+ Reply to Thread
Page 1 of 2 1 2 LastLast
Results 1 to 20 of 25

Thread: Can anyone dispute this...

  1. #1
    Join Date
    Jan 2010
    Posts
    5

    Can anyone dispute this...

    I'm looking for hard evidence, i.e. something directly stated by Blizzard that contradicts what I'm about to propose. Admittedly I haven't really cared to research the subject myself, I'm just going off of what people have told me versus what I have observed myself and read from wowtips on the loading screens.

    It seems to me that there's there's only a theoretical basis for this whole 102.whatever avoidance that everyone talks about when tanking, I've never seen a post from a Blizzard employee directly verifying it. On the contrary, I believe mitigation to be based on a diminishing returns system for each individual mob attacking the character, such as is stated on the tooltip when you mouse over a stat in the character menu. Diminishing returns being similar to crowd control effects, but not exactly the same (obviously the game wouldn't allow a person to expend a particular mitigation technique one hit after the other, and probably not multiple types one after the other either). I come to this conclusion based on two things I have observed while both playing a tank and a healer:
    1) I've noticed that, both when healing and tanking, a tank tends to take progressively more and more damage in waves. At the start of a fight, a properly geared tank will take damage slowly, but as the seconds tick by, they'll start to take more hits and require more attention from the healer(s). Then, after some length of time, the damage will seem to slow down, the healer(s) will have more leeway for a short time before the damage starts to stack up again, and it seems to cycle like that.
    2) I've also noticed, being a paladin tank, that as far as blocks go I tend to expend more charges of Holy Shield on multiple mobs versus a single one. For a single mob I tend to expend somewhere between two to five charges, depending on what is attacking me. When tanking two mobs, I tend to expend somewhere between five to all charges, and for more than two mobs I tend to expend all charges before the spell comes off cooldown.

    This all seems to fit a system of diminishing returns, which would mean that the effectiveness of the each stat is calculated individually, not as a total. It also explains why there only seems to be a slightly noticeable difference when I'm tanking one trash mob versus four in a raid; each mob is being affected individually by the diminishing returns of my mitigation stats and therefore the damage isn't becoming overwhelming.

    I had also theorized that crushing blows are based solely on character level, and in a raid scenario, gear level. This is simply because I don't buy into the whole avoidance theory, as well as the only mention of crushing blows by Blizzard that I've seen being the loading screen tip saying that mobs four levels above you can hit you with them.

    I'm just spitballing here, like I said this is based on what I've observed and not what I've read, if anyone'd care to dispute me please reply with a link to something directly stated by Blizzard to the contrary. Peace!

    PS: Thank you Lichslapped of HJPU on Chromaggus for your support of my idea and for encouraging me to put it out there, you're a champ!

  2. #2
    Join Date
    Jul 2009
    Posts
    580
    You just blew my mind. I'm not quite sure if you meant to say avoidance in a few places you put mitigation, but then again i'm no math guru.

    But to comment on one thing I understood in there: your damage intake as a paladin is going to be a bit diffirent than another type of tank.

    For instance, a warrior who pops shield block off the pull, like your paladin with holy shield, is going to start with little dmg incoming and build to higher damage as the effect drops (if it does). Conversely, a deathknight who did not drop two blood runes before the pull is going to miss 5% of his mitigaion while the first hit or two lands (takes at least 2 globals to dump the blood runes, but most dk's will start with plagues, so you end up with an average of 4 globals unless he pre-emptively boiled them off). A druid will take more damage on the first hit or two in a similar manner, unless his first hit is a crit (since he must crit to pop savage defense).

    That's all very basic stuff and I won't pretend to have studied models of tank damage based on class, but as a raider and raid leader who regularly plays three diffirent classes of tank, i'm pretty familiar with their mechanics.

    Per the crushing blow deal. What would lead you to question it? Are you suggesting that some mobs can crush a target that is not 4 levels below them, or simply implying that level isn't the sole factor in calculation?

  3. #3
    Join Date
    Jan 2010
    Posts
    5
    Ok, to start with I didn't explain my thoughts on the crushing blows too well. All I'm going off of is the wow loading screen tip stating that, paraphrasing, "A monster that is four levels above you or more can hit you for a crushing blow." All this says to me is that the basis for whether you catch a crushing blow is your level versus the mob attacking you, and in the case of raids, where the boss is a skull to everyone, the only method I could fathom of augmenting your level would be some system based on your gear level, as stats can be changed to accommodate what you're your facing.

    Also, as for the initial damage (and your right I've no experience raid tanking with another class beside a paladin) I'm popping holy shield and rarely anything else so that all my o.s. buttons are free for when something goes wrong and my healers have to pull off me. On the other hand, I've noticed when healing with my shaman that the damage pattern for a tank of any class tends to obey the trend that I stated above; they start off strong, peter off, and then go back to being strong again, and usually in a fairly regular rhythm. In fact I've found that identifying that rhythm for each individual tank that I'm healing gives me a crude basis for when I can pull off and toss heals elsewhere and when I should be focusing solely on the tank and letting others fend for themselves.

    Edit: Mind if I ask if I'm blowing your mind because this seems contrary to what you've observed or if because some part of it seems to ring true?

  4. #4
    Join Date
    Jun 2009
    Location
    Houston, TX USA
    Posts
    4,360
    I'm not sure what you're trying to suggest here. Yes, you will burn through more Holy Shield charges when more mobs are on you simply because you burn a charge whenever you block and there are more mobs attacking you and thus more chances to block.

    As for mitigation, I think you're suggesting from your post that as you are hit more, you receive less benefit from armor and other mitigation over time? If that's what you're suggesting, then it's simply not true. Look at any combat parse over time, and see how much damage is done to the tank over the course of the fight, and you'll see the damage intake is a function of the amount of mitigation the tank has, and not a function of the number of hits the tank takes.

  5. #5
    Join Date
    Oct 2009
    Posts
    19
    bit confused - Crushing blows do not exist in current raid content anymore... the 102.4% thing is combined avoidance/block/miss rate at which point you will no longer take a normal hit at all. Skull mobs are set to playerlevel +3, meaning you won't take a crushing blow from one I believe.

    I have no experienced anything like what you are talking about RE: damage coming in waves, except whne you're talking about absorbs being beaten down etc. (PW:S fading, the proc from Sacred Shield getting used up, AMS going down, whatever)...

  6. #6
    Join Date
    Jul 2009
    Posts
    580
    TBH the blowing my mind bit was because i'm unfamiliar with the concepts you were debating (mitigation calculations and possible variation on their returns based on mob count, etc). I didn't really have math to answer anything you said in the affirmative or negative.

    As far as regular patterns of damage, I can understand that. Though what you may percieve as a regular pattern for all classes is something i'm not familiar with. I make a point to monitor the damage patterns of our tanks (usually myself and a paladin, but we do have a prot warrior OT who is a regular fill-in, and a feral cat who has a substantial tanking set when needed aswell) and I don't see any more regularity than what the encounter mechanics dictate.

    For example: On festergut our paladin tanks first, rides to the pre-determined number of stacks and I taunt and repeat the process. During that time we both take an increasing amount of damage from festergut due to the way the encounter works. However, the paladin's damage intake ramps up far more cleanly than my own. He consistantly blocks off a few thousand damage per swing and thus the incline is not as steep. While my dk is tanking, I get spikes of damage, espcially when using my trinket or VB as a cooldown. The relative increase in damage taken over time is similar, but i'll more often be missed and take a huge hit, causing a mini-peak in damage taken.

    Our healers tend to group the paladin, and to a lesser extent the warrior in the "predictable" catergory of tank damage, while i'm out somewhere in spike territory. That isn't to say either of us makes a poor tank (nor am i suggesting we debate class balance), but our damage profiles have a lot more peaks than those that can regularly activate a blocking mechanism. My cooldowns have a similar effect, but as a blood dk only 1 of my cooldowns actually mitigates, so I don't really have the regularity of damage intake control that the paladin does (for example).

    I agree, that you can find a rhythm for each tank and on each encounter, but I'm not sure I would go so far as to say tank damage profiles are mirrior images. There are most definitely diffirent patterns in the larger picture of: boss hits, boss hits harder, boss soft-enrages, etc. for each tank and that is something that doesn't so much have to do with the theory behind mitigation as how each tank gets their mitigation, how each of those sources is calculated, and how often it is avalible.

  7. #7
    Join Date
    Jan 2010
    Posts
    5
    Well for starters, I'm not suggesting that armor obeys this diminishing return rule. The tooltip states that it simply reduces the amount of physical damage you take and nothing more, and common sense-wise I'm not inclined to think anything beyond that. What I've noticed is that, outside of special circumstances such as debuffs from raid bosses or spells cast on the tank like PW shield, once a tank starts taking damage they tend to take it slowly at first, and as the seconds tick on the rate of damage taken begins to increase up to a point (provided the tank doesn't die of course), where they go back to slowly taking damage. Looking at the overall damage a tank takes at the end of a fight tells you nothing about when and how all that damage was taken; say you were to graph that particular information. You'd have a straight line from zero seconds and zero damage taken right to whenever the fight ends with whatever value you ended with. This is of course not true. In reality, were you to somehow graph the real-time damage taken, you'd end up with a graph with steep rates of change and less steep rates of change, regardless of whether I'm right or not. What I'm suggesting is that you'd see a fairly consistent pattern with the rates of change; they'd increase and decrease / level out in regular intervals. Does putting it like that clarify what I'm saying a little better?

    Oh, and I'm certainly not suggesting that there's regular patterns for each class, I've noticed that the pattern is unique to each individual tank in each individual tanking scenario, i.e. heroic five man instances vs. ten man raids vs. heroic raids, etc. I'd argue that this further supports my theory since virtually each individual tank is going to have their own individual mitigation values.
    Last edited by VWD; 01-15-2010 at 10:13 PM.

  8. #8
    Join Date
    Nov 2009
    Location
    WI, USA
    Posts
    2,614
    Your logic is flawed simply because the numbers will never make sense when you consider them in a small window of time. You know those caster DPS that get 20,000 DPS on Hodir, if you look at the World of Logs parse for them, you might see it spike all the way up to something like 60,000 DPS. Well naturally if they are able to do 60,000 DPS, they should be able to do it all the time right, why did they only average out to 20,000? This is simply because you greatly reduce the accuracy of your measurement by reducing the sample size. Almost all damage dealing classes have a pattern in their DPS where it moves up and moves down based on the point in their rotation, buff changes, procs, and so on. The same rules apply to tanks. They could potentially dodge 10 attacks in a row. They can also potentially be hit by 10 attacks in a row. When you view a graph which measures the samples in small intervals like World of Logs you see patterns of it going up and down. However when you look at the actual numbers, over an extended period of time, they trend towards the actual mathematically calculated values as fight duration goes towards infinity.

  9. #9
    Join Date
    Jul 2007
    Location
    Canadia
    Posts
    3,523
    It seems to me that there's there's only a theoretical basis for this whole 102.whatever avoidance that everyone talks about when tanking
    Well, the only theoretical part about that is in extrapolating the measurable (through combat log data collection and then parsing said data) fact that there you gain an extra 0.2% avoidance for each level of an opponent is below you, to having a reduction of 0.2% for every level a mob is above you at the maximum level. That actually is backed up by more than a few people doing a lot more analysis of combat logs over the past 5 years to first verify it and then re-verify every once in a while to see if it's changed.

    I've never seen a post from a Blizzard employee directly verifying it
    You almost certainly never will. Only in very rare circumstances have Blizzard employees come right out and said "this is how this works". They have, on the other hand, stated that they find it enjoyable to watch the community efforts to reverse engineer the game mechanics, and often enough once the theorycrafting community has it right will discuss in terms of the correct mechanics -- but they still pretty much never come out and say "yep, you got it". It's like a metagame between the designers and the theory community.

    I believe mitigation to be based on a diminishing returns system for each individual mob attacking the character, such as is stated on the tooltip when you mouse over a stat in the character menu
    Assuming by "mitigation" you mean avoidance, the mechanics of the diminishing returns are well known at this point, again by a great deal of combat log analysis doing controlled experiments to pin down exactly what's going on. You can find the full details of Whitetooth's work here:
    Combat Ratings at level 80 - Page 6 - Elitist Jerks
    Combat Ratings at level 80 - Page 8 - Elitist Jerks
    Combat Ratings at level 80 - Page 8 - Elitist Jerks
    Combat Ratings at level 80 - Page 9 - Elitist Jerks
    Combat Ratings at level 80 - Page 9 - Elitist Jerks
    Combat Ratings at level 80 - Page 10 - Elitist Jerks
    Combat Ratings at level 80 - Page 14 - Elitist Jerks

    You can find my summary and guide to avoidance DR here: http://www.tankspot.com/forums/f63/4...avoidance.html. The explanation of what's going on with the poorly worded tooltip can be found here: http://www.tankspot.com/forums/f63/4...tml#post228997

    The additive nature of dodge+parry+miss (+block) is long since accepted as true by data collection and parsing; the combat table was one of the earliest pieces of theory to be completed (see http://www.tankspot.com/forums/f63/3...bat-table.html and http://www.tankspot.com/forums/f111/...-analysis.html). If you have evidence to the contrary, there are a lot of people who would love to see your collected data and analysis, since that'd be a fairly fundamental change in our understanding of the game mechanics!

    I've noticed that, both when healing and tanking, a tank tends to take progressively more and more damage in waves. At the start of a fight, a properly geared tank will take damage slowly, but as the seconds tick by, they'll start to take more hits and require more attention from the healer(s). Then, after some length of time, the damage will seem to slow down, the healer(s) will have more leeway for a short time before the damage starts to stack up again, and it seems to cycle like that.
    Again, I am certain that a lot of people would be very interested in seeing your collected data and analysis to show this trend, since it would have a pretty big impact in how guilds assign healers and healing roles over the duration of indivudual encounters.

    I've also noticed, being a paladin tank, that as far as blocks go I tend to expend more charges of Holy Shield on multiple mobs versus a single one.
    Well, yes, of course it works this way. How could it not?

    I had also theorized that crushing blows are based solely on character level, and in a raid scenario, gear level. This is simply because I don't buy into the whole avoidance theory, as well as the only mention of crushing blows by Blizzard that I've seen being the loading screen tip saying that mobs four levels above you can hit you with them.
    The mechanics of crushing blows have been well known for a long time now, in one of the very rare instances where a Blizzard employee came out and said "this is how this works". Even then, he wasn't telling the entire truth in that the chance for crushing blows is not actually based on a simple difference in level. The "+4 levels" thing is a convenient way to explain the mechanic without confusing the hell out of people. More on that here: http://www.tankspot.com/forums/f63/4...ing-blows.html. The loading screen tips are great and all, but also consider this one: "You can send mail to other players or even to your other characters from any mailbox in game." Except that an alliance character cannot use a horde mailbox and vice-versa. As with the crushing blows one, it conveys the basic idea just fine, but the explanation may not be complete -- it is a single line, after all. Somewhat related, note that the tooltip for defense was flat out wrong for more than a year at one point. The moral of the story? Data doesn't lie, but the UI does sometimes -- good to remember when working on the Next Big Thing

    Now before you go and fire off a reply, heated or not, about me bashing your ideas, please don't get the wrong idea. I'm not bashing you. A fair bit of what we consider "common knowledge" today came about because someone came and said "I think it's this way!" when we the common knowledge of the day was totally different. WoW gives us a great deal of power in the ability to record combat logs and write addons to collect data, and we have a long line of really smart people who have taken that data and analyse the hell out of it to figure out what's up. Sometimes "I think it's this way!" ends up turning what we thought we knew totally upside down, and then everyone gets excited to figure out exactly what that means. However, those big changes are always on the strength of collected data and good analysis. "I think it's this way!" is great and all, but you can't prove it with "I feel", you can't show it to anyone else with "I notice", and you can't convince anyone with "I believe". It'd be pretty exceptional to find that Blizzard went and changed something on the sly that we have to figure out from scratch. I'm pretty sure you're wrong, as will most people who read this thread, but we're also excited to see if you can prove yourself right!
    Got a question? Try here: Evil Empire Guides and here: Tankspot Guides and Articles Library first!

  10. #10
    Join Date
    Nov 2009
    Location
    WI, USA
    Posts
    2,614
    Oh, I did a Moonfire once with two piece tier nine bonus that did 28,000 damage. They should all do that right? Wait, I'm not going to always luck out and have every DoT tick crit? Damn you law of large numbers... as I trend towards casting an infinite number of Moonfires, my DoT trends to crit only as high as my crit chance is for Moonfire.

  11. #11
    Join Date
    Mar 2009
    Location
    melbourne.au
    Posts
    543
    Quote Originally Posted by Satrina View Post
    "I think it's this way!" is great and all, but you can't prove it with "I feel", you can't show it to anyone else with "I notice", and you can't convince anyone with "I believe".
    <3 Satrina

  12. #12
    Join Date
    Nov 2009
    Location
    WI, USA
    Posts
    2,614
    Here is a good way to look at how the Law of Large Numbers plays out in World of Warcraft. These numbers are made up, but are to illustrate the point.

    Have you ever walked up to a crab along a coastline somewhere and whacked it with a melee weapon? Did it hit you back for like 2 damage? Open up your character sheet... say it says you have 20% chance to dodge. Did you dodge 20% of that one attack? What, you can't partially dodge an attack? Clearly then you fail as a tank because your chance to dodge is no where near what your character sheet says.

    Now repeat the same test a million times. You dodged 202,678 times? That's 20.2678% dodge.

    That's the Law of Large Numbers... as the number of cases trends towards infinity, the result trends towards a predictable (and calculable) value.

  13. #13
    Join Date
    Feb 2009
    Location
    AZ (Zonie)
    Posts
    376
    Well, this explains why, in the course of a 5-man dungeon, my Shadow priest will do anywhere from 1600k to 4k, on the dps meter...and, if I have two different meters going, the numbers will be anywhere from 150-300 dps different, from each other, and I will hear choked laughter, from somewhere in my room, usually coming from me...

    It's definitive proof that I believe, notice, and feel, therefore, I am...

    And, they aren't Large numbers, at all...they're just metaphysically big-boned...

    /groan

    Ok, I'll stop being irrelevant now...or,

    now...

    ;>)
    -"Just like a buzzin' fly, I come into your life, I'll float away, like honey in the sun..."--Tim Buckley

  14. #14
    Join Date
    Jan 2010
    Posts
    5
    Satrina in the link you provided to your post Diminishing Returns - Avoidance I'm encountering problems understanding how you came to several things. For starters, could you inform me what addon you're using to come to such precise values for change of rating to skill as well as the values of 'k' and 'c'? Also, later in the post there are values that you multiply or divide by that seem to come out of nowhere, for example in the formula for calculating dodge gained you stated "1b) Convert defense skill to base dodge chance: 113 * 0.04 = 4.52%". What is the 0.04 and how did you come to that value? There's no mention of it before you use it there, probably on another post.
    Last edited by VWD; 01-16-2010 at 05:25 PM.

  15. #15
    Join Date
    Jun 2009
    Posts
    8
    This man has destroyed every bit of tanking knowledge we know today and has redefined tanking stats as we know it because he fought 4 mobs and his Holy Shield charges were eaten faster than when he was tanking 1.

  16. #16
    Join Date
    Jul 2007
    Location
    Canadia
    Posts
    3,523
    Well, the 0.04 is the basic conversion for defense skill. 1 point of defense skill raises each of miss, dodge, parry, and block by 0.04% (see http://www.tankspot.com/forums/f63/41796-defense.html)

    I think that at this point, you'd be best served by doing some reading and understanding to get a handle on the basic elements of tanking theorycraft:
    - the articles in Evil Empire Guides - TankSpot
    - http://www.tankspot.com/forums/f200/...questions.html
    - http://www.tankspot.com/forums/f200/...nce-guide.html (dated but relevant)
    - There's also a lot of good stuff to read over at Maintankadin

    Once you're done there, you'll want to read the entirety of Combat Ratings at Level 80 to see Whitetooth's derivations of k and c.
    Got a question? Try here: Evil Empire Guides and here: Tankspot Guides and Articles Library first!

  17. #17
    Join Date
    Sep 2009
    Location
    Fort Worth, Texas
    Posts
    92
    Quote Originally Posted by VWD View Post
    What I've noticed is that, outside of special circumstances such as debuffs from raid bosses or spells cast on the tank like PW shield, once a tank starts taking damage they tend to take it slowly at first, and as the seconds tick on the rate of damage taken begins to increase up to a point (provided the tank doesn't die of course), where they go back to slowly taking damage. Looking at the overall damage a tank takes at the end of a fight tells you nothing about when and how all that damage was taken; say you were to graph that particular information. You'd have a straight line from zero seconds and zero damage taken right to whenever the fight ends with whatever value you ended with. This is of course not true. In reality, were you to somehow graph the real-time damage taken, you'd end up with a graph with steep rates of change and less steep rates of change, regardless of whether I'm right or not. What I'm suggesting is that you'd see a fairly consistent pattern with the rates of change; they'd increase and decrease / level out in regular intervals. Does putting it like that clarify what I'm saying a little better?
    You've described the timing of mob special abilities.

  18. #18
    Join Date
    Jan 2010
    Posts
    5
    Ok, thanks for clarifying the 0.04 value for me Satrina. While I do appreciate you giving me those further links which I intend to read in time, would you mind just explaining to me how such precise numbers used in your formulas are obtained? I'm sure the information is in there somewhere but frankly I don't have the time right now to read through it all.

  19. #19
    Join Date
    Nov 2008
    Posts
    228
    I think I can summarize the OP's statements in one sentence. Seems like a trolling attempt to me, but I tend to see the worst in people ^^

    I done looked for a long time and can't find that gravity thing them smart people talk about, that there must be a lie.

  20. #20
    Join Date
    Jul 2009
    Posts
    580
    Well, to be fair, regardless of whether or not it's a troll (and i personally doubt it, despite the fact i'm a cynical arse) there is certainly no harm in questioning what is often assumed to be true.

    Science depends on people who continue to question everything we think we know.

    I don't think anything was disproven here, but the fact he questioned something and tried to gain clarity on the situation isn't something to lampoon.

+ Reply to Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts