Remove Advertisements

Hit Rating and Expertise

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Hit Rating and Expertise

Postby Dem » Thu Mar 10, 2011 1:55 am

You're making the assumption that every raid has a healer with an offspec dps or a spare healer and also probably that we're talking a 25 man situation as 3->2 healers in 10 is a much bigger deal.

None of these assumptions hold for everyone and equally general gearing guidelines do not apply to everyone.

Last time we had this debate, the counter argument was that the general readership of this site was not sufficiently experienced/intelligent to make their own decisions on this so we should only put up simplistic recommendations. I prefer to put a bit more faith in people and let them decide for themselves.

You haven't addressed the substantive points and we're not just talking about personal dps. There are situations for many raids where the tank threat caps the dps output. Not just at the start of a fight, but on new pickups or threat resets.
Dem
 
Posts: 137
Joined: Fri Jul 18, 2008 5:39 am

Re: Hit Rating and Expertise

Postby Aerron » Thu Mar 10, 2011 8:33 am

Dem wrote:You're making the assumption that every raid has a healer with an offspec dps or a spare healer and also probably that we're talking a 25 man situation as 3->2 healers in 10 is a much bigger deal.


Going to chime in here and say that going from 3 to 2 healers in a 10 man is not that big of a deal, if we're talking farm content. And if you're running with 3 healers, the chances are pretty good that at least one of them has a DPS offspec.

My guild has begun moving away from 3 healers to 2 on farm content. Our healing corps actually appreciates the change as it makes some of the boring fights fun for them again.

My viewpoint for the current discussion is, it will always be more beneficial to have another full DPS spec such as an Elem/Enh Shaman, Shadow Priest, Ret Pally etc etc, rather than to just have the Tank put on a few dps pieces. And if you do go with 2 healers, I'm sure they'd appreciate a tank more focused on survival than dps to make their job easier.
Aerron
 
Posts: 475
Joined: Sat Oct 13, 2007 6:33 pm
Location: Richmond, VA

Re: Hit Rating and Expertise

Postby theckhd » Thu Mar 10, 2011 8:36 am

Dem wrote:Take a fight where your chance of dying is zero. I assume you would have no issue with gearing all out for threat/dps?
Now take a fight with a 1 in million chance of dying. I'd suggest that most if not all reasonable people would say in this case gearing towards dps was sensible.


Except that this is never the case. There is always the chance that your healer will die, disconnect, or have to move out of fire. You will always have a reasonable chance of dying on "at-level" raid content, even the farm portions of it. We "farm" all of the bosses that we aren't already killing or attempting on hard mode. None of them are so trivial that I'd want to gear for less survivability. Ony/Nef/Constructs/Cho'gall/Al'akir are all perfectly capable of still killing me in certain situations. I am maximizing our chance of killing those bosses far more by gearing survivability in case those situations occur than I am by adding 1-2k DPS to the 320k+ DPS that the actual DPS players are putting out. Not to mention that survivability lets my healer relax a bit and do more than just heal me, which increases the chance another player doesn't die.

Dem wrote:You're making the assumption that every raid has a healer with an offspec dps or a spare healer and also probably that we're talking a 25 man situation as 3->2 healers in 10 is a much bigger deal.

There are still many 10-man fights where the tank's gearing can make the difference between 3 and 2 healers. 10-man Cho'gall is a very good example. Our alt run had much better luck 2-healing than 3-healing with a well-geared tank (me one week), but needed to 3-heal when we used an actual alt tank. Our alt tanks are just now getting to the point where we can comfortably 2-heal that fight.

That said, most of my insight about tanking does come from 25-man raids.

Dem wrote:You haven't addressed the substantive points and we're not just talking about personal dps. There are situations for many raids where the tank threat caps the dps output. Not just at the start of a fight, but on new pickups or threat resets.


That's because this isn't a substantive point. There is absolutely no fight in this tier of content where tank threat is not 100% irrelevant past the first 30s-1m, provided you have the proper raid support (tricks or MD). Absolutely none. And even without tricks, a tank who knows what they're doing has no trouble with threat within those 30 seconds. With proper planning, you have multiple attacks to use any time an add spawns, and an automatically successful taunt for the rare case where multiple things miss. The few adds that DPS do need to switch to immediately hit so weakly and die so quickly that you don't even need a tank for them.

If threat is a problem for you at any point in this tier of content, then you're doing something wrong. Other than hit-capping to cover interrupts (irrelevant after 4.1), there's literally no good argument for gearing towards threat in this environment. There are plenty of arguments, some of which I even agree with ("I don't like missing abilities, it's annoying"). But none of them are "good" arguments if your metric is "how do I best increase our chance of killing this boss."
"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: 7800
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby Dantriges » Thu Mar 10, 2011 9:01 am

Reading the replies, gearing for threat seems to involve tanking without pants, switching to all out DPS gear or in blue tank gear.

Seems to me that proper execution of tatics, not dying and actually knowing what to do by everyone will have a bigger impact. We had to gear replacements, and them not knowing the encounters had a much bigger impact than me switching to convenience gar.
Dantriges
 
Posts: 1252
Joined: Wed Jan 30, 2008 12:39 am

Re: Hit Rating and Expertise

Postby theckhd » Thu Mar 10, 2011 9:07 am

That's not particularly relevant though when discussing what one tank should be doing with his gear, since presumably changing gear doesn't suddenly make you forget encounter mechanics.

The difference between heavy survivability and "convenience" gear, as you described it, can be fairly small given the right gear choices. But that doesn't mean you should delude yourself into thinking it's anything but convenience. It's certainly not magically increasing your chances of success in the encounter.
"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: 7800
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby Dantriges » Thu Mar 10, 2011 9:57 am

Yeah, perhaps all these discussions lately, you get the impression that many people think tanks somehow turn into Mr Squishy, if they take some expertise/Hit.

The convenience is that I don´t have to use a sledgehammer to crack open the skull of some DPS to insert brain. :lol:
God forbid that you can´t go all out in the first five seconds or actually have to know that the boss wipes threat occassionally. If one rogue DPS complains about threat after Magmaw switches phases again, I petition Blizzard for a "over the knee and spank" emote.
Dantriges
 
Posts: 1252
Joined: Wed Jan 30, 2008 12:39 am

Re: Hit Rating and Expertise

Postby tlitp » Mon Mar 14, 2011 11:01 pm

theckhd wrote:The difference between heavy survivability and "convenience" gear, as you described it, can be fairly small given the right gear choices. But that doesn't mean you should delude yourself into thinking it's anything but convenience. It's certainly not magically increasing your chances of success in the encounter.

Disclaimer : I'd be willing to forget that I'm a Rogue misplaced amidst a bunch of sissy tanks. However, I'm constantly reminded the dire truth. As such, I'll specify from the very beginning that what people (generalizations, yeah) understand by "threat" is "damage" to me.

That said, if (e.g.) 100 more "damage" is merely a convenience, do explain how often 100 more HP (or 1% more block) is not another convenience (a "psychological buffer" of some sort). After all, we've had this discussion before.
Look, this seemingly endless debate is rather futile. While most people appear to be convinced that their standing is the only valid one, it'd be lovely to make use of the "dubito, ergo cogito" saying a bit more freely. :P
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Hit Rating and Expertise

Postby theckhd » Tue Mar 15, 2011 6:19 am

It's a matter of consequences. If I die during a progression encounter, it's a wipe (or at the very least, a waste of a battle res). If I do 100 less DPS, we still kill the boss. Sure, there have been wipes at 100k left on the boss, but in my experience they're much less common than wiping because of a tank death. And on many occasions, I've died in low-overkill situations; probably not by 100 health, but certainly by 1-2k, and often due to a collection of unblocked hits.

Quoting you for a moment:
tlitp wrote:Trading (e.g.) 1000 DPS for 1000 HPS is worth it if, and only if, it can be associated with an increase in the probability to successfully complete a given scenario (from "just defeat the fucker, however you do it" to achievements, to power runs). What is available, in order to assess it ? Neither the tools, neither the knowledge. The said probability, the "holy grail" of tanking, is all but unquantifiable. Which is fine, mighty fine - barring one side effect; it opens the door to the realm of "gut feelings".


The simple fact is that we do have tools with which to assess it. You can look at logs and make very clear determinations about many of the factors: Do we have enough RDPS to beat the enrage mechanic? How often are wipes caused by me dying, and by how much overkill? Do I require two healers or one? And so forth. Even without the benefit of logs, you can estimate a lot of these things in situ during an encounter. Are adds dying fast enough? Do I keep dying during a particular section of the fight (maybe when a healer has to move)? Is it taking more than one healer to keep me alive?

Can we quantify or analytically express the probability of success with either gear choice? Maybe not to great accuracy, but certainly well enough to determine which is the smarter choice for a given encounter. Note that it may depend on your specific circumstances as well - in my old guild, DPS was always a little low, but my survivability was fine. There were times that I geared/specced specifically for more DPS simply because the barrier to killing a boss wasn't my survivability, but whether we could hit certain enrage mechanics or not.

That has so far not been the case in my new guild, especially not in Cataclysm. If we're wiping, it's because a tank died or we screwed up an encounter mechanic so badly that half the raid died. So we try and minimize those chances - in the tank's case, by gearing as much survivability as possible and using cooldowns and WoG intelligently to counteract burst. This is the same thing that other reasonably-progressed tanks are telling me as well, so it's not just my guild.

It's good to doubt and critically analyze situations. But making up situations that don't exist in current raid content ("1 in a million chance of dying," "tank threat capping raid dps") isn't critical analysis. It's coming up with scenarios for the sake of justifying an argument. For there to be any useful discussion of hit/exp as viable progression stats, we'd need to come up with specific situations in current content where threat/dps-gearing gives you some significant, tangible advantage over survivability gearing. And so far, nobody has presented such a situation.
"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: 7800
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby tlitp » Tue Mar 15, 2011 8:35 am

What effects has a higher "damage" output ? Easy, no false dichotomies here :
a. decreasing execution time
b. dying the death of the Reckless (often wiping the group in the process)
Conversely, what effects has a higher "survivability" (primarily the health pool, but also CTC/armor) ?
a. decreasing the chance of "accidental" deaths (bursts/disconnects/suboptimal execution)
b. nothing, if none of the above actually happen

All in all, I'd add a few arguments to your rebuttal :
  • Tools to assess the "survivability" ? It very much depends on what we understand by assessment. Take your regular set for encounter A. Now remove exactly one Stam gem, being left with one empty slot. Are you (or anyone else) able to numerically assess how much higher is your chance of dying the "poor man's death" ?
  • Gradients. Do remember that tanks do not need to invest 10k itemization points in anything non-"survivability". Reaching 8% phys.hit and 26 exp skill (example chosen because "capping" is such a cool topic) doesn't imply the loss of 50k HP and 50% CTC.
  • Gut feelings. Do you remember the times when people like Mel said "screw it" and ran Festergut 25H with a tiny 35k unbuffed HP ? Granted, we're talking about "farm versus progression" here. Still, 35k was significantly lower than the generally recommended value (50k+). I think this illustrates fairly well that people do not actually gauge the "survivability" requirements of a certain encounter all that accurately. "OMG, OMG, I need more of this or that, else I'll die horribly"... sounds familiar ?
  • Absolute thresholds. Going from (e.g.) 205k HP to 195k when the encounter's mechanics indicate a maximum burst of 200k is a binary process (no chance of dying to bursts -> certain death if a burst does occur). Similarly, being able to reduce the number of healers from (e.g.) 3 to 2 is another example of an absolute threshold. The question is, "how often do these scenarios really occur ?".
  • Objective. You've explicitly stated : "For there to be any useful discussion of hit/exp as viable progression stats, we'd need to come up with specific situations in current content where threat/dps-gearing gives you some significant, tangible advantage over survivability gearing." I haven't actually argued that (e.g.) hit/exp are the better choice in encounter A or B. I've merely questioned the (generally accepted) consensus that "survivability" is universally better than any other alternative. It's a "can you prove it ?" instead of "you're wrong, dude".

The above "wall" can be basically reduced to one simple statement : we're not politicians/demagogues, disliking axiomatic truths should "come with the job". :P
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Hit Rating and Expertise

Postby Dem » Tue Mar 15, 2011 9:06 am

@Theck

My argument, to reiterate, is that a tank has a primary responsibility for survival and a secondary responsibility for threat. I have suggested that a tank should begin to move towards ‘threat’ stats from full survival gear when they feel it’s appropriate to do so, on an encounter basis.

This is an example of my wider point which is you should ‘Gear to your raid’s requirements’ rather than blindly following general recommendations.

You’ve made my argument for me by noting that there were times (in the past) when you judged it necessary to gear for more dps.

The arguments made against this are full of assumptions that actually argue for gearing to the raid e.g. (inferred) ‘assuming your raid can replace a healer by a dps” and “provided you have the proper raid support (tricks or MD)” i.e. include a rogue and/or hunter.

Dismissing the ‘1in a million chance of dying’ point is specious – the point is that every encounter will fall somewhere between 0% chance of dying and a 100% chance of dying. Many tanks will choose to modify gearing depending on their view of where the encounter falls along this line. You can argue that this ‘is never the case’ (which is incorrect if you accept infinite dimensions), but when a tank chooses to change out of full survival gear is a matter of judgement. You will never have enough facts and figures at your disposal to be able to generalise your approach to all situations.

When assertions are made such as "tank threat capping raid dps" “doesn’t exist in current raid content” this simply means that it’s not within your experience, not that it doesn’t happen. 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.

We can assume they weren’t ‘doing it wrong’ which is what you are implying in some of your comments.

If you’re looking for clear, tangible, specific evidence for these kind of choices then you’re probably out of luck. We could argue all day about how that loss of holy power from a miss killed me, that 30k wipe would have been prevented by extra tank dps or that dps died because of their lack of aggro control but it’s all about judgement.

Someone once said that ‘Intuition is the clear concept of the whole at once’ and Albert Einstein noted that ‘The only real valuable thing is intuition’.

Experienced tanks will continue to be guided by, but not limited by the excellent theorycrafting done at Maintankadin.
Dem
 
Posts: 137
Joined: Fri Jul 18, 2008 5:39 am

Re: Hit Rating and Expertise

Postby theckhd » Tue Mar 15, 2011 3:23 pm

tlitp wrote:
  • Gut feelings. Do you remember the times when people like Mel said "screw it" and ran Festergut 25H with a tiny 35k unbuffed HP ? Granted, we're talking about "farm versus progression" here. Still, 35k was significantly lower than the generally recommended value (50k+). I think this illustrates fairly well that people do not actually gauge the "survivability" requirements of a certain encounter all that accurately. "OMG, OMG, I need more of this or that, else I'll die horribly"... sounds familiar ?

I want to address this in particular. Mel's parse hunting is not a good indication of "gearing to your raid." It's an example of "I'm stunting because clearing ICC for the 50th time is incredibly boring." He specifically decreased his survivability to hunt for better parses. Nor is it a good exmaple of working below the "generally recommended value," because he did so with the support of a 30% health/healing buff and extra attention from his healers.

Again, there are plenty of reasons to do this (fun), but it's decidedly not increasing his chance of killing the boss. Which also, not coincidentally, makes it terrible advice to be giving to other players. Experienced/smart players will already be able to recognize when breaking from "conventional wisdom."

Dem wrote:My argument, to reiterate, is that a tank has a primary responsibility for survival and a secondary responsibility for threat. I have suggested that a tank should begin to move towards ‘threat’ stats from full survival gear when they feel it’s appropriate to do so, on an encounter basis.

This is an example of my wider point which is you should ‘Gear to your raid’s requirements’ rather than blindly following general recommendations.

You’ve made my argument for me by noting that there were times (in the past) when you judged it necessary to gear for more dps.


So to reiterate, my point is that the Vengeance mechanic and 3x threat multiplier have made threat trivial in every raiding situation you'll encounter this tier. So moving towards 'threat' stats is a non-issue for any tank with a minimum level of talent.

I have no problem agreeing that there's no universal rule that says "survivability is always better than threat in all situations." But in this content, the content that's currently relevant to progression, there really isn't any situation where threat is difficult, provided the player is playing their class well. A poorly-played paladin may certainly run into threat problems, but my suggestion to such a player is not going to be to stack hit/exp to compensate for their poor play. It's to practice and improve so that they don't need to gear for threat, just like the rest of us. To say otherwise is doing a disservice to the advisee, because all you're doing is encouraging them to play badly and compensate for it with gear.

Certainly, gearing to your raid's needs is the best course of action. But this is by most accounts the hardest tier of raiding since Sunwell, and tank death is still a very real possibility on every one of the current encounters. There are no "one in a million chance of tank death" encounters this tier, which is why considering such a scenario is not helpful.

If there are situations where an extra 2k-4k DPS (which is all you get from hit/exp, by the way) makes or breaks an encounter for you or your raid, then by all means consider it. But my suggestion is that for the vast majority of paladins out there, the ~900-1200 rating needed to cap hit and reach expertise soft-cap is going to prevent more wipes if it's converted to survivability stats wherever possible. Because that 1.2 million extra damage you do in a 5-minute encounter can be covered in 4 seconds by 15 DPS players pulling 20k each.

Perhaps more pertinently for a lot of players out there, when was the last time a tank got blamed for failing to hit an enrage timer? How much more likely is it that they'll be blamed for causing a wipe if they die? Remember that the advice we're giving here is what novice tanks are using to guide them. As I said earlier, an experienced tank should already have a decent idea of whether it makes sense to break with conventional wisdom for a certain encounter.

Dem wrote:When assertions are made such as "tank threat capping raid dps" “doesn’t exist in current raid content” this simply means that it’s not within your experience, not that it doesn’t happen. 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.

We can assume they weren’t ‘doing it wrong’ which is what you are implying in some of your comments.


Again, that was in a period before Vengeance. This time around, those same contributors are complaining that the threat game is incredibly boring because, as one of them put it, "we want the buttons we hit to matter." And in the current environment, threat is such a joke that they simply don't matter.

The only way a paladin is threat-capping their raid in the current tier of content is if they're doing something wrong. That's not an opinion, it's a simple fact based on simulations and analyzed parses. If you're performing the correct rotation, you're putting out nearly twice the threat that even the best-played DPS is capable of at this point.

I'm really not sure what point your quote about "intuition" serves. The type of intuition old Albert is talking about is not "gut feeling based on no concrete evidence." It's "what does the math behind this phenomenon really mean. And I should note that Einstein's intuition failed him pretty miserably on more than one occasion - one of the first things you learn about in an upper-level Quantum Mechanics course is a situation where Einstein was blatantly wrong because his intuition told him that the world shouldn't work the way Quantum Mechanics suggested it did.
"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: 7800
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Hit Rating and Expertise

Postby Firann » Tue Mar 15, 2011 7:21 pm

I think it still largely depends on your raid composition.

http://eu.battle.net/wow/en/character/f ... ann/simple

Thats my profile. By no means am i boasting my gear as its average (still no drops in our boss kills o.O) so i'm still wearing my crafted / purchasable gear.

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. The reason I am hit/exp capped is because i need all the interupts to land. Unfortunately I do not trust any of the dps in the 10mans to interupt except 1 shammy when she is dpsing.

So far whipes have not been caused by tank deaths, nor have healers complained about mana issues. We wipe because of people not following instructions :( (still only 5/12 /sigh)
Firann
 
Posts: 58
Joined: Thu Feb 12, 2009 8:41 am

Re: Hit Rating and Expertise

Postby tlitp » Tue Mar 15, 2011 11:25 pm

theckhd wrote:Experienced/smart players will already be able to recognize when breaking from "conventional wisdom."

You're not a random Paladin, chit-chatting with me or A or B or C; you're the Bringer of Numbers and Pounding Headaches, are you not ?
/teasing off

As a content author (guides a.s.o.), you have the responsibility to present things exactly as they are (i.e. dodging periphrases/ellipses). People like you should explicitly state something like this : "For this particular encounter you'll want _"survivability" stats_ that should put you comfortably outside of the danger zone; however, do keep in mind that the survival threshold is actually _value(s)_, therefore you shouldn't panic if you fail to meet the recommended stats".

PS. Parse hunting is, indeed, useful; at the most basic level, it's an effort to numerically determine (trial and error) the absolute "survivability" threshold of a specific encounter. Do not shrug it off with a loltankdps tag, please; it's more than simply "having fun".
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Hit Rating and Expertise

Postby baleogthefierce » Wed Mar 16, 2011 6:23 am

Firann wrote:I think it still largely depends on your raid composition.

http://eu.battle.net/wow/en/character/f ... ann/simple

Thats my profile. By no means am i boasting my gear as its average (still no drops in our boss kills o.O) so i'm still wearing my crafted / purchasable gear.

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. The reason I am hit/exp capped is because i need all the interupts to land. Unfortunately I do not trust any of the dps in the 10mans to interupt except 1 shammy when she is dpsing.

So far whipes have not been caused by tank deaths, nor have healers complained about mana issues. We wipe because of people not following instructions :( (still only 5/12 /sigh)


In my mind, a tank covering for DPS who play poorly isn't the same as a tank running a restrictive raid comp (read: 10-mans) where they are forced to interrupt (and that will be fixed in 4.1 anyway).

Fix your DPSers' performance and suddenly you will see how marginal threat stats are.
baleogthefierce
 
Posts: 291
Joined: Sun Jul 13, 2008 2:44 am
Location: Farming Molten Core for my Eye

Re: Hit Rating and Expertise

Postby theckhd » Wed Mar 16, 2011 6:30 am

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. The reason I am hit/exp capped is because i need all the interupts to land. Unfortunately I do not trust any of the dps in the 10mans to interupt except 1 shammy when she is dpsing.


This is one of the few cases where stacking hit does make sense. If you need to be a designated interrupter in a 10-man raid, which isn't that uncommon, then getting to 8% hit can be a reasonable course of action. Note that you don't need to be exp-capped for this though, as mobs can't dodge/parry while casting anyway.

Also note that the bosses don't hit all that hard on normal mode, so your gear is probably at an acceptable level for clearing those bosses. More survivability won't hurt, but it may not help enough to make a noticeable difference either.

To put some perspective on it, here's a death log from heroic 25-man Magmaw last week, I was tanking two constructs:
Code: Select all
[00:02:55.598] Blazing Bone Construct hits Theck 23147 (A: 1136, B: 16875)
[00:02:57.245] Magmaw Magma Spit Theck 40935
[00:02:58.014] Blazing Bone Construct hits Theck Dodge
[00:02:59.127] Blazing Bone Construct hits Theck 29016 (B: 20164)
[00:03:00.942] Blazing Bone Construct Fiery Slash Theck 81489 (R: 10060)
[00:03:01.185] Blazing Bone Construct hits Theck 27573 (B: 19161)
[00:03:02.816] Blazing Bone Construct hits Theck 3935 (O: 14793, B: 13015)
[00:03:03.539] Theck dies


I've stripped out all of the healing here to make it easier to see what sort of raw damage input heroic bosses and mobs are capable of. I didn't get any direct heals during this period, because the healers had to move out of lava spew, and I didn't have a cooldown up. The melees are about 50k after armor, though you can block around 20k of that. The Fiery Slashes hit for 80k after resistances, and can be blocked. Note that the overkill is only about 15k, so blocking the Fiery Slash or dodging any one of those hits instead of blocking would have let me survive one or two more seconds, which might have been enough to get an LoH, GS, or what not.

That said, I was using Seal of Truth here. So it's clear that even I believe there are trade-offs to be made between DPS and survivability. In this case, I have a lot of face time with the constructs while the DPS are busy burning the head, and killing them 4 or 5 seconds earlier is fairly helpful. But seals are something that can be quickly switched during an encounter to adapt to the appropriate situation, while gear is not.

Which I guess is part of the reason I question hit/exp as a gearing strategy. Sure, there might be some situations where the extra DPS is useful, but by gearing as such you're forcing yourself to have lower survivability in the situations where it's not. Whereas if you gear for max survivability, you can always switch seals and dump Holy Power into SotR instead of WoG to raise your DPS in the situations you want it. The difference in DPS between SoT/SotR and SoI/WoG is pretty huge (4-6k DPS), while the difference between 2% hit / 10 expertise and 8% hit / 26 expertise is only about 2k.

The on-the-fly versatility of switching seals and finishers puts the static gearing choices, which are both less versatile and less efficient, in a much different light. Assuming you don't change your gear that significantly from fight to fight, would you rather sacrifice survivability on the fights where it's helpful to raise your DPS ceiling by ~2k for the fights where the DPS is helpful? Or sacrifice 2k DPS from your DPS ceiling to have better survivability on the fights where you could use more survivability? In hard modes so far, my experience is that it's always been the latter. Rarely have I come across a fight where I didn't feel at least a little squishy.

tlitp wrote:As a content author (guides a.s.o.), you have the responsibility to present things exactly as they are (i.e. dodging periphrases/ellipses). People like you should explicitly state something like this : "For this particular encounter you'll want _"survivability" stats_ that should put you comfortably outside of the danger zone; however, do keep in mind that the survival threshold is actually _value(s)_, therefore you shouldn't panic if you fail to meet the recommended stats".

PS. Parse hunting is, indeed, useful; at the most basic level, it's an effort to numerically determine (trial and error) the absolute "survivability" threshold of a specific encounter. Do not shrug it off with a loltankdps tag, please; it's more than simply "having fun".


The problem with this approach is that it's very hard to quantify the "danger zone." It will depend on many factors, some of which are completely outside your control. The strength of your healers, DPS, and what tactics you employ on a given encounter all make the thresholds different for every raid composition. That's part of the reason I don't try and give out general advice for what stat levels you should aim for - the values are going to be arbitrary no matter how you come up with them.

Which is another reason parse hunting isn't a good representation of how to advise other players. You might be numerically determining the absolute survivability threshold of an encounter, but you're not doing so under conditions that are relevant to players attempting that content for the first time. 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. The numerical thresholds you come up with in a parse-hunting experiment are only relevant to other parse-hunters, and almost irrelevant to anyone trying to progress on a given boss.

Also keep in mind that I have to think carefully about the advice I give out, because I get parroted in a lot of places. I've gotten angry private messages and e-mails chiding me for suggesting people cap hit and expertise because of the MATLAB thread, despite the fact that I don't do so in any part of that thread. If I say, "go ahead and cap hit and expertise if you want to," it will get repeated in blogs and other forums, and a small subset of players will blindly follow it even when it's not in their best interest "because Theck said it's OK." So I try to be conservative with my advice. And gearing for survivability will very, very rarely be a bad choice in the current content environment.

I think the more responsible approach is to advocate a solid strategy for approaching bosses. Threat is incredibly easy at this stage of content, so the conservative approach is to start out with a full survivability kit every time you learn a new encounter. As you get more comfortable with the encounter, you may decide to change things up - maybe shift towards more DPS if the encounter rewards it, or stack hit for interrupts, or what not. But those decisions should depend on you and your raid group, not what some random dude on the internet suggests.
"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: 7800
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

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 { ?