Remove Advertisements

Armor vs HP Math Help

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Armor vs HP Math Help

Postby Eminai » Tue Jan 26, 2010 8:37 pm

lralexl wrote:None of that included unmitigateable damage such as magic or bleeds..... taking 3 melee hits in a row is one thing, taking one melee, one bleed, and one magic damage is another. This is why stam is usually valued higher to tanks(such as me), it is universal.


Magic or bleed damage is just a smokescreen. Even if we assume that 50% of a tank's damage is magic based, which we could do with my formula by increase tank 2's damage per swing by 500, he still takes 500 less damage every 1.5 seconds, which means that the healing needed is increased by 5000 over the course of 15 seconds, which still leaves him needing 2811 less healing than tank 1. Assuming 15 seconds is the longest you would go without avoiding a hit, you would need for 78% of the spike damage taken to be unaffected by armor for Juggernaut's vitality to be better than the armor trinket, ignoring the use effects.
theckhd wrote:Philosophically speaking, wiping your raid is slightly less serious than handing a baby off to the wrong set of parents.
Eminai
 
Posts: 13
Joined: Tue Jan 26, 2010 5:53 pm

Re: Armor vs HP Math Help

Postby lralexl » Tue Jan 26, 2010 8:40 pm

Eminai wrote:
lralexl wrote:None of that included unmitigateable damage such as magic or bleeds..... taking 3 melee hits in a row is one thing, taking one melee, one bleed, and one magic damage is another. This is why stam is usually valued higher to tanks(such as me), it is universal.


Magic or bleed damage is just a smokescreen. Even if we assume that 50% of a tank's damage is magic based, which we could do with my formula by increase tank 2's damage per swing by 500, he still takes 500 less damage every 1.5 seconds, which means that the healing needed is increased by 5000 over the course of 15 seconds, which still leaves him needing 2811 less healing than tank 1. Assuming 15 seconds is the longest you would go without avoiding a hit, you would need for 78% of the spike damage taken to be unaffected by armor for Juggernaut's vitality to be better than the armor trinket, ignoring the use effects.


This is why EH is situational. No one is arguing armor's usefulness, we all have some bonus armor pieces. We are arguing the first remarks made that armor is the only stat needed hands down. I think this thread is taking a turn for the better.
lralexl
 
Posts: 9
Joined: Sun Jan 24, 2010 11:43 pm

Re: Armor vs HP Math Help

Postby Njall » Tue Jan 26, 2010 8:43 pm

Wrathy wrote:You have provided us with a Theory:
Armor is superior to Stamina in all situations


Its more properly a hypothesis. He needs to test for a null-case.

Sorry, being a bit pendantic but I'm busy with my thesis. :)
Image
Flight to Peru: £1000
Camping gear: £200
Native guide: £50
Sledgehammer to break down stone door: £12.99
Awakening one of the Great Old Ones: priceless.

There are some things man was not meant to know. For everything else,
there's Mastercard.
User avatar
Njall
Maintankadonor
 
Posts: 4328
Joined: Mon Oct 05, 2009 1:13 pm
Location: Thank heavens! Heavy Metal!

Re: Armor vs HP Math Help

Postby Kihra » Tue Jan 26, 2010 8:48 pm

I am not really sure what we're even arguing about at this point.

If a tank is not sufficiently geared to survive a worst-case burst scenario, then it's easy to mathematically determine the relationship between armor and stamina for the particular type of burst damage you're taking. Even in a best case scenario where that damage is 100% physical and can be mitigated by armor, the Corroded Skeleton Key will help you more than any of the bonus armor trinkets.

If a tank is sufficiently geared to survive a worst-case burst scenario, then stack what you want. Theck's TEH formulas and the concept of EH are about gearing when you're under the burst threshold. You don't have to keep stacking max EH gear once you're over the worst-case burst threshold. Stacking armor is fine if you want to help reduce HPS required. Adding more threat is fine if you want to hold aggro more easily. Stacking avoidance is fine if you want to reduce your damage intake that way.

What approach you take once you're over the worst-case burst scenario is up to you, and I don't think anyone disagrees that continuing to stack armor in this case is an excellent strategy.

If you're trying to say that armor trinkets beat the skeleton key when you aren't over the burst threshold, well, then you're just mathematically wrong. It sounds like that's not what you're saying though, and this argument is just being caused by a disconnect where you think people are blindly applying EH rules when over burst thresholds. We aren't.
Kihra
 
Posts: 554
Joined: Mon Dec 31, 2007 12:01 pm

Re: Armor vs HP Math Help

Postby Koatanga » Tue Jan 26, 2010 8:55 pm

Eminai wrote:Gladly. Using http://www.wowarmory.com/character-sheet.xml?r=Mal%27Ganis&cn=Grim as the tank in question. For the purposes of this assumption, we shall assume his stats are exactly as shown in the armory, on a boss that hits for 25k on average, at his armor level, with a 1.5 second swing timer. Given those assumptions, he, at his gear level will require a heal of at least 4436 within 3 seconds of engaging the boss, and will then require 25k healing every 1.5 seconds, if we assume a worst case scenario.

If he had the armor trinket, instead of having 45564 health and 29613 armor he would have 43375 health and 31584 armor, which would reduce festergut's swings to roughly 24000 per swing, which would change the healing needed to 4625 within 3 seconds, followed by 24000 healing per 1.5 seconds. The benefits of this depend on how we view our worst case scenario. If we assume that a tank is going to have a streak of up to 15 seconds where he does not avoid or block a hit, using the armor trinket will reduce his total healing needed for one of those streaks by 7811 over 15 seconds, or 520 healing per second. In order for stamina to be more valuable than armor, we need to assume that the tank will never get hit more than three times in a row, otherwise he will actually require less healing by getting the armor trinket instead of regular juggernaut's vitality. Now, this is an unrealistic scenario that will never occur in a raid, because actually predicting everything that could happen to gib a tank in a raid would require more effort than I'm willing to put towards something I'm not getting paid for.

EDIT: Because this is the post with all the math in it, I'll address theck's "worst case scenario" using the example above. Assuming a 5 minute fight with the above damaging attacks, the tank with the armor trinket will require 659 less healing per second than the tank using the stamina trinket, for a total of nearly 200,000 less healing, assuming a true worst case scenario of a tank that never avoids a hit or blocks one.

Edited again to fix the URL


Even that difference will not be sufficient for healers to swap to a lower-healing more mana-efficient healing spell. They will still see the tank taking 24000-damage hits, and still be spamming away with their large heals the entire fight. The difference will be expressed as overhealing.
Retired. Koatanga, Shapely, Sultry, Doominatrix of Greenstone - Dath'Remar
Koatanga
 
Posts: 1971
Joined: Mon Nov 17, 2008 12:46 pm

Re: Armor vs HP Math Help

Postby Eminai » Tue Jan 26, 2010 8:55 pm

theckhd wrote:1) Do you even know the definition of EH? Because stamina is a part of it, as is armor.
2) If you think that EH is a worthless stat, you're a worthless tank. Seriously. It is certainly not the end-all, be-all, one stat one must try and maximize. But then again, nobody ever said it was, including me.
3) There are several prominent theorycrafters who might disagree with your assessment of EH as worthless. Perhaps you sholud brush up on one of them.


1)Yes, I know the definition of EH. You're the one who seems to be confusing it for stamina.
2)Nice ad hominem there. Glad we could keep things civil.
3)I don't really care what they think. However, I'm glad you could follow your ad hominem fallacy with an appeal to authority fallacy

theckhd wrote:Are you seriously this stupid? He asked the question, presumably to make the point that Armor is somehow "better" for this scenario. But he's wrong for a few reasons reasons.
1) Any time you are asking whether a tank can survive a burst scenario, you are asking "Does this tank have enough EH to survive this burst." I'm pointing out that we both have almost exactly the same EH (in fact, his is ahead by less than 0.5%), but I do it with mostly stamina right now, while he does it with armor. However, on any fight with a magical or bleed burst, I come out far ahead. And when I finally do manage to get the badges for the chest, gloves, and cloak, I'll be even farther ahead.
2) More pertinently, he would have higher Effective health with the stamina trinket than the armor trinket as well.
3) Is that irrelevant in current content? Probably. We're both above the EH threshold for the most demanding fight, which is unquestionably Festergut. But it will be very relevant once hard modes come out if there are any serious tank checks involved.


I wasn't asking why you feel EH is important, although you do make a good use of the straw man fallacy here. I was asking why you felt it was necessary to say you could survive Festergut immediately before saying that it didn't matter whether or not you could survive Festergut.

theckhd wrote:Wait, what? What does this even mean? My formula doesn't even describe a particular worst case, it's valid for all worst cases, including the ones from completely physical damage.


The first part means that saying "stamina is the best stat" is not almost as valid as saying "armor is the best stat" just because stamina is almost as good as armor.

theckhd wrote:I would suggest that without data, your conjecture is just that - conjecture. I understand what you're saying. My posts should be enough to prove that I'm a pretty smart guy. It's not comprehension that's lacking here. The simple fact is that you're incorrect.

In real situations, where real raid tanks die, they do not do so because they didn't maximize the healing done to them. They die because they take enough damage in a short period of time without heals to kill them. This is well established tanking theory here, at tankspot, at Elitist Jerks, and basically everywhere else.

If you're going to come in here and suggest that somehow, all of those people are completely wrong, and without providing anything to justify it other than "I said so," then you're wasting your time. Because I, and most of the other experienced tanks on this forum can see right through you to see that you don't know what you're talking about.


Actually, in most real raid situations, the reason a tank dies is because he took a large hit, did not quite get healed to full health, then took another large hit, not because he goes "5 seconds without heals" or whatever your worst case scenario is. Yes, with holy paladins being able to hit a tank for absurdly large heals during damage spikes, the two things are almost the same. But in real raid situations, almost is the difference between getting world first Tribute to Insanity and wiping on NRB for 3 hours. I will grant you that I did not post proof of my assertions, and I have since rectified the situation.



Kihra wrote:I am not really sure what we're even arguing about at this point.

If a tank is not sufficiently geared to survive a worst-case burst scenario, then it's easy to mathematically determine the relationship between armor and stamina for the particular type of burst damage you're taking. Even in a best case scenario where that damage is 100% physical and can be mitigated by armor, the Corroded Skeleton Key will help you more than any of the bonus armor trinkets.

If a tank is sufficiently geared to survive a worst-case burst scenario, then stack what you want. Theck's TEH formulas and the concept of EH are about gearing when you're under the burst threshold. You don't have to keep stacking max EH gear once you're over the worst-case burst threshold. Stacking armor is fine if you want to help reduce HPS required. Adding more threat is fine if you want to hold aggro more easily. Stacking avoidance is fine if you want to reduce your damage intake that way.

What approach you take once you're over the worst-case burst scenario is up to you, and I don't think anyone disagrees that continuing to stack armor in this case is an excellent strategy.

If you're trying to say that armor trinkets beat the skeleton key when you aren't over the burst threshold, well, then you're just mathematically wrong. It sounds like that's not what you're saying though, and this argument is just being caused by a disconnect where you think people are blindly applying EH rules when over burst thresholds. We aren't.


Basically, the cause of this argument is that the burst threshold doesn't actually exist. No encounter in current content is designed around a tank taking their entire health pool in such a short period of time it is impossible for a heal to land on them, other than in situations where cooldowns will be available to mitigate the damage every time.


EDIT:

Koatanga wrote:Even that difference will not be sufficient for healers to swap to a lower-healing more mana-efficient healing spell. They will still see the tank taking 24000-damage hits, and still be spamming away with their large heals the entire fight. The difference will be expressed as overhealing.


You're assuming that during the "worst case scenario" a tank is actually getting healed for more health than he is losing. If that is the case, then the fight isn't designed to stress the tank's survivability, and he'd be fine stacking strength, or agility, or, realistically, whatever he feels like.
theckhd wrote:Philosophically speaking, wiping your raid is slightly less serious than handing a baby off to the wrong set of parents.
Eminai
 
Posts: 13
Joined: Tue Jan 26, 2010 5:53 pm

Re: Armor vs HP Math Help

Postby Kihra » Tue Jan 26, 2010 9:12 pm

Eminai wrote:Basically, the cause of this argument is that the burst threshold doesn't actually exist. No encounter in current content is designed around a tank taking their entire health pool in such a short period of time it is impossible for a heal to land on them, other than in situations where cooldowns will be available to mitigate the damage every time.


I agree that in ICC normal we haven't seen any significant burst yet. The developers have been especially generous about not stunning us in ICC either, so that we can continue to avoid the melee attacks.

Isn't it better to look at TOGC though? Some very real burst scenarios did exist in TOGC. Goremok, Icehowl and Anub all had interesting burst combos where you could take a triple hit from melee and special abilities in under a second.

Is it true that the burst threshold wasn't particularly high in TOGC? Yes. However, we don't know what the worst-case burst threshold will be for the ICC fights, so EH tanks like myself, Meloree, Theck, etc. tend to be gearing right now as though we are under that burst threshold. If we end up being over the threshold for a fight, then fine, we'll stack more threat, more armor, more avoidance, wear more skirts when tanking, etc.
Kihra
 
Posts: 554
Joined: Mon Dec 31, 2007 12:01 pm

Re: Armor vs HP Math Help

Postby Rhiannon » Tue Jan 26, 2010 9:19 pm

I'm curious how often in wrath tank death occurs due to the heals that landed not being big enough, rather than them not landing fast enough. Eminai and Jamesvz seem to be advocating that tank survival is balanced on a knife edge where healer throughput is struggling to match the average incoming damage profile, when from my experience every few seconds you're topped off to full health.

If longterm healing throughput didn't vastly overshadow the longterm damage intake, then yes, reducing the damage taken by as much as possible would be highly desireable, but it's not the case that tank deaths, at least in 25 man content, are a long, drawn-out battle where healers slowly but surely loose the struggle to outheal the damage coming in. This averaging game only makes sense when the total HP pool is massively larger than the incoming damage and heals profile - the closest analogue I can think of in WoW is the obviously trivial example of Loatheb, where you have no chance of getting spiked down, and what matters is taking little enough damage that the few seconds of healing throughput you get every cycle can outpace the damage you took in that cycle.

Edit: Reading Eminai's last posts this is precisely where the disconnect lies. He is of the opinion that the majority of tank deaths are due to a whittling down process with lots of back and forth, with the healers getting throughput into the tank but it being insufficient to top the tank off. If this is indeed the case, then armor does perform very strongly - from my own recollection and perception of personal experience I've never felt it was though, but that has the potential to be fairly flawed/biased.

Armor is wonderful and all, but let's not kid ourselves - it's not some sort of panacea that will transform us from requiring constant attention to allowing the healers to go read a book like the post quoted by Jamesvz tried to suggest.

DR% = Armor / (Armor + 400 + 85 * (AttackerLevel + 4.5 * (AttackerLevel - 59)))

Vs level 83 mobs we can just say:

DR% = Armor / (Armor + 16,635)

With 40,000 armor, DR% = 70.63%

With 38,000 armor, DR% = 69.55%

So if a physical attack would hit for 100,000 post-reduction from talents, stances, abilities, glyphs, buffs:

on 40k armor it would hit for 29,370

on 38k armor it would hit for 30,450

If a healer has to give constant attention to a tank taking 30.5k hits, they need to give just as much attention to a tank taking 29.4k hits. Your healers will barely notice the difference between these two setups in all honesty, and certainly shouldn't change their behaviour.

Now, glyph of indomitability doesn't even give 2,000 armor, but let's say it does. The skeleton key gives 228 stamina, which after multipliers ends up at around 2860 hp. We're being a little generous with the glyph, so let's round this up and call it 3k.


So let's say tank A has the glyph, with 40k armor and 57k raid buffed health. After this 100k pre-armor hit, he has 27.63k health. Without additional heals, the biggest pre-armor hit he can survive now is 94.40k.


Tank B has the skeleton key, with 38k armor and 60k raid buffed health. After this 100k pre-armor hit, he has 29.55k health. Without additional heals, the biggest pre-armor hit he can survive now is 97.04k.


So in this healing vacuum, the next raw hit has to be 2.80% bigger to kill tank B than to kill tank A.

Now, this is obviously the worst case metric that EH was designed to try to model, and in real raiding it's unlikely to happen that you receive absolutely nothing in the way of healing (though it can and does happen).



If one tries to refine or expand the model, different factors push the balance in favour of armor or stamina. It should go without saying that if, rather than 100k armor-mitigatable damage, the incoming damage is of more than one sort, this will push one towards the stamina trinket, but to illustrate:

Each tank has a bleed that ticks for 10k damage, post-mitigation from abilities, talents and so on. A 90k (pre-armor, post other mitigation) physical attack lands just after the bleed has ticked.


Tank A (40k AC, 57k HP): bleed takes him to 47k HP, physical hits for 26,343, leaving him at 20,657 hp. Without additional heals, the biggest physical pre-armor hit he can survive now is 70,574.


Tank B (38k AC, 60k HP): bleed takes him to 50k HP, physical hits for 27,405, leaving him at 22,595 hp. Without additional heals, the biggest physical pre-armor hit he can survive now is 74,204.


So in this healing vacuum, with a mixture of damage types that can and cannot be mitigated by armor, the next armor-mitigatable hit has to be 5.14% bigger to kill tank B than to kill tank A.

So, obviously, adding some source of damage that armor doesn't mitigate into the burst tips the scales in favour of more stamina.



On the other hand, what factors play in favour of mitigation over buffer? Well, healing and absorption affects are the obvious ones. The problem is that we're trying to address the worst case scenario here, where healing is, for whatever reason, stretched very thin. You can only receive one Power Word: Shield every 15 seconds at maximum, Divine Aegis gets chewed through very fast and is dependent on the disc priest critting, Sacred Shield has an internal cooldown, and with Icecrown Radiance not even paladins can guarantee that every hit will have to pass through our other form of mitigation, block value, in a typical gearset. Look at a log of a typical reasonably hard hitting boss, even with a disc priest and Val'Anyr-wielding holy paladin healing the tank, and you will see series of hits where for some combination of reasons nothing gets absorbed at all.

So yeah, absorption shields and block value make armor more effective, but neither of them are 100% reliable, and they are guaranteed to not be there from time to time.

It is probably reasonable to assume some level of passive healing at all times, from hots, earthshield (though this has an ICD), etc - one could argue that if your raid isn't able to maintain this passive throughput it's probably going to fail anyway regardless of 2k armor here or there, but what value we pick will strongly determine how strong armor appears to be. Too high a value, and risk of tank death becomes minimal anyway. Let's say after the hit we're interested in, 6k worth of heals land before the next hit. That's about consistent with a rejuv, and an earth shield charge.


For whatever reason, the main healers are unable to do their jobs for a gcd or two (e.g. one has to move from fire while other is refreshing a non-healing spell). Going back to purely physical damage intake:


Tank A (armor trinket), with 40k armor and 57k raid buffed health. After this 100k pre-armor hit, he has 27.63k health. Before the next hit lands, he regains 6k health, leaving him on 33.63k health. Without additional heals, the biggest pre-armor hit he can survive now is 114.90k health.


Tank B (stamina trinket), with 38k armor and 60k raid buffed health. After this 100k pre-armor hit, he has 29.55k health.Before the next hit lands, he regains 6k health, leaving him on 35.55k health. Without additional heals, the biggest pre-armor hit he can survive now is 116.75k health.


So, if we include this small amount of healing into the model, the next raw hit has to be 1.16% bigger to kill tank B than to kill tank A.

Now obviously, the more we scale up the healing between hits, the better the armor trinket becomes. However, situations where large amounts of healing are landing between hits are not usually the situations where tank survival is a problem or an interesting issue to worry about. Edit: at least, that's my perception, but that seems to be the main thing that Eminai is proposing.



So you have two factors working in opposite directions - one being that often damage intake isn't all mitigatable by armor, and the other being that true EH healing vacuum scenarios are rare, and if they do occur you will, apart from in an impossibly narrow band of damage ranges, live or die with either setup. In more realistic situations, there is almost always some amount of healing throughput going on, and every point of healing that lands but does not lead to both tanks being topped off will favour the armor trinket more. We're all about gearing for the worst case scenario as tanks though, when your PW:S has been eaten, sacred shield and earth shield are on ICD, paladin's moving out of fire with holy shock on cd and priest's just been silenced, and all you've got healing you for a a couple of seconds are a couple of druid hots and a flash of light hot. And for lasting a couple of seconds with minimal throughput, having a bigger buffer is more likely to be beneficial.
Last edited by Rhiannon on Fri Mar 12, 2010 10:55 pm, edited 5 times in total.
Rhiannon
 
Posts: 1061
Joined: Mon Mar 03, 2008 8:17 am

Re: Armor vs HP Math Help

Postby Fridmarr » Tue Jan 26, 2010 9:26 pm

That's just saying if you take less damage you'll need less healing. That's a given, but it doesn't necessarily correspond to survivability. Higher EH means the Boss needs to do more DPS in order to kill you, it increases your TTL. If two tanks had the same EH, then the one with the higher armor would have an advantage. That is why armor is weighted beyond the EH that it gives, but in the case of the trinket, with all else equal the one with the skeleton key, has more EH.

Also, I have a hard time with the concept that non physical damage is a smokescreen when it can exist in very large hit sizes, meaning during any given death scenario it could be a significant contributor to the damage. It's counter intuitive to suggest mitigation is the end all be all, at the same time as saying damage that isn't mitigated shouldn't be counted.
Fridmarr
Global Mod
 
Posts: 9665
Joined: Sun Apr 08, 2007 1:03 am

Re: Armor vs HP Math Help

Postby Eminai » Tue Jan 26, 2010 9:40 pm

RE:Rhiannon

Basically, you've hit the nail on the head, and all we disagree on is the amount of healing your healers can give you between hits. The only issue I would take with your post is that you're assuming that the scenarios that contribute to tank death are more often going to be where no healers can heal the tank and he only goes 2-3 hits under this situation and then is immediately healed to full, where I'm assuming that the tank is actually having a long streak of bad avoidance, but is consistently getting healed, just not for enough to top him off.

RE:Fridmarr

The reason I'm saying it's a smokescreen is from the math I posted in the same post. Yes, magic or bleed damage can contribute to tank death, but the sizes that it would have to hit simply don't exist in current content. It's also worth noting that in the examples both myself and Rhiannon used, both tanks had the exact same time to live with no healing received, despite that being the scenario that EH is supposed to be the ideal solution to.
theckhd wrote:Philosophically speaking, wiping your raid is slightly less serious than handing a baby off to the wrong set of parents.
Eminai
 
Posts: 13
Joined: Tue Jan 26, 2010 5:53 pm

Re: Armor vs HP Math Help

Postby Fridmarr » Tue Jan 26, 2010 9:48 pm

Eminai wrote:RE:Rhiannon

Basically, you've hit the nail on the head, and all we disagree on is the amount of healing your healers can give you between hits. The only issue I would take with your post is that you're assuming that the scenarios that contribute to tank death are more often going to be where no healers can heal the tank and he only goes 2-3 hits under this situation and then is immediately healed to full, where I'm assuming that the tank is actually having a long streak of bad avoidance, but is consistently getting healed, just not for enough to top him off.

RE:Fridmarr

The reason I'm saying it's a smokescreen is from the math I posted in the same post. Yes, magic or bleed damage can contribute to tank death, but the sizes that it would have to hit simply don't exist in current content. It's also worth noting that in the examples both myself and Rhiannon used, both tanks had the exact same time to live with no healing received, despite that being the scenario that EH is supposed to be the ideal solution to.
Yeah but there's maybe one encounter in ICC that's a threat to the tank at all. In ToGC though there was non physical damage that could be very dangerous, some of which was illustrated above.

Yes in your scenario TTL was probably equal. However, more EH can increase TTL depending on the hit sizes relative to the EH pool, but the opposite can never happen. Generally, the rule of thumb is acquire the EH to live through a given worst case generally constrained by a reasonable time frame depending on the damage components, and once you can reach that minimum, then start looking at reducing damage. No one really cares about that now because the content isn't demanding it.
Fridmarr
Global Mod
 
Posts: 9665
Joined: Sun Apr 08, 2007 1:03 am

Re: Armor vs HP Math Help

Postby Rhiannon » Tue Jan 26, 2010 9:49 pm

Fridmarr wrote:That's just saying if you take less damage you'll need less healing. That's a given, but it doesn't necessarily correspond to survivability. Higher EH means the Boss needs to do more DPS in order to kill you, it increases your TTL. If two tanks had the same EH, then the one with the higher armor would have an advantage. That is why armor is weighted beyond the EH that it gives, but in the case of the trinket, with all else equal the one with the skeleton key, has more EH.


Well, no. To illustrate his point, which I don't necessarily agree with, I'll use extreme numbers. Imagine a tank with 100k health, and enough armor to give himself 50% physical damage reduction. So one would describe this tank as having 200k (classical) effective health (ignoring all other sources of damage reduction for this hypothetical). Imagine a tank with 45k health, and enough armor to give himself 75% physical damage reduction. One would describe this tank as having 180k effective health.

Now, imagine the only source of damage is a 20k pre-armor attack every second, and that the only source of healing is an 8k heal every second.

The first tank would take 10k dps after his 50% reduction, while the second tank would take 5k dps after his 75% reduction.

First tank:

100
90
98
88
96
86
94
...
0

Second tank:

100
95
(103) 100
95

Now, obviously this is taking things to extremes, but here you can see that the tank with the higher effective health will inevitably die while the one with less can carry on forever. This is the crux of what Eminai is saying.

Now I have to say my personal recollection (and to be honest, with paladin tanks the way they have been since 3.2, it's actually my druid or warrior co-tanks that were dying first on heroic beasts type situations) is that tank death occurs more often in a 100% to 0% burst over a couple of seconds, with only a couple of significant heals landing, than this whittling down type scenario with tanks not getting topped off for significant periods of time, but I couldn't swear it to be the case.

Edit: I'm suggesting that tanks typically get topped off frequently, every few seconds. Eminai is proposing that a large portion of tank deaths are because of tanks not getting topped off due to failing to avoid a string of hits. If a hit had been avoided it would, I guess, have allowed the healers to catch back up and reset to full health. If this idea has merit then Chill of the Throne has increased the frequency of these unavoided hit strings, leading to an increase in the relative value of armor - on the other hand Icecrown has featured far fewer unavoidable attacks so far than ToGC did.

Anyway, while the tone that Eminai and Jamesvz and the rest of the theck fan club adopted when they decided to troll the boards has been pretty obnoxious, I'll be looking at death logs when hardmodes come out with another gearing option to consider - if I do see a lot of these slow, drawn-out back and forth tank deaths then I'll definitely consider the alternative points of view raised in this thread. Going into them though I don't intend to stop stacking EH as efficiently as I can, until I have a better idea of the specifics of the fights.
Last edited by Rhiannon on Fri Mar 12, 2010 10:58 pm, edited 1 time in total.
Rhiannon
 
Posts: 1061
Joined: Mon Mar 03, 2008 8:17 am

Re: Armor vs HP Math Help

Postby Mex » Tue Jan 26, 2010 10:30 pm

Reading this thread is like watching people trying to use a^2 + b^2 = c^2 to calculate their mortgage repayments.
User avatar
Mex
 
Posts: 476
Joined: Wed Aug 06, 2008 11:52 pm

Re: Armor vs HP Math Help

Postby Flitter » Tue Jan 26, 2010 10:45 pm

Somehow you're trying to say, that the tank who takes less damage over a given timeframe during a bossfight is the better tank, because she requires less healing and therefore receives less healing. Do you really think that's the case? If you take 5k less damage per hit (and thats exaggerated due to the amount of armor we're arguing about) do you really think that resto druid will roll one hot less on you? That the Paladin will stop spamming Holy Light?

Yes, the tank with stamina trinkets will take higher physical hits but will still be overhealed by a lot! It doesn't matter if you wear armor or stamina trinkets, you will still die in 3 hits at 90%-Festergut. But the stamina trinket tank will take more use out of said overhealing than the armor one, therefore increasing her TTL.

Yeah, i'm somehow stating the obvious, but i don't know how to get that even clearer.
Image
Flitter
 
Posts: 34
Joined: Tue Sep 15, 2009 2:31 am

Re: Armor vs HP Math Help

Postby Eminai » Tue Jan 26, 2010 11:02 pm

Flitter wrote:Somehow you're trying to say, that the tank who takes less damage over a given timeframe during a bossfight is the better tank, because she requires less healing and therefore receives less healing. Do you really think that's the case? If you take 5k less damage per hit (and thats exaggerated due to the amount of armor we're arguing about) do you really think that resto druid will roll one hot less on you? That the Paladin will stop spamming Holy Light?

Yes, the tank with stamina trinkets will take higher physical hits but will still be overhealed by a lot! It doesn't matter if you wear armor or stamina trinkets, you will still die in 3 hits at 90%-Festergut. But the stamina trinket tank will take more use out of said overhealing than the armor one, therefore increasing her TTL.

Yeah, i'm somehow stating the obvious, but i don't know how to get that even clearer.


I'm not saying the armor tank should get less heals, I'm saying that they will both get the same heals, but the armor tank will get more use out of both those heals and his current health pool, and because armor double dips and stamina doesn't, armor is quite significantly better than stamina. Also, the idea that either tank gets any use of overhealing is fallacious, as the whole point of over healing is that it is not effective healing.
theckhd wrote:Philosophically speaking, wiping your raid is slightly less serious than handing a baby off to the wrong set of parents.
Eminai
 
Posts: 13
Joined: Tue Jan 26, 2010 5:53 pm

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