I've heard that the general rule is to choose mitigation over endurance whenever possible. I'm hoping that somebody can explain why this is preferable. It seems to me that once you reach a certain point, endurance is the way to go.

In an attempt to understand I'm using some very basic numbers. I've put together a mitigation gear set and an endurance gear set. The "fully optimized" numbers for each set are probably slightly off, but I'm hoping that they are close enough for comparison purposes.

Optimized Mitigation:

End: 2634

Def: 1214 (+13.92%)

Shi: 868 (+16.78%)

Abs: 548 (+13.32%)

Optimized Endurance:

End: 3190

Def: 1121 (+13.13%)

Shi: 740 (+14.71%)

Abs: 341 (+8.76%)

For a typical guardian this provides:

-------------------------------------

Optimized Mitigation:

+27657 health

Def: 21.92%

Shi: 40.78%

Abs: 33.32%

Optimized Endurance:

+35175 health

Def: 21.13%

Shi: 38.71%

Abs: 28.76%

To keep the math simple and skewed as much as possible to favor mitigation, lets assume 4500 DPS all defensible and shieldable. I'll also arbitrarily use 35% armor reduction to favor mitigation.

Mitigation Build takes 1973 damage/second

Endurance Build takes 2050 damage/second

Over the course of a 1 minute interval the Endurance Tank takes 4620 extra damage, roughly the equivalent of single medium heal. So the Endurance Tank requires one extra heal per minute to compensate for the damage taken, but has 7500 extra health to compensate.

I'm having difficulty understanding why mitigation should be favored over endurance. I'd greatly appreciate any insight.

Thanks in advance!

To use your own numbers. Assuming a bossfight is 5 minutes long (many are longer, but don't keep damage up at all times so 5 mins is a good place to start) a high endurance tank will 23000 extra damage over a high mitigation one.

That's healing that could be spent on healing the rest of the raid, the other tank or help dpsing to beat enrage (in case there are fights where the enrage timer is very tight, like NiM SC & FB was).

You're trading a slight buffer for more overall survivability.

In a word, healing. In a few more words, when you being healed, maximum endurance means very little compared to the DPS vs HPS ratio. That extra 7500 HP of yours never comes into play. In many more words: http://www.swtor.com/community/showthread.php?t=616779

In a vacuum (i.e. HPS greater than incoming DPS 100% of the time) the extra health wouldn't matter, but in actual encounters I think it's worth consideration. There are quite a number of times when the tank dies due to lack of healing (raid wide damage that needs to be healed, unlucky RNG, healer incapacitated, etc) where that 7500 HP could make a difference.

It seems to me that it's the difference between 1 extra heal per minute vs. providing a buffer that could sustain a tank through those tough situations.

To use your own numbers. Assuming a bossfight is 5 minutes long (many are longer, but don't keep damage up at all times so 5 mins is a good place to start) a high endurance tank will 23000 extra damage over a high mitigation one.

 

That's healing that could be spent on healing the rest of the raid, the other tank or help dpsing to beat enrage (in case there are fights where the enrage timer is very tight, like NiM SC & FB was).

 

You're trading a slight buffer for more overall survivability.

I think your points contradict one another. On a time-averaged basis, the extra health buffer allows for the healer to perform other tasks more than extra mitigation does. Using my original numbers, the Endurance Tank can more easily survive 10+ seconds while the healer is off doing other things, whereas the Mitigation Tank could not.

In a vacuum (i.e. HPS greater than incoming DPS 100% of the time) the extra health wouldn't matter, but in actual encounters I think it's worth consideration. There are quite a number of times when the tank dies due to lack of healing (raid wide damage that needs to be healed, unlucky RNG, healer incapacitated, etc) where that 7500 HP could make a difference.

 

It seems to me that it's the difference between 1 extra heal per minute vs. providing a buffer that could sustain a tank through those tough situations.

It's a good thought, but practically speaking, the numbers just don't bear it out. Average boss pre-mitigation DPS is 4200 in HM S&V. A well geared shadow has about 73.5% net survivability. Thus, speaking in averages, a well geared shadow is going to require about 1113 HPS to remain health-neutral (note: this is very close to what I see in my combat logs). Think about how *long* that is w.r.t. standard health pools. That same shadow will have about 36500 HP. If they start from 100%, they can go for 32.79 seconds without any healing whatsoever. Longer if they pop a cooldown.

Of course, this is all speaking in averages. What's more interesting is spike damage. Every boss has some sort of ability that hits disproportionately hard. These abilities are generally mitigable though. Additionally, they are always very infrequent. Dash'roode's Gutwrenching Kick hits for about 12.5k post-armor on a shadow if it isn't shielded or defended. So let's assume terrible RNG and just subtract that value from the hitpoints (basically viewing 12.5k as a minimum HP). That still gives us 21.56 seconds without a healer. Notably, Gutwrenching Kick only happens once every 27.72 seconds, which is to say that healers have more than enough time to pop us back up.

The bosses in this game simply don't hit hard enough to make effective HP a particularly exciting metric. Nightmare Kephess is really the only boss I've ever tanked where I wanted more of a buffer, but I also knew that I needed higher probability mitigation at the same time. Eventually, I decided that the mitigation was more important, but it was a somewhat involved and slightly subjective decision path.

The other thing to keep in mind (always!) is that mitigation stats in TOR don't have hard caps (aside from armor, where the hard cap is entirely unreachable). It's not like in other MMOs where you get your percentages to a certain point and just stop stacking that stat. In TOR, you can *always* get better mitigation by increasing your mitigation stat budget. Thus, you will never reach a point where your mitigation is "maxed" and HP is the only stat worth increasing.

Overall, it's a good argument, and I think everyone needs to think about these sorts of questions, but ultimately the game design itself rewards mitigation far more than it rewards endurance. The combination of low boss DPS and self-managed energy pools (for healers) means that you simply want to get your HP to the point where you're not going to get one/two-shot within the cast time of a heal, after which point you do your healers the most service by improving your healing efficiency (i.e. mitigation).

It's a good thought, but practically speaking, the numbers just don't bear it out. Average boss pre-mitigation DPS is 4200 in HM S&V. A well geared shadow has about 73.5% net survivability. Thus, speaking in averages, a well geared shadow is going to require about 1113 HPS to remain health-neutral (note: this is very close to what I see in my combat logs). Think about how *long* that is w.r.t. standard health pools. That same shadow will have about 36500 HP. If they start from 100%, they can go for 32.79 seconds without any healing whatsoever. Longer if they pop a cooldown.

 

Of course, this is all speaking in averages. What's more interesting is spike damage. Every boss has some sort of ability that hits disproportionately hard. These abilities are generally mitigable though. Additionally, they are always very infrequent. Dash'roode's Gutwrenching Kick hits for about 12.5k post-armor on a shadow if it isn't shielded or defended. So let's assume terrible RNG and just subtract that value from the hitpoints (basically viewing 12.5k as a minimum HP). That still gives us 21.56 seconds without a healer. Notably, Gutwrenching Kick only happens once every 27.72 seconds, which is to say that healers have more than enough time to pop us back up.

Hi KBN, and thanks for the response. If there are times when the heals cannot be provided to maintain a health-neutral state, wouldn't it benefit the healer more if the tank had longer survivability? In your example, wouldn't the ability to survive 28.3 seconds without a healer be better than 21.56 seconds without a healer?

Say the tank takes a 12.5k hit at the same time that two party members return from being lost in the sandstorm. HPS on the tank will be reduced while other party members are healed. Add to that the constant stacking damage from the sandstorm debuff and that's more HPS taken away from the tank. Obviously this isn't a very intense fight so it's not the best example to illustrate my point, but I feel the argument is still valid.

With the constant up-and-down intensity of most fights, doesn't a larger health pool serve better than overall mitigation, within reason? It provides for longer survivability during the times that it's most crucial, and only slightly less mitigation on average.

Given that tanks die more often during "abnormal" (that is, not average) situations, wouldn't it be more beneficial to increase your survivability during those abnormal situations as long as you are not making a significant sacrifice to average mitigation?

I've heard that the general rule is to choose mitigation over endurance whenever possible. I'm hoping that somebody can explain why this is preferable. It seems to me that once you reach a certain point, endurance is the way to go.

In a nutshell that extra HP total is useful for the first hit you take, after which it provides nothing for the remainder of the fight, while mitigation provides repeated returns.

If the choice is between an extra 1,000 HP and taking 250 less damage per hit, you take the mitigation. The fight's going to last more than four hits, so mitigation does more for you.

There's more nuance and math to it, but that's the bare bones reasoning behind it.

If there are times when the heals cannot be provided to maintain a health-neutral state, wouldn't it benefit the healer more if the tank had longer survivability?

Yes, but if you go the high mitigation route, there will be less times where you reach that point where you can't keep up with incoming damage.

From the healers perspective, I would rather have a tank with high mitigation. My challenge as a healer is to keep up with the damage the whole raid is taking. The less damage we take, the easier my job. If I can't keep up, cause my tank is taking too much damage, its not going to really matter how long it takes before all die. I suppose if the difference in health pools was something huge, like double or triple, it might be worth it. But its really pretty small all thing considered.

Hi KBN, and thanks for the response. If there are times when the heals cannot be provided to maintain a health-neutral state, wouldn't it benefit the healer more if the tank had longer survivability? In your example, wouldn't the ability to survive 28.3 seconds without a healer be better than 21.56 seconds without a healer?

Sure, 28 seconds is obviously better than 21. But how often are you actually without a healer for that long? It would be optimizing for a raid scenario which never arises.

Say the tank takes a 12.5k hit at the same time that two party members return from being lost in the sandstorm. HPS on the tank will be reduced while other party members are healed. Add to that the constant stacking damage from the sandstorm debuff and that's more HPS taken away from the tank. Obviously this isn't a very intense fight so it's not the best example to illustrate my point, but I feel the argument is still valid.

Actually, this scenario can't happen. The 12.5k hit only happens during the steady-state portion of the fight. Dash'roode's damage is much lower during the transition phase between shields. There is *some* risk that you will lose two healers simultaneously. The odds of this happening on any given raid are 12.97%, which is about once every 8 runs. However, even when this happens, you're still fine. It takes about 12-15 seconds for a competent player to return to the shield after getting lost. Even without the damage being reduced, the tank will still have more than enough life left in him. Not to mention the fact that you actually have two tanks, so the TTK without a healer assuming both are topped off is in the 50 second range.

With the constant up-and-down intensity of most fights, doesn't a larger health pool serve better than overall mitigation, within reason? It provides for longer survivability during the times that it's most crucial, and only slightly less mitigation on average.

This has been the traditional argument for eHP over mitigation, and there's some validity to it. However, fights are not as "up and down" to the extent that you're suggesting. The damage in most fights looks fairly homogeneous for the duration. There are spike attacks, but very few heavy spike phases (Styrak's soft enrage is a good counter-example). Thus, the average case carries a lot more weight than it would with different boss design.

Given that tanks die more often during "abnormal" (that is, not average) situations, wouldn't it be more beneficial to increase your survivability during those abnormal situations as long as you are not making a significant sacrifice to average mitigation?

Tanks do die during abnormal situations, but their probability of dying during those situations is greatly heightened by two factors: low mitigation probability and pre-stressed healers. If the healers are able to baby their energy pool throughout the majority of the fight, they will be in much better shape when called upon to handle a major spike. Remember, self-managed energy pools!

As for the low mitigation probability… As Kitru as pointed out many times, by trading away mitigation to optimize for the "abnormal" case, you are actually making the "abnormal" case happen a lot more frequently (due to the reduction in mitigation probability). That seems like a poor trade. As long as a tank *can* survive the abnormal phases without boxing the healers into an incredibly tight burst window, more HP is just low-value padding.

I think your points contradict one another. On a time-averaged basis, the extra health buffer allows for the healer to perform other tasks more than extra mitigation does. Using my original numbers, the Endurance Tank can more easily survive 10+ seconds while the healer is off doing other things, whereas the Mitigation Tank could not.

I don't think my points contradict eachother at all. A tank taking 23000 extra damage over the course of five minutes will require more heals to stay alive, even if that tank has a buffer of 7500 hit points.

That is more time spent healing that particular tank, rather than doing any other task.

I will say this though, your priorities should be discussed with the healer you normally run with. I, as a healer ,prefer tanks to take less damage rather than having large healthpools, but that's my playstyle. If your healer likes having larger margins of error, then you should gear for endurance to a higher degree.

Tanks really shouldn't be in control of their own gear

Thanks everyone for the discussion! It is helpful to see the other side of things. While I'm still not convinced, this has all been very helpful and informative.

If we accept that fights mostly adhere to some sustained level of incoming DPS (as KBN suggested), it seems that it would be more beneficial to gear for mitigation to the point that you can comfortably survive those sustained levels and then min/max to provide more room for error during the heavy burst damage phases.

Again using KBN's numbers, he only needs to receive 1113 HPS to remain health-neutral. If he were to gear more towards endurance, he might need 1200 HPS (still easily maintainable) to remain health-neutral, but he would also be able to maintain greater survivability during those critical high damage phases (which are the times that a tank is mostly likely to die).

When we look at the two builds, the difference in DPS taken is so small (75 DPS), it wouldn't even be noticed by a healer. A single heal can fluctuate by hundreds of points just based on RNG, and we're talking about a difference of less than 100 between the two builds. In exchange for that very small decrease in average damage taken, you gain a relatively significant increase in the ability to withstand a heavy burst phase.

Hi KBN, and thanks for the response. If there are times when the heals cannot be provided to maintain a health-neutral state, wouldn't it benefit the healer more if the tank had longer survivability? In your example, wouldn't the ability to survive 28.3 seconds without a healer be better than 21.56 seconds without a healer?

 

Say the tank takes a 12.5k hit at the same time that two party members return from being lost in the sandstorm. HPS on the tank will be reduced while other party members are healed. Add to that the constant stacking damage from the sandstorm debuff and that's more HPS taken away from the tank. Obviously this isn't a very intense fight so it's not the best example to illustrate my point, but I feel the argument is still valid.

 

With the constant up-and-down intensity of most fights, doesn't a larger health pool serve better than overall mitigation, within reason? It provides for longer survivability during the times that it's most crucial, and only slightly less mitigation on average.

 

Given that tanks die more often during "abnormal" (that is, not average) situations, wouldn't it be more beneficial to increase your survivability during those abnormal situations as long as you are not making a significant sacrifice to average mitigation?

You are forgetting that if you stack more health then you are essentially reducing mitigation stats which means you are taking more dam and need more heals to cover that dam. So you have to remember that more health doesn't equal more survivability in this game, UNLESS we see a boss fight that does 100% elemental/Internal and then and only then will your health pool be the deciding factor on how long you survive. And since from what KBN has been reporting the total amount of internal/elemental dam from HM S & V is less than 2% health doesn't matter as much as it would say it be 10%.

so what do u think is an acceptable trade off?

at best you can trade 1 point of mitigation for 2 points of endruance.

shadows: .000039 mitigation for 21.6 health

PT: .00002 mitigation for 21.6 health

jugg: .000022 mitigation for 21 health

assuming boss is doing 4.5k dps

change in time to kill dTTK= hf/(mf*dps)-h0/(m0*dps)

where

hf is health with mitigation traded for health

mf is mitigation after that trade

h0 is health before that trade

m0 is mitigation before that trade

hf=h0+dh

where dh is amount of health gained from trade

mf=m0-dm

where dm is mitigation lost due to trade

dTTK=-(m[0]*dh+h[0]*dm)/((-m[0]+dm)*dps*m[0])

using 2400 stat pools:

shadows: m0=0.3193, dm=.000039, dh=21.6

jugg: m0=0.3257, dm=.000022, dh=21

PT: m0=0.2807 , dm=.00002 , dh=21.6

this gives dTTK:

shadow: 0.150e-1+8.5020*10^(-8)*h[0]

jugg: 0.143e-1+4.6090*10^(-8)*h[0]

PT: 0.171e-1+5.6420*10^(-8)*h[0]

so about 0.15 seconds longer for every point of mitigation you trade for 2 points of endurance. and you take an extra 0.12 dps per point that you trade (average between classes) (before self heals/absorb).

so if you trade 100 points of mitigation for 200 points of endurance you will take 12 more damage per second, but will live 15 seconds longer.

think i did that right

Basically it comes down to 2 things. Your healers, and how well you use your cool downs. Neither option is going to save or cost you a wipe if both you and your healers play well and use your cool downs correctly so talk to your healers about which they prefer, since they're the ones it actually affects.

In terms of effective HP, the usual argument for it is to maximize survivability against a string of unmitigated hits. KBN and I discussed exactly that here: http://www.swtor.com/community/showthread.php?t=548875

The conclusions of that are:

- On average, that string of unmitigated hits takes over 20 seconds

- On average, you see that string of unmitigated hits every 3 weeks if you raid 10 bosses a week.

In a nutshell that extra HP total is useful for the first hit you take, after which it provides nothing for the remainder of the fight, while mitigation provides repeated returns.

 

If the choice is between an extra 1,000 HP and taking 250 less damage per hit, you take the mitigation. The fight's going to last more than four hits, so mitigation does more for you.

 

There's more nuance and math to it, but that's the bare bones reasoning behind it.

Actually the extra HP is only used at the end of your 'life' because it's only useful if the hit would have killed you at a lower HP level.

Eg. Two tanks, one has 20k health the other 25k. That extra 5k health is only useful to the second tank if they have already taken 20k damage first. If both tanks have taken 19.5k health then the first tank will be on 500hp and the second 5500hp. The next hit will probably kill the first tank, while the second could survive it. This is ignoring the Self heal mechanics (shadows TkT) etc.

If you're expecting every hit to take a full 50% of your hp bar, then sure total HP starts to matter, but only inasmuch as it allows you to live to needing that third hit to take the last 1 hit point.

Besides, I said in my previous post that was a nutshell explanation, not a detailed analysis. The thing that matters is the rate at which hp is going up and down, not how much of it there is when topped off.

I like to test on my shadow and confirm what our theorycrafters find. In addition, I like to add some things as well and peruse my logs. Having said that, I have tried quite a few setups (43k hp and 38khp with BIS budget mitigation stats). In the 43k HP scenario, my logs show a lot more damage taken. I can even give you the numbers, roughly 200k more if i go high endurance versus BIS mitigation. Now that 200k more needs to be healed and I have done almost 300k self healing during the whole fight. Now there are moments when in 16 man people are careless and pull some adds or get damage from other sources.

When the above happens, the healers are giving me a heads up and i react accordingly to help them. Now we ran the same fight with BIS mitigation stats and when we take unnecessary damage the healers got a bit edgy because of my lower HP, at first but then they noticed that my damage taken was so little sometimes they had no healing to do at all on me. They said and i quote 'we have nothing to heal irisa, we are waiting for the spike damage, there it is, bang we gotta top you back up'. So 43k hp, did give me more self heals and a higher instant battle readiness heal but way more damage taken. The theory crafters know what they are doing and it is good that we get the chance to confirm their findings.

Here is a different way of looking at it....

You have two cars. The goal is to drive as far as you can. One car gets better gas mileage. The other has a larger gas tank. All others things being equal, which one do you want?

There isn't any right answer per say here, just different approaches. I'm a fan of the high gas mileage (i.e. high mitigation) option.

I've always felt as though the Endurance route presupposes that the other people in your raid group are incompetent.

All of the "if" scenarios center around resource mismanagement on the part of the healers, or excessive healing on other raid members, meaning those other raid members didn't know the fight mechanics and how to minimize their damage taken.

If you are in a raid group with underskilled members and have a compelling reason not to leave, building for Endurance to accommodate the derpiness makes sense. If you're in a group with 8 or 16 people who all know what they're doing, it's virtually unnecessary.

Maybe I'm just an elitist schmuck, but I'd much rather raid with a group of 8 people who know what they're doing.

...If you are in a raid group with underskilled members and have a compelling reason not to leave, building for Endurance to accommodate the derpiness makes sense....

ROFL! That is soooo now gonna be my quote of the week!

ROFL! That is soooo now gonna be my quote of the week!

Glad I could make you smile.

In seriousness, though, there are plenty of raid groups (just, usually not the ones hunting server firsts) who have membership criteria beyond pure skill, and that's totally fine. If you need to change your tactics to accommodate them, there's nothing wrong with that.

Just, recognize what you're doing and why, and own up to the fact that your healers aren't the strongest and benefit from extra cushion, or that your DPS don't pay as much attention as they should. No one's going to demonize anyone for saying that, it's just the way things are sometimes. It's not optimal, but it is entirely sensible depending on the group.