Jump to content

Tanking Stat Weights


dipstik

Recommended Posts

2.2 Info here: http://www.swtor.com/community/showthread.php?t=645948

Updated 5/30/2013: Using simulator for kinetic bulwark, found time averaged contribution to absorb of 5.52%, which switched around 20 defense to absorb.

Updated 5/14/13: AMR for Guardians changed back to underworld enhancements. at 4500 dps, it turned out to be taking 2 less dps per second at the cost of 2k health pool. 32.7k vs. 34.7k according to AMR.

Updated May 10 2013: Updated total mitigation data for vanguards since the stance provides extra 1.5% armor rating than stated (no difference to stat weights, just overall mitigation, so the relic numbers are a little off). added numbers under the slope and intercept to show how mitigation per point changes at higher values of stat pool.

Updated May 6 2013: Changed the AMR profile for guardian to use crafted 66 immunity and sturdiness, which gives better overall mitigation. also updated all profiles for using a nano stim.

Update 4/30/2013: New calc for Kinetic bulwark used. makes shadows want a little more absorb for 790 gearing.

Updated 4/29/13 New askmrrobot links where sins/shadows now use high mitigation due to high dps numbers (very low internal/elemental). due to implants and ears affecting the minimum shield rating, i redid the relic numbers with these profiles as a launching point (where the 2399 stat budget was the starting point, with itemization limitations). suugest just shooting for the AMR profiles.

Updated 4/27/13 for static defense on defense proc relic

Updated 4/20/13 for 2.0

 

HOW TO USE THE TABLES: Depending on your gear level, you will have a different mitigation pool (in the tables this is "total"). To find the values of your mitigation stats, such as defense, shield and absorb, for which the damage you take is minimized, add up your mitigation ratings (found by hovering over your character sheet) and use that as your total (assuming you are not stacking accuracy). In the total column, find your total, and use the mitigation ratings in that row to achieve minimum squishiness (d is defense, s is shield and a is absorb)

 

I have also provided tables for the relics as they affect each class. I used the mitigation (post armor) while the relic was "on" and applied that mitigation to the uptime (assumeing 23 seconds between procs), and used the AMR profile stat pools as the baseline for the downtime. the weighted value is the expected damage over an extended period of time. You want to use the lowest possible, but remember you can only use one proc relic (unless you go with PVP relics which were not calculated).

 

DISCLAIMERS:

In the builds linked (askmrrobot.com), the last remaining points of each build are up for debate, but they do not matter for the calculations. most choices are bertween different force regen or damage concerns.

 

I have not taken "Oh Crap!" cooldowns, such as oil slick and the like into account because assuming that you have a 7% defense buff at all times is drastically different then having a 20% buff for 20 seconds every 60 seconds. I have tried to include all buffs that come from skilltree as well as from debuffs on the boss from the tanks rotation, such as slow time and the like. I do not attempt to use mitigation stats with unmittigatable damage in mind.

 

I am hoping that people will offer suggestions as to how these assumptions could be altered to better estimate optimal stat weights. Perhaps the 50/50 split between accuracy-damage types is naive, If statistics are presented, or better yet, revised formulas, we can get better data.

 

examples of formulas used to minimize squaishiness are presented in the shadow section. I am using a revised version of keyboard ninjas spreadhsheet, where i have modified many things and double checked the eqautions to actual characters in game to make sure everything lines up.

 

I may have missed something ... so review my work

 

all of this assumes:

38.965% of damage is melee/ranged kinetic/energy at 90% accuracy

38.965% of damage is melee/ranged kinetic/energy at 100% accuracy

20.7% of damage is force.tech kinetic/energy damage

1.37% of damage is force/tech internal/elemetal damage

 

base armor rating for light armor at arcanian is 2894

base armor rating for light armor at underworld is 2999

base armor rating for heavy armor at arcanian is 5466

base armor rating for light armor at underworld is 5666

 

post armor mitigation:

 

(0.7793*(1-dr)*(0.5*(1-d-0.1)*(1-s*a)+0.5*(1-d)*(1-s*a))+0.207*(1-dr)*(1-r)*(1-s*a)+0.0137*(1-ir)*(1-r))

 

where d-0.1 from is from 90% accuracy attack, 0.5 from fraction melee/ranged kinetic/energy at 90% accuracy

dr is damage reduction form armor, to kinetic and energy damage

d is defense fraction

s is shield fraction

a is absorb fraction

ir is internal resistance

r is resiatance

 

____________________________________________________________

shadow tank 2.0 http://swtor.askmrrobot.com/character/65b987d7-c58f-4cde-bcb3-a4087603dcb5

 

health= endurance*10*1.08, 0.05 from warrior/knight buff, and 0.03 from skill tree. remember you get a

base of 360 endurance before mods etc.

 

armor= armor_rating*(1+1.15), where 1.15 comes from stance

 

kinetic/energy resist (damage reduction from armor and talents) = (armor/((armor)+240*55+800) + 0.02

+0.02 +0.05, where 0.02 comes from skill tree, 0.02 come from 4 set bonus and 0.05 from slow time

 

internal resist (internal/elemental damage reduction) = 0.10 + .02 + 0.09 + 0.02 +0.05 , where 0.1 is from willpower buff, 0.09 is from skill tree (tech mastery), 0.02 form 4 set bonus, 0.02 is form skill tree (jedi resistance) and 0.05 from slow time

 

force/tech resist = 0.02

 

defense = 0.1 + 0.05 +0.06 + 0.30*(1-(1-(0.01/0.3))^((Def_Rating/55)/1.2)) , where 0.1 from base, 0.05 from accuracy debuff of force breach, 0.06 from skill tree

 

shield = 0.15 + 0.2 + 0.05 + 0.5*(1-(1-(0.01/0.5))^((Shield_Rating/55)/0.78)) , where 0.15 comes form stance, 0.2 comes from kinetic ward (with 4 peice set bonus), 0.05 is base shield from shield offhand.

 

absorb = 0.04 + sum_i=0..8(.01*i*binom(i,15,shield)) + 0.2 + 0.50*(1-(1- (0.01/0.5))^((Abs_Rating/55)/0.65)) , where 0.04 comes from skill tree, shield dependent portion comes from kinetic ward stacks eaten to give bulwark stacks (15 comes from (1-d)*20/.87, 20 is length of ward and .87 is swing time for snv hm), 0.2 comes from base from shield offhand

 

self heals:

health*0.08 every 12 seconds from 3 stack telekinetic throw

368 every 6 seconds or so from combat technique.

 

you get 0.072 hps per point of endurance and 3.9E-5*dps mitigation per second per point of mitigation

 

total d s a mit slope/intercept

2000 378 790 832 0.3310 -0.000039

2100 431 790 879 0.3268 0.40867

2200 484 790 926 0.3227

2300 537 790 973 0.3187 3.9598E-05

2400 590 790 1020 0.3149 3.83053E-05

2500 643 790 1067 0.3112 3.70595E-05

2600 661 825 1114 0.3076

 

 

 

 

amr: 531,868,1050

 

 

 

 

SHADOW RELICS:

healing does 634/23 s... need 4519+ dps to make any relic better than healing relic (see table of dps values)

 

base relic weighted diff mit=634 heal dps

def proc 0.32468 0.30595 0.31979 -0.0061 4519

abs proc 0.32594 0.30728 0.32107 -0.0049 5625

shd use 0.32594 0.30739 0.32130 -0.0046 5945

pvp(ewh) 0.32594 0.32128 0.32128 -0.0047 5913

use def 0.32594 0.31077 0.32341 -0.0025 10902

 

 

 

 

 

 

______________________________________________________________________

 

Guardians http://swtor.askmrrobot.com/character/f4f059f3-a83f-4d82-8bb1-ff52c0cf84e6

 

skill tree

 

dust storm: 5% defense

guard stance: 15% armor rating boost

blade barricade: 5% defense. 5% resistance

blade barrier: (0.05 * 11180 + BonusHealing) * 1.2. bonus healing = 0.14*strength+0.17*power+0.17*force power. remember you get a 5% buff to power and force power and a 11% buff to strength.

shield specializtion: 4% shield

inner peace: 5% internal resistance

guardian slash: 3% kinetic/energy resistance, 3% internal resistance

single saber mastery: 3% defense

 

sorusu form: 61.5% armor rating, 15% shield, 6% kinetic resistance, 6% internal resistance

offhand: 5% shield, 20% absorb

base 5% defense.

 

get 10% internal resistance from willpower buff.

 

get 0.01554 absorb per second for each point of strength (with 11% boost) and 0.01785 absorb per second from power. you get 3.08E-5*dps mitigation per second form each point of mitigation.

 

absorb around 117 per second.

 

total d s a mit slope/int

2000 850 790 360 0.3385 -3.08111E-05

2100 903 790 407 0.3351 0.399834577

2200 956 790 454 0.3319

2300 1009 790 501 0.3288 3.12649E-05

2400 1062 790 548 0.3257 3.02887E-05

2500 1115 790 595 0.3228 2.93461E-05

2600 1168 790 642 0.3200 2.84358E-05

 

 

amr: 1178,868,623

 

GUARDIAN RELICS:

base relic weighted diff

def proc 0.32565397 0.311133188 0.321865939 -0.0048

abs proc 0.326631781 0.311936685 0.322798277 -0.0038

shd use 0.326631781 0.313514168 0.323352377 -0.0033

pvp(ewh) 0.326631781 0.322987574 0.322987574 -0.0036

use def 0.326631781 0.314868838 0.32467129 -0.0020

 

_____________________________________________________________

 

Vanguard http://swtor.askmrrobot.com/character/42750c58-32fc-4f35-92dd-03637ae825f7

 

combust: 5% kinetic reduction, 5% internal reduction

rebraced armor: 16% armor rating

shield vents: 2% shield

ablative upgrades: 4% absorb

ion screen: 2% kinetic reduciton, 2% internal

deflective plating: 4% defense

 

heat screen: 3% for 6 seconds after heat blast

heat blast: during heat blast: (25+0)*1.5+(25+1)*1.5+(25+2)*1.5*1+(25+3)1.5

taking time averaged for both gives (3*6+25*1.5+26*1.5+27*1.5+28*1.5)*(1/12) = 14.75% absorb. KBN got 15.33. so im just going to call this number 15%.

 

power armor: 2% kinetic reduction, 2% internal reduction

integrated cardio package: 3% endurance

 

stance: 5% kinetic damage reuction, 5% internal damage reduction, 60% increase armor rating, 15% shield chance,

 

extra 1.5% armor rating from bug for stance.

 

set bonus: 2% resistance, 2% defense

resist buff: 10% internal resistance

 

because you want shield here, i didnt need to run numbers for 720 and 790 separately.

 

total d s a mit slope/int

2000 846 740 414 0.29058481 -2.77459E-05

2100 828 802 470 0.28777525 0.346038276

2200 814 862 524 0.284975389

2300 803 921 577 0.282189185 2.7862E-05

2400 794 978 628 0.279419981 2.7692E-05

2500 788 1034 679 0.276670619 2.74936E-05

2600 783 1089 728 0.273943515 2.7271E-05

 

 

amr: 798,1052,719

 

VANGUARD RELICS:

base relic weighted diff

def proc 0.281093161 0.267712039 0.277602433 -0.0044

abs proc 0.281994229 0.268823737 0.278558448 -0.0034

shd use 0.281994229 0.267568957 0.278387911 -0.0036

pvp(ewh) 0.281994229 0.277392254 0.277392254 -0.0046

use def 0.281994229 0.269959988 0.279988522 -0.0020

Edited by dipstik
Link to comment
Share on other sites

  • Replies 328
  • Created
  • Last Reply

Top Posters In This Topic

And it's still completely incomprehensible..

 

How are we supposed to read your post? Are the numbers for the stats what we should aim for, or? And which line then - or why more than one line for optimum stats to aim for?.. It makes absolutely NO SENSE..

 

I don't mind math heavy posts, when they are actually well written and are comprehensible by people other than the author..

Link to comment
Share on other sites

I don't mind math heavy posts, when they are actually well written and are comprehensible by people other than the author..

 

They're comprehensible. The values at the left of each column are the total stat budget and the 3 values to the right of it are the ratings that you should stack up to with that given budget. It's actually pretty simple.

Link to comment
Share on other sites

As far as I know, all boss attacks beyond "special" attacks have 90% accuracy. Special attacks (e.g. Kephess's Empowered Slash) have an insane amount of accuracy baked in as they are designed to be undefendable (even with cooldown). These conclusions are based on a fairly large amount of log analysis that I did as part of building my corpus of boss DPS information. I didn't specifically look at accuracy though, so more rigorous work here would probably give a better answer than I'm able to at present.
Link to comment
Share on other sites

They're comprehensible. The values at the left of each column are the total stat budget and the 3 values to the right of it are the ratings that you should stack up to with that given budget. It's actually pretty simple.

 

Why on earth would I write that a post is incomprehensible if it actually wasn't?.. And yes, I can add numbers together, that doesn't make the post any more well written.

 

What's my statsbudget at any given point in the game? From the outside it just looks like a bunch of random lines with the numbers added up on the side..

 

So NO, it's not comprehensible or useful in ANY way.. And you writing "yes it is", doesn't make it so..

Link to comment
Share on other sites

What's my statsbudget at any given point in the game? From the outside it just looks like a bunch of random lines with the numbers added up on the side..

 

Your stat budget is equal to the sum of all of your secondary stat ratings, which means it can very pretty heavily depending on the types of mods and enhancements you choose to use. At the moment, if you're using 2 passive relics and full mitigation mods and enhancements with 27 grade gear, you'll pop right a bit over an 1800 stat budget. With 2 non-passive relics, it's manage just ahead of a 1600 stat budget.

 

These numbers can actually be useful if you're looking into how to tweak your stats for better survivability. When someone says "I have XX% abs, XX% shield, and XX% defense; how can I tweak those to be more appropriate?" you can simply ask what their ratings are, inform them of their stat budget, and tell them how to retweak their numbers to a more appropriate ratio. Shield will almost always be a bit higher than what is listed in the optimal stat budget (since Shield doesn't compete with anything else in the stat budget

 

So NO, it's not comprehensible or useful in ANY way.. And you writing "yes it is", doesn't make it so..

 

It's incomprehensible to you because you don't know what you're looking at. If you did, it would be perfectly comprehensible. Sure, it's not formatted perfectly, but it's more than enough to get the point across. Just because *you* have no clue what it means, does not equate to information being incomprehensible or useful since, to someone that *does*, it's entirely comprehensible *and* useful. I was simply telling you *how* to use it and *how* to read it for the relevant information so that you could remedy your incomprehension. Apparently you're not interested in that and more interested in just criticizing him for forgivably bad formatting (he's drawing info from Excel or Googledocs and that's notoriously difficult to put into a forum friendly format without a lot of extra work).

Link to comment
Share on other sites

Good write up. I can confirm that for at least PTs the info is correct since ive done the math on them and got the same results.

 

The one thing that i did was also compare the value of endurance vs defensive stats once you get into that 1700 and 1800 bracket. Even with the DR getting that high on my shield and absorb, it still beats endurance in the 1700 range but by the time you're getting into the high 1800s it becomes more debatable.

 

I am in no way saying that endurance stacking becomes viable, but i have swapped out a few mods for the lettered ones with higher stam.

 

As tanks, we only have maybe 1 more tier of gear before we'll be looking at endurance more and more from hitting the higher DR ranges on some of our stats. By then we'll need an xpac like wow does to "reset" our stats.

Link to comment
Share on other sites

The one thing that i did was also compare the value of endurance vs defensive stats once you get into that 1700 and 1800 bracket. Even with the DR getting that high on my shield and absorb, it still beats endurance in the 1700 range but by the time you're getting into the high 1800s it becomes more debatable.

 

Given that Endurance doesn't actually affect survivability for a powertech, I'd be curious to see on what grounds you were comparing them. If your goal is to minimize healing required, endurance will never be the best stat for a powertech, since it only serves to provide a buffer. This is distinct from a shadow/assassin, where stacking endurance actually increases survivability (decreases healing required) due to the self-heal mechanic.

Link to comment
Share on other sites

I was assigning an actual stat weight for endurance for a while but in the end i just went with my experience in mmos and some common sense . Since i duel-main healer/tank in every mmo, i have a pretty good feel for dealing with some of the intangibles that's hard to model. I can post some of my old notes where i tried to give endurance a defined value later.

 

If you compare 2% shield rating to 2k hp, shield will always win. But if you take 1900 total stats, which breaks down to D:435 S:817 A:647, your shield rating is so high that you can trade part of that for endurance by using a lettered mod or other means. Now you're looking at a ratio of stat loss to endurance gain that can give you some options when you look at the different boss fights. For a fight like nim kephess, im willing to trade a little more mitigation for endurance since kephess at 100-60% is the most healing intensive for us and i want to give our healers a little more wiggle room.

 

Again, i am in no way saying stack endurance, just that those lettered mods become more attractive as you work your way up the total stat list.

Link to comment
Share on other sites

Again, i am in no way saying stack endurance, just that those lettered mods become more attractive as you work your way up the total stat list.

 

The lettered mods trade Absorb/Defense for Endurance, not Shield. Shield only exists on enhancements, so you are choosing to use the high Endurance enhs rather than the high Shield enhs to trade Shield for Endurance.

Link to comment
Share on other sites

Approaching the question of how to actually value endurance versus a trade-off in mean mitigation lead me to make a spreadsheet evaluating the best cast scenario of Time To Killed (highly slanted in favor of MM, as there is no seperate calculations for attacks that bypass D/S/A).

 

I found that even with the formulas biased in favor of mitigation, the B variant mods outperform the unlettered variants in TTK by a very good margin (over 10k EHP, with only a 0.56% mean mitigation loss). And this is on a vanguard tank, which has the least to gain from a larger endurance pool.

 

My numbers (I use all 27B's):

Defense: 377

Shield: 748

Absorb: 587

D+S+A = 1712

MM: 77.69

HP Buffed: 27,950

EHP: 113,557 hp

TTK Buffer: 14.20 Seconds

HPS Required: 1740 HPS

 

Number if swapped out for unlettered 27 mods:

Defense: 407

Shield: 776

Absorb: 619

D+S+A = 1802

MM: 78.25

HP Buffed: 24,961

EHP: 103,411 hp

TTK Buffer: 12.93 Seconds

HPS Required: 1784 HPS

 

TTK Buffer and Min HPS based on a boss with a raw dmg output of 8,000 DPS.

 

Spreadsheets used to figure this out:

 

https://docs.google.com/spreadsheet/ccc?key=0Ar2yXaOUZhwCdHVyY01zVExmV29pWHR6cHc2Y2hzYWc

https://docs.google.com/spreadsheet/ccc?key=0Ar2yXaOUZhwCdEVhbjFiY0pHZjF4bnExRXJ3cU1VWGc#gid=0

Edited by Redklaw
Link to comment
Share on other sites

Unfortunately, all that changes when you are getting healed. The ability to handle 230 more DPS with mitigation heavy set up and the same constant healing reflects this. Further, eHP is a terrible (ok maybe not terrible, but redundant) metric in this game. Check out this post for why: http://www.swtor.com/community/showthread.php?t=548875

 

I also assume by looking at your spreadsheet that you're using stock Campaign/Dreadguard gear given that as you go up a tier, your Mean Mitigation drops for both Shadow and Guardian. Yet you appear to be comparing it to your actual (presumably re-itemized) set up.

Link to comment
Share on other sites

No neither the stock nor the compared theoretical set are stock, both are optimized, my spreadsheet gets the same values as the OP's Lagrange points, so it seems legit. And compare HPS differences. The point is you can actually develop a metric to analyse for the value of EHP versus impact of MM on total tankability.

 

230 extra tankable dps sounds like a lot...

 

Until you realize that it's about 8.9k and 9.1k dps for a single 2k HPS healer. You have to look at the scales, NOTHING in the game hits near that hard outside of an enrage timer.

 

Once again the numbers are very skewed in favor of MM builds, actual mitigation numbers are probably closer to 10-12% lower across the board, which will further skew the numbers in favor of a balanced build.

 

I am not saying Stack Endurance. I'm saying if you can get an extra 10% HP buffer and only sacrifice half a percent of MM, you should probably do it.

Edited by Redklaw
Link to comment
Share on other sites

No neither the stock nor the compared theoretical set are stock, both are optimized. And compare HPS differences. The point is you can actually develop a metric to analyse for the value of EHP versus impact of MM on total tankability.

 

Your PT set may be optimized, but I guarantee your Sin and Jugg ones aren't.

 

The point of the linked thread was two fold:

- The argument for eHP is that a series of unmitigated attacks are what kill tanks. In this game, the string of attacks to kill the tank is approximately 20 seconds long and occurs on average once ever 2 weeks if you are running HM EC, TFB and NMP each week.

- eHP is only useful when not being healed. Add a healer in, and the window for eHP to be relevant is so long that stacking HP isn't needed in the first place. Unless your healer wants to ignore you for 20+ seconds AND you have no CDs left.

Link to comment
Share on other sites

Ignore the other stuff, it's left over from playing with something else, I didn't create that spreadsheet for this thread so the non-highlighted stuff is all pretty random. Also if you're looking at this thread in a couple weeks, that link is probably totally changed for some other purpose.

 

Also note that in practice, both the numbers for MM and EHP don't matter to much, the differences are really to small to notice much in a fight with the way the game is currently tuned. You really aren't going to visibly see a difference in 10k ehp or 0.5% mean mitigation in a fight.

 

I've been fully mitigation stacked and performed just fine in top end raids and seen tanks endurance stack (over 30k buffed) and perform just as well. Most of your tanking effectiveness will be determined by proper rotation, cooldown usage, and mechanics handling.

 

The argument is academic, but I really like tuning so I wanted to take a look at an additional metric.

 

Note: I misspoke earlier, when I said the "stock set" I meant what I'm actually wearing (hey, it's stock to me!). I see where the confusion is coming from.

Edited by Redklaw
Link to comment
Share on other sites

They're comprehensible. The values at the left of each column are the total stat budget and the 3 values to the right of it are the ratings that you should stack up to with that given budget. It's actually pretty simple.

 

Hey Kitru - I think I remember you saying you have a jugg tank? If so how closely do you follow the numbers here? I have roughly 1800 budget now stacked to 29/50/50. If I were to follow this, I'd have roughly 33/45/45, is this really the way to go for juggs? Doesn't stacking so much defense end up with the jugg having very spiky dmg? It's a question for anybody but Kitru always gives the most thorough answers so wanted to try picking his brain a little :p

Link to comment
Share on other sites

I think I've got it now then.. But if the OP really wants to help out the community, a little bit more of an explanation is probably worth putting into the posts..

 

Now the question is, are these stat weights even possible in game? Or are they simply the mathmatically optimal ones?.

 

Right now my VG is 176 727 604.. Without sacrificing any shield/absorb I can get something like 49 defense more with just BH gear.. I don't have the means to kit out my VG in rating 150 gear in all slots. I only have BH rating 146 to play with still. Then I could go a bit more defense happy, but that would most of all affect my absorb rating, and not my shield chance.. And that just seems like a dumb thing to do. Ofcourse I could just boost my accuracy in order to reach a perfect stat weight - but really?? :)

 

For fun I looked at the DG supercommando set, and the mitigation stats on the core supercommando set are pretty horrific, and even has Defense as the highest stat.

Link to comment
Share on other sites

If so how closely do you follow the numbers here?

 

I've got a Guardian tank, but my Shadow is my main so I don't focus as heavily on gearing her perfectly as I do my main. As such, I do actually try to follow the general gist of the optimal weights, mainly because it's true that defense is substantially useful to a Guardian than Shield/Abs.

 

Doesn't stacking so much defense end up with the jugg having very spiky dmg?

 

When calculating spikiness, you first have to look at each of the comparative chances of mitigation: with 29/50/50, you've got a 29% chance of taking no damage, a 35.5% chance of taking half damage, and a 35.5% chance of taking full damage; with 30/45/45, you've got a 33% chance of taking no damage, 30.15% chance of taking half damage, and a 36.85% chance of taking full damage. Since your DR should be completely static (same talents and same armor rating), the only difference in spikiness arrives from the difference in those percentages (Shadows have higher spikiness than either of the other tanks specifically because they have *way* lower DR). You're only 1.35% more likely to take an unmitigated hit so you're *slightly* spikier, but it's such a small amount that its not likely to be noticeable. In the same vein, you'll end up taking less damage over the entire course of a fight: the first setup would have you take ~53.25% damage pre-DR; the second would have you take ~50.42% damage pre-DR. That nearly 3% difference is pretty massive, so you're trading off a bit of spikiness for a pretty large increase in total mitigation.

 

The reason that Guardians care so little for Shield/Abs is pretty simple actually: Shield and Absorb are synergistically useful. To make Shield useful as a mitigation stat, you need a lot of Absorb; to make Absorb useful as a mitigation stat, you need a lot of Shield. Shadows and VGs both get plenty of Shield and Absorb from their respective tank specs. Guardians, on the other hand, do not. In fact, they get a lot of Defense, which ends up just making Defense even more valuable (the higher a mitigation stat gets as a percentage, the more valuable from a comparative mitigation standpoint further increases get). Once you realize that, it becomes pretty obvious that Defense stacking is the way to go.

Link to comment
Share on other sites

Now the question is, are these stat weights even possible in game? Or are they simply the mathmatically optimal ones?.

 

They're not explicitly possible, mainly because it assumes that Shield directly competes with Absorb and Defense rating, which doesn't happen except in augs, which should never be devoted to Shield rating. The way that I use them is to determine how to tweak my relevant Absorb/Defense ratio to get closer to the optimal loadout. They're not explicit numbers that I aim for, but they're decent guidelines for what general ratio to approach while getting kitted out.

Link to comment
Share on other sites

When calculating spikiness, you first have to look at each of the comparative chances of mitigation: with 29/50/50, you've got a 29% chance of taking no damage, a 35.5% chance of taking half damage, and a 35.5% chance of taking full damage; with 30/45/45, you've got a 33% chance of taking no damage, 30.15% chance of taking half damage, and a 36.85% chance of taking full damage. Since your DR should be completely static (same talents and same armor rating), the only difference in spikiness arrives from the difference in those percentages (Shadows have higher spikiness than either of the other tanks specifically because they have *way* lower DR). You're only 1.35% more likely to take an unmitigated hit so you're *slightly* spikier, but it's such a small amount that its not likely to be noticeable. In the same vein, you'll end up taking less damage over the entire course of a fight: the first setup would have you take ~53.25% damage pre-DR; the second would have you take ~50.42% damage pre-DR. That nearly 3% difference is pretty massive, so you're trading off a bit of spikiness for a pretty large increase in total mitigation.

 

The reason that Guardians care so little for Shield/Abs is pretty simple actually: Shield and Absorb are synergistically useful. To make Shield useful as a mitigation stat, you need a lot of Absorb; to make Absorb useful as a mitigation stat, you need a lot of Shield. Shadows and VGs both get plenty of Shield and Absorb from their respective tank specs. Guardians, on the other hand, do not. In fact, they get a lot of Defense, which ends up just making Defense even more valuable (the higher a mitigation stat gets as a percentage, the more valuable from a comparative mitigation standpoint further increases get). Once you realize that, it becomes pretty obvious that Defense stacking is the way to go.

 

Thanks Kitru, if that's the case and we prioritize defense over shield/abs, does that mean you would advocate using the PT set bonus for the +2% defense over the warrior tank set bonuses? Since that's taking a full 2% defense over a boost in our absorb ability.

Link to comment
Share on other sites

does that mean you would advocate using the PT set bonus for the +2% defense over the warrior tank set bonuses? Since that's taking a full 2% defense over a boost in our absorb ability.

 

No, I wouldn't. The 20% increase in your absorb is a phenomenal boost to your survivability, not to mention that the 2 pc with the Supercommando is wasted whereas the 2 pc for the War Leader set makes your already powerful CDs even better. It's a question of 2% Defense being nice, but not as awesome as 20% more absorb (which is useful against *all* attacks, not just M/R) *and* 1 extra second of 100% Defense w/ Saber Ward and extending Warding Call's duration to be equal to the standard CD duration of 12 seconds.

Link to comment
Share on other sites

The number are VERY possible to reach in a way that you reach a statistical parity, my numbers for 1712 posted earlier are my actual numbers with gear equipped and I use a spreadsheet to balance, so I am only a bit off (less that 20) on all the stats listed for 1700.

 

One of the things that makes this easier is using the appropriate mods and enhancements in order to make stat balancing with augments easier.

 

For example, as a Vanguard / Powertech you can balance your stats a bit easier if you always try to have absorb on your mods and defense on your enhancements (viceversa is nearly as good, but you need a bit more absorb augments than the rest to balance out). The reason you want to have balanced mods is in the event of an upgrade that doesn't conform to your gear scheme, so you can compensate via swapping augments to re-balance later.

 

The pattern is probably different with the other tanks, I haven't bothered stat balancing my Jug or Assassin, since one is on a different server than my main and the other I never got to that gear level.

 

Question for Kitru:

 

Why would you never devote augments to shielding? My vanguard uses 3 of them as they net more mean mitigation than the other two will. Or is this statement referring to the other two tank classes that have a lower gain from shield rating?

 

As a note: I don't really get what is so hard to read about that data, there isn't really a way to display it other than a wall of numbers heh.

Edited by Redklaw
Link to comment
Share on other sites

The number are VERY possible to reach in a way that you reach a statistical parity, my numbers for 1712 posted earlier are my actual numbers with gear equipped and I use a spreadsheet to balance, so I am only a bit off (less that 20) on all the stats listed for 1700.

 

It really depends on the class you're talking about. Remember, Shield doesn't compete with Defense or Absorb through normal gearing means (re: everything except for augs) so, assuming you're stacking your mitigation and using high Shield enhs, you're going to have a functionally static amount of shield. You can't lower that shield to increase your absorb or defense, so, sometimes, the explicitly listed optimal ratio isn't really accessible without dropping your total stat budget (this is going to happen a lot with Guardians and Shadows). VG don't really have this problem because they want to stack the living hell out of their Shield.

 

Why would you never devote augments to shielding?

 

You can get all of the Shield rating you need by simply using high Shield enhancements (which you're probably using anyway to get to said high stat budgets). Since Shield only competes with Endurance (and Accuracy, but who cares about Acc on a tank) but Defense and Absorb directly compete with each other as well as Endurance, you should be auging for Defense and Absorb to reach the appropriate level of parity within your given stat budget.

Link to comment
Share on other sites

Well, to be fair, the PT bonus is 2% against everything, not just wait. I know i've shrugged off spits from TFB with it, so not certain how that weighs in there. I don't know how an extra 22 HPS from the shield stacks up to 2% chance to resist anything, but I would miss my two more seconds on invincible.

 

Also, I'd feel weird looking like my PT on both Chars.

Link to comment
Share on other sites

But even with full bastion or bulwark 27 enhancements (the highest 27 enhancements for mitigation I can find / make) PTs and VGs have to add Augments for shielding to hit stat optimization, even with 27B absorb / defense mods (unlettered variants just push the value of shield higher). There just isn't enough shielding available on our gear otherwise.

 

Take a look at the values in the OP again Kitru, the tech tanks have an unusually high optimal shield, an average absorb, and a very low defense value as their sweet spots.

 

Force tanks though, yeah I agree totally on that assessment.

 

Note: I forgot the numbers and source, but I remember someone worked or the 4 piece set bonus for jugs as being one of the best in game, will try to dig it up later or do independent math. I wouldn't touch the tech tank bonus.

Edited by Redklaw
Link to comment
Share on other sites


×
×
  • Create New...