant risk to the tank. If you're clearing normal modes in 346 gear with a few 359 epics, you're in danger of dying. If you're clearing hard modes in 359 epics with a few 372's, you're definitely in danger of dying.

There might be a discussion to be had about DPS vs. survivability under certain conditions, but those conditions tend to be rare and very specific in the current tier of content. If we're going to have that conversation, let's be specific about it so that it's a meaningful conversation.

"Gearing survivability is sub-optimal for farm content" is far too broad and general to have any meaning, because "farm content" is too vague. Especially in this tier, where "farm content" can still be quite dangerous. We've been "farming" heroic Halfus and Chimaeron for almost two months or more, yet I still take a pounding on both of those fights and the threat of dying is very real. I know that the enrage timer won't be a problem, so what incentive do I have to decrease my survivability in that situation? What makes doing so "optimal?" I'd argue that if anything, it's two specific examples where survivability is "optimal" for farm content. Or at the very least, just as valid as any other choice if you're willing to accept that the chance of wiping is small.

Or perhaps more pertinently, can we come up with examples in the current tier where a tank's personal DPS is valuable enough to trade some survivability for? And is there some advantage to modifying your gear strategy rather than just switching seals/rotation? And what about the negative implications of that gear strategy for harder, more survivability-oriented bosses?
"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: 7763
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby Meloree » Wed Mar 16, 2011 10:17 am

Parse-hunter here (although not so far in T11, but I've been referenced).

My traditional approach to parse-hunting, which Theck has correctly diagnosed as "boredom relief" has been to trade survivability stats for threat stats down to roughly the level of stats that I had during progression/first kill. 35k hp at Festergut, with a 20% buff (I think that's when I first started toying with ICC farm stuff) had me pretty damned close to the tank stats (including avoidance/armor) that I had at our first kill at 0%. So, my general approach is to start by matching progression stats, and then going further from there incrementally. We've never added a healer, though, to help create tank-parses. That feels like cheating, a bit.

Dem wrote:The last time we discussed this, some people were surprised to find that this did happen and contributors such as Worldie and Meloree noted that they did gear for more threat.


I believe I'm being misquoted. I spent the vast majority of Wrath suggesting that you NEVER needed to gear for threat, the incidental stats that you picked up were more than enough. Unless we're talking about the parse-hunting, but that's something completely different. Certainly I've never geared for threat in Catalcysm, and in Wrath I only did it for laughs.

Firann wrote:However comparing it to a much better geared tank such as Meloree you can see that I have 5% more block while 9% less avoidance while being hit and EXP capped at the same time.


10-mans also tend to have a slightly different gearing paradigm than 25mans.

It's possible that I'll go parse-hunting sometime in T11, when we're done, and before T12 hits. In long progression instances, I do tend to do stuff like that. But it's not likely to happen anytime soon, and I've yet to see a boss that requires any kind of threat-stat-stacking - even now that Vengeance doesn't reliably sit at 100% anymore.

@tlitp: Parse-hunting for me has always been about challenging myself to plan well beforehand (correctly keeping survival stats above a minimum threshold), and executing well during. It's "fun" because it creates a challenge where none previously existed. Certainly it has some value to me, because it helps me to better understand my healers, and to prepare for the next tier of hard content by being artificially beaten on, and it helps me to shave off the fractions of a second in the rotation that give me extra breathing room for when threat does matter. So, it's a bit more than "loltankdps", for me personally. But I'm not sure how much value it has to the wider community.

EDIT: I ninja-edited. I typed 30% buff when I meant 20% buff.
Last edited by Meloree on Wed Mar 16, 2011 4:35 pm, edited 1 time in total.
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: Hit Rating and Expertise

Postby Dem » Wed Mar 16, 2011 12:04 pm

@Meloree - sorry if you feel misquoted, the relevant thread is here:
http://maintankadin.failsafedesign.com/ ... hilit=tank levels&rb_v=viewtopic&start=30
and you did mention it was more for 'stunts'.

@Theck

I think we're mostly arguing about definitions. I said that gearing without hit/exp is sub-optimal (definitions again) in farm content. Your contention is basically that there is no farm content yet. I guess that's up for debate and there will be some people who say they haven't wiped on some encounters for ages and some who say it's hit and miss. 'Farm' is a matter of judgement as we haven't agreed any definition for it.

In any case, experience from heroic 25 raids isn't that relevant to a discussion on 'farm' content. You mention giving advice to inexperienced tanks, but how many of these will realistically be raiding at that level?

Just so you know where I'm coming from, I did a lot of serious progression stuff in Wrath and I've gone casual for Cata - about 8 hours a week in a 10 man with others who don't want to raid 5 nights a week. My guild is well inside world top 200 (25 man) and my 10 man has cleared all normals and Halfus heroic.

You talk a lot about dps being insignificant but that's not really my point, it's partly threat and partly predictability and control which I'll get to shortly.

First though, threat. You mention that after the first 30 seconds of a fight this 'is never' an issue, but it's a factor in most fights. If you think about it, most current raid encounters require at least a second pickup. Atramedes, Maloriak (boss tank not add tank) and Al'Akir are exceptions. It isn't a big issue, but it's there. Chogall is a simple example - with enough hit it's possible to easily pickup the adds and do all the required interrupts even with a slacking/dead/MC'd interrupt dpser. Equally the tentacles get interrupted better with shield. Failing at these mechanics is a lot more dangerous (imo) than slightly lower defensive stats.

Really though, my point is about predictability and control. Before holy power, misses were annoying but if threat was fine and interrupts were covered then they weren't a big problem.
In Cata however, we have a key mechanic that is directly linked to our ability to hit something. We've been given the control to decide if we need more threat (single target or aoe) or more healing, every time we get holy power. Misses reduce our control. Missing on a regular basis significantly reduces this control.
Predictability is a factor in successful encounters. When people know what to expect, they act more rationally. This is going to be very difficult to factor into your calculations which is where intuition comes in.

If you want to talk about the benefit in progress fights then perhaps Halfus HM will be of interest. Good threat at the start it pretty important on multiple targets. An extra bit of block might be useful but equally it may be more important to get some of the drakes down before the tanks run out of cooldowns.

I'm not suggesting that people go out and get hit/exp capped, just that it should be a factor in their thinking and the decision is made with due regard to the strengths and weaknesses of the particular raid group.

We need to guard against 'groupthink' and make sure we're not dogmatic about gearing (or anything else). This game can't be wholly reduced to numbers and to make decisions purely based on what can be calculated is to miss the whole picture.
Dem
 
Posts: 137
Joined: Fri Jul 18, 2008 5:39 am

Re: Hit Rating and Expertise

Postby Meloree » Wed Mar 16, 2011 12:16 pm

Dem wrote:@Meloree - sorry if you feel misquoted, the relevant thread is here:
http://maintankadin.failsafedesign.com/ ... hilit=tank levels&rb_v=viewtopic&start=30
and you did mention it was more for 'stunts'.


I see. I guess I did technically say that I wore threat gear on deep farm - but I also specifically noted it had nothing to do with threat ceiling. I think that supports the convenience argument, in this case. Carry on.
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: Hit Rating and Expertise

Postby Malthrax » Wed Mar 16, 2011 12:43 pm

theckhd wrote: I will try to be more careful about this in the future, since "rarely" doesn't have the same absolutist connotations that "never" and "always" do.


Only a Sith deals in absolutes - Obi-Wan Kenobi, Jedi Knight

I have nothing constructive to offer to this discussion. I just love the delicious irony of the above quote.
User avatar
Malthrax
 
Posts: 727
Joined: Tue Feb 03, 2009 11:23 am

Re: Hit Rating and Expertise

Postby Kelaan » Wed Mar 16, 2011 4:06 pm

theckhd wrote:I contend that there is no proof that heavy survival gearing is sub-optimal for farm content. You could argue that heavy survival gearing is not needed, so you can afford to boost your DPS.

Also, if it's "farm" or "easy mode" for you and your DPS, your wearing more defensive-oriented gear allows it to be easier and more-farm for your healers, too. (:

Meloree wrote:My traditional approach to parse-hunting... has been to trade survivability stats for threat stats down to roughly the level of stats that I had during progression/first kill.

That makes a lot of sense. I don't yet have the gear to be able to trade out to "go back" to that level, as I'm just about at the same level of gear. ;)

Anyways .. thanks to you all for keeping this a civil and informative thread, even for tanks like me who are not rolling heroic modes (or have not even beaten Nef).
User avatar
Kelaan
 
Posts: 4036
Joined: Thu Jan 03, 2008 12:01 pm

Re: Hit Rating and Expertise

Postby theckhd » Thu Mar 17, 2011 7:17 am

Dem wrote:I think we're mostly arguing about definitions. I said that gearing without hit/exp is sub-optimal (definitions again) in farm content. Your contention is basically that there is no farm content yet. I guess that's up for debate and there will be some people who say they haven't wiped on some encounters for ages and some who say it's hit and miss. 'Farm' is a matter of judgement as we haven't agreed any definition for it.


I can agree with that. My definition of farm seems to be different than yours. To me, "progression" is a boss we haven't killed yet or that we're still struggling to kill reliably, and "farm" is something we go in and one-shot most weeks. So my definition of "farm" for my raid group includes 4 heroic modes. That makes survivability in "farm" content seem a lot more important than if you're limiting the definition to burning through normal modes in record time.

I can see the argument that once you're one-shotting all the normal modes, it doesn't really matter how you gear, so hit/exp are just as viable as anything else. If we're restricting "farm" to mean content where the chance of death is insignificant, your assessment of my attitude is correct - there really isn't any farm content yet unless your raid group is steamrolling the easier normal modes without trying them on heroic. I'd still argue against calling it "sub-optimal" to keep gearing for survivability in that situation though, simply because the metric isn't well defined. "Sub-optimal" loses a lot of its relevance, because there's no good metric with which one can define an "optimal" strategy when all roads lead to victory.

Dem wrote:First though, threat. You mention that after the first 30 seconds of a fight this 'is never' an issue, but it's a factor in most fights. If you think about it, most current raid encounters require at least a second pickup. Atramedes, Maloriak (boss tank not add tank) and Al'Akir are exceptions. It isn't a big issue, but it's there. Chogall is a simple example - with enough hit it's possible to easily pickup the adds and do all the required interrupts even with a slacking/dead/MC'd interrupt dpser. Equally the tentacles get interrupted better with shield. Failing at these mechanics is a lot more dangerous (imo) than slightly lower defensive stats.


Raid size seems to be a big influence on our disparate opinions. I don't usually get the "easy" job of staying on the boss, I'm generally the one dealing with add mechanics. So the 30-second 'threat period' is something I'm well acquainted with. And the truth is... it just isn't usually an issue in 25's. For anything the DPS will be focusing on immediately, I'll have a TotT or MD, and for anything they don't, I have two taunts and two ranged pick-up abilities. Judgement and AS can only miss, so failure rate on ranged pick-ups is fairly low. And there's no situation where I'm ever expected to be the only interrupter, because the majority of the 15 DPS players have one.

However, I can see how several of those factors may not be the case in 10-man, which makes hit/exp a much stronger argument. A tank that can reliably interrupt and/or reliably pick up adds without Tricks/MD can make an enormous difference on a 10-man fight, where your choices may be more limited.

Despite all of that, I still fail to see the wisdom in building hit/exp into your "survivability" gear set. It would make more sense to me to have one "all-out survivability" set as your default gear, and swap a few hit/exp pieces in for the fights where you might want them. I have almost an entire second gear set made up of leftover pieces where I've reforged the avoidance to hit/exp for the sole purpose of fooling around in heroic 5-mans. Reputation gear makes this relatively easy and cheap to do without relying on extra raid drops.

Dem wrote:Really though, my point is about predictability and control. Before holy power, misses were annoying but if threat was fine and interrupts were covered then they weren't a big problem.
In Cata however, we have a key mechanic that is directly linked to our ability to hit something. We've been given the control to decide if we need more threat (single target or aoe) or more healing, every time we get holy power. Misses reduce our control. Missing on a regular basis significantly reduces this control.
Predictability is a factor in successful encounters. When people know what to expect, they act more rationally. This is going to be very difficult to factor into your calculations which is where intuition comes in.


I think you're over-estimating the importance of that predictability. I originally thought it was incredibly important as well, not least because it makes it very easy to model the rotation. But in practice it ended up only being a small annoyance rather than a serious difficulty.

The restrictive nature of the rotation makes the threat cost of that predictability very small. You're talking about 2k-3k DPS at most by going from 8% hit / 26 expertise to 2%/10. And if you're using Seal of Truth, you have a very steady threat baseline from just auto-attacks, Seal/Censure, and the 3-second Crusader Strike heartbeat. All the predictability does is change what you do with the alternating cooldowns (SotR instead of other fillers, or an empty GCD).

Similarly, very few (successful) raid groups are planning their strategy around a the tank as a steady HPS source. And WoG/SoI healing is fairly insignificant compared to what a real healer is capable of. If anything, the tanks are better off not using WoG on cooldown, and saving it to react to burst damage, or to smooth spots where their real healer is incapacitated or has to move. This often means you're sitting on 3 Holy Power for a few seconds or more before using it anyway. The only thing that hit/exp buys you in that scenario is the ability to use it more frequently, which while convenient, isn't that big a survival boost. And since WoG scales linearly, you can always burn it at 1 or 2 Holy Power if you're in immediate danger.

Dem wrote:If you want to talk about the benefit in progress fights then perhaps Halfus HM will be of interest. Good threat at the start it pretty important on multiple targets. An extra bit of block might be useful but equally it may be more important to get some of the drakes down before the tanks run out of cooldowns.


Sure, but Halfus is also a serious tank check, since you're likely to be tanking either a drake and whelps at the same time or Halfus and a drake at the same time. Certainly, our first attempts at Halfus HM v1.0 were not limited by my ability to maintain aggro, but by my squishiness even in full survival gear. Also note that Vengeance stacked almost instantly on that fight because of the ridiculous amount of tank damage that one drake + whelps could put out.

I didn't change my gear for that, I just ran Seal of Truth for the extra threat/dps during the first drake burn. Glyphed Hammer of the Righteous is beastly enough that threat on whelps was never a serious issue, and pre-arranged Tricks of the Trade ensured that everything made it to me cleanly. After that, I switched back to Insight for the dangerous "tank a drake during Ferocious Roar" phase.

Note, however, that the new v2.0 mechanics have made this fight significantly easier. The drakes are significantly less dangerous, and you no longer need to release 3+whelps on the pull to have any chance of success. Even our 10-man alt run was able to complete heroic Halfus, despite having two very under-geared tanks the first week we did it. And even with those tanks, threat was never a problem (though we did have Tricks for the whelp tank). Tank-related wipes were always due to tank death during the damage-intensive first drake burn. To succeed, we ended up having to go with an unorthodox pull strategy - mage Pyroblast pulls Halfus and then pops Mirror Images to tank Halfus for 10-15 seconds, giving us more time to burn the drake and not build debuff stacks on the Halfus tank.

Dem wrote:I'm not suggesting that people go out and get hit/exp capped, just that it should be a factor in their thinking and the decision is made with due regard to the strengths and weaknesses of the particular raid group.

We need to guard against 'groupthink' and make sure we're not dogmatic about gearing (or anything else). This game can't be wholly reduced to numbers and to make decisions purely based on what can be calculated is to miss the whole picture.


Fair enough. I don't disagree with the sentiment, and I've certainly fallen victim to groupthink at least once before. It's also obvious that my attitudes and opinions are heavily influenced by my raid environment, since I can reliably expect Tricks/MD (or at least, know when not to expect them) and am rarely required to interrupt.

Though be aware that the same logic applies to the statement you made that started this whole argument. Calling survivability gearing "sub-optimal" for farm content assumes a lot of things about the raid group and content. I certainly don't feel like I'm doing my raid group a disservice by continuing to focus on survivability on our "farm" encounters.
"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: 7763
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby Dem » Fri Mar 18, 2011 6:48 pm

I think much of this can be resolved by agreeing that 10 man raiding has different requirements. There are a lot more 10 man raids now and perhaps we should give more consideration to this in advice give.
We don't gear in 25 and faceroll 10's anymore, these raids are separate and equal if not more challenging due to class and number limitations.

I never suggested that 'standard' gear should be modified however, as you suggest. Everyone should have a survival set for progress encounters. Threat pieces are easy enough to come by – expertise/hit trinkets, engineering head, exp/mastery drops etc. I have enough alternative pieces to customise my gear as needed and usually stick with 4 set for the (survival) tier bonus.

It’s also worth considering the bigger picture. I generally use elixirs and food along with gear changes. Elixirs are strong because the mastery flask is a battle elixir but in fact is defensive for us – say 3% block. Add to this a second defensive elixir, prismatic for around 10% less spell damage. This is a strong combination vs the stam flask for many situations and it’s easy (and cheap) to swop in hit or expertise elixirs and food where needed. By doing this, the need to change gear is reduced.

I still don’t think we’ve bottomed out the definition of 'farm'. Your definition of this including a number of heroic modes seems to be more about supporting an argument from authority (along with a few other comments in the above post). If I was still raiding full time in our 25 man then I'd be 10/13 heroics. Farm is not about normal or heroic mode. It’s about the ease with which you complete the encounter and an assessment of the chance of you dying (and possibly wiping the raid). Equally it’s not an all or nothing proposition. You don’t say ‘now this is on farm’ and instantly put on your 5 man heroic set. It’s about considering the encounter and then moving towards your ideal set for that rather than being forced into uber-survival gear as you usually are in progression fights. I have no issue with people’s judgement differing on how and when to do this but it is a judgement call.

I disagree with your assertion that “there's no good metric with which one can define an "optimal" strategy when all roads lead to victory.”

When all roads lead to victory you have many choices – take the quickest, take the easiest, take the most scenic and so on. Each of these may be optimal depending on your definition. Some of these such as ‘fastest’ can obviously be measured but other can't.

This gets into the realm of the art of tanking rather than the science. Some aspects of our craft can't be easily measured and it's probably worthy of a distinct post. The intelligent use of cooldowns, variation in task loading, impact of self healing options vs threat, predictability of rotation and so on should all contribute to our considerations.

The bottom line however is, if you feel squishy then throw on the progress gear. If not so much, then what can you change to make the encounter work better for your particular raid setup?
Dem
 
Posts: 137
Joined: Fri Jul 18, 2008 5:39 am

Re: Hit Rating and Expertise

Postby theckhd » Fri Mar 18, 2011 7:33 pm

Dem wrote:I still don’t think we’ve bottomed out the definition of 'farm'. Your definition of this including a number of heroic modes seems to be more about supporting an argument from authority (along with a few other comments in the above post). If I was still raiding full time in our 25 man then I'd be 10/13 heroics. Farm is not about normal or heroic mode. It’s about the ease with which you complete the encounter and an assessment of the chance of you dying (and possibly wiping the raid).


That seems a little below the belt. I think that "any encounter we can reasonably expect to one-shot on a weekly basis" is a pretty fair and widely-used definition of "farm." It doesn't make an overt distinction between heroic and normal modes. I don't see how that turns into an argument from authority.

If you're going to artificially limit "farm" to "any encounter we can beat even if I gear sub-optimally," then I'd say that you're the one tailoring the definition to suit your argument, not me. Or we can just agree to disagree on our definition of "farm," at which point this whole argument basically becomes irrelevant because we're talking about completely different things.


Dem wrote:I disagree with your assertion that “there's no good metric with which one can define an "optimal" strategy when all roads lead to victory.”

When all roads lead to victory you have many choices – take the quickest, take the easiest, take the most scenic and so on. Each of these may be optimal depending on your definition. Some of these such as ‘fastest’ can obviously be measured but other can't.


Except that none of those are universally agreed upon metrics. I think most people would agree that "what are our chances of killing the boss" is a useful metric that's relevant to the majority of raid groups. And it's also arguably the most-used for defining what an "optimal" approach to an encounter is.

However, if your success is guaranteed, the rest doesn't really matter to most people. You're not going to find much agreement on which is better - killing the boss quicker, killing him easier, killing him in a "scenic" fashion, or whatever other way you can think of. For the vast majority of players, none of that matters - what matters to them is that the boss dies.

Again, if you want to start making the argument that what metric we're using to define "optimal" is arbitrary, then what point is there in the discussion? Once you do that, you can't make any true statement that "X is sub-optimal," because it's going to depend on what particular metric your raid group prefers.
"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: 7763
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby Cascadian » Wed Mar 23, 2011 3:49 pm

theckhd wrote:However, if your success is guaranteed, the rest doesn't really matter to most people. You're not going to find much agreement on which is better - killing the boss quicker, killing him easier, killing him in a "scenic" fashion, or whatever other way you can think of. For the vast majority of players, none of that matters - what matters to them is that the boss dies.


There are some raid bosses or dungeon bosses, and especially some achievments, that are DPS races or timed against the clock. "Kill X number of mobs within 10 seconds". If survival is no longer a risk, then gearing for more Hit and Expertise could make sense in those scenarios. I am not doing this, I am maximizing Parry/Dodge/Block. I am merely offering a scenario where someone might gear for that extra tank DPS if their experience tells them survival is not an issue (because of skill of tank and/or healers).

If I recall, the primary reason for going heavily Parry/Dodge/Block was because of healer mana issues early in Cata. Everyone was getting gut checked in heroic dungeons and CC was considered mandatory in December 2010. Now we overgear the encounters and it is an AOE party like it's 2009.

At some point in Cata, I would like to see a change where we consider Hit and Expertise relevant again. Missing on interupts and losing holy power to misses is a step in the right direction. It caused people to go recheck the math. But the math still seems to say ignore Hit and Expertise. I think making rebuke interupts always land starting in 4.1 is going to be a mistake.

They should do something to give a reason to make tough choices in terms of gearing. Right now it doesn't exist.
Cascadian
 
Posts: 48
Joined: Tue Mar 15, 2011 12:44 pm

Re: Hit Rating and Expertise

Postby Syncognition » Wed Mar 23, 2011 4:25 pm

It surely has been said before, and I don't necessarily disagree with you in principle Cascadian, but rather than try to maximize the DPS of your Tank, wouldn't it just make more sense to futher improve your survival? This would potentially allow your group to bring an extra DPS in lieu of a 3rd healer and obtain an additional 15-20k dps rather than the paltry improvement in gearing your tank for extra damage.

The interrupt change is perfectly reasonable due to the importance of preventing raid wiping abilities in 10 man groups. A DPS should not have to babysit the boss to interrupt during vital add burn phases/target changes (Maloriak, Omnitron, Halfus, etc.). You may think this is simply a matter of opinion, but blizzard has discouraged tanks gearing for threat stats due to the vengeance mechanic, tank gear generally lacking hit, having both hit and stam as blue gems, and removing parry haste from bosses (among other changes). The only "encouragement" to gearing for hit/expertise that blizard has put forth is the holy power change, which is now being addressed with the new Grand Crusader holy power mechanic.
Syncognition
 
Posts: 44
Joined: Mon Jul 05, 2010 2:32 pm

Re: Hit Rating and Expertise

Postby Cascadian » Wed Mar 23, 2011 5:06 pm

Syncognition wrote:It surely has been said before, and I don't necessarily disagree with you in principle Cascadian, but rather than try to maximize the DPS of your Tank, wouldn't it just make more sense to futher improve your survival? This would potentially allow your group to bring an extra DPS in lieu of a 3rd healer and obtain an additional 15-20k dps rather than the paltry improvement in gearing your tank for extra damage.


I completely agree with that concept. But if survival is no longer an issue to the point where you are bring 2 healers to a 10 man raid, then the tank also gearing for hit/expertise is likely also a non-issue for survival.

For many of these bosses, where you stand and when you pop trinkets and damage reduction CDs is far more important than 5% more CTC coverage. Of course, people like me should do both. Learn where to stand AND gear for the worst case scenario of me screwing up.

Just so nobody misinterprets my position. I am massively gearing for avoidance and mitigation, because for the most part I am still learning these raid bosses on normal mode. We have maybe 6/12 on farm and we can get through 10/12. I don't consider any raids to be automatic.

I have even started switching to Seal of Insight after 30 seconds once vengeance has been established. To heck with DPS, I want to survive. On 25 man it takes the raid about an extra 1.5 seconds to cover the 6 minute loss of DPS from me changing seals from SoT to SoI.

I am just saying that there are scenarios when hit and expertise could be relevant (speed kills for achievments) when survival is not an issue.
Cascadian
 
Posts: 48
Joined: Tue Mar 15, 2011 12:44 pm

Re: Hit Rating and Expertise

Postby PsiVen » Thu Mar 24, 2011 5:08 am

Part of farm content involves your raid being highly trained at how much threat there is on X at time Y, at least in general terms such as "I can full burn here without worrying" or "I need to be careful here". Missing a bunch of attacks and causing serious problems by letting something loose is an issue which you will encounter on farm just as often as progression, and may be an area where you simply hoped it wouldn't happen. That's a good candidate for making a change on farm mode -- once you can afford to lose some survivability, add more hit to make sure the thing you hope doesn't happen, never happens. But most of the time, this just isn't an issue. Increasing your threat slightly isn't going to automatically cause your DPS to use that extra threat in a meaningful way.

I don't see much of a difference between 10/25 in terms of situations arising where you will want more hit/exp for threat purposes. The difference comes when you think about RDPS contribution from the tank. Say you have 5 DPS in a 10-man and 15 in a 25-man: your DPS contribution from the tank is three times as valuable towards meeting the enrage timer and other DPS requirements. Sometimes it is useful even on progression to squeeze out some extra DPS, if that's what you need, and the tank is far more able to contribute in a 10-man setting.
Gladiator Psiven, 90 Tankadin
90 Druid, 90 Mage, 85 Monk, 85 DK, 70 War, 70 Pal, 60 Priest, 60 Lock, 64 Rogue
Longtime addict of Space - Glory Through Conquest
User avatar
PsiVen
Moderator
 
Posts: 4363
Joined: Fri Jun 01, 2007 5:28 pm
Location: On a Boat

PreviousNext

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: No registered users and 1 guest


Remove Advertisements

Who is online

In total there is 1 user online :: 0 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: No registered users and 1 guest
?php } else { ?