Tanking Stat Weights

[SkySly]

I have question regarding VG/PT relic numbers. When you reach about 20% defense rating defense stacking pretty much stops and even decreases while absorb rises steadily. Yet according to your numbers defense relic is still better than absorption one. Shouldn't it be the other way around? Or am I missing something?

 

Second question is about trading mitigation for endurance. Is there any ratio or simple formula to calculate if this trade is worthwhile (I've seen 1:2,5 mentioned somewhere earlier in the thread)? For example swapping enhancement from arkanian to verpine would trade 10 points of mitigation for 21 point of endurance. I don't think it's really worth it, I just want to make sure.

[danosaurusrex]

Sorry for nub question. When we talk of the pvp relic, which one exactly are we meaning? I'm a jugg tank. Thanks in advance.

[Whizle]

Sorry for nub question. When we talk of the pvp relic, which one exactly are we meaning? I'm a jugg tank. Thanks in advance.

 

CMIIW, but I believe they mean the old static elite war hero defense relic (+120).

[Whizle]

I have question regarding VG/PT relic numbers. When you reach about 20% defense rating defense stacking pretty much stops and even decreases while absorb rises steadily. Yet according to your numbers defense relic is still better than absorption one. Shouldn't it be the other way around? Or am I missing something?

 

Second question is about trading mitigation for endurance. Is there any ratio or simple formula to calculate if this trade is worthwhile (I've seen 1:2,5 mentioned somewhere earlier in the thread)? For example swapping enhancement from arkanian to verpine would trade 10 points of mitigation for 21 point of endurance. I don't think it's really worth it, I just want to make sure.

 

Even if you've hit diminishing returns, the proc is so large that it usually adds 4-5% to your defense chance.

[dipstik]

I have question regarding VG/PT relic numbers. When you reach about 20% defense rating defense stacking pretty much stops and even decreases while absorb rises steadily. Yet according to your numbers defense relic is still better than absorption one. Shouldn't it be the other way around? Or am I missing something?

 

Second question is about trading mitigation for endurance. Is there any ratio or simple formula to calculate if this trade is worthwhile (I've seen 1:2,5 mentioned somewhere earlier in the thread)? For example swapping enhancement from arkanian to verpine would trade 10 points of mitigation for 21 point of endurance. I don't think it's really worth it, I just want to make sure.

 

the numbers are posted in the OP regarding the relics. if u want to know why all i can say is thats how the math turned out. the defense relic also has a static 32 defense that the absorb proc does not have.

 

the mitigation for endurance at 2.5 assumes you are a shadow/sin tank with self heals and damage is 4.5k/s. vanguards do not get enough benifit from their self heals to make thinking about trading mitigation for endurance seem reasonable. for juggs there is a strength/power to mitigation trade since they have self absorb, but i think it is too high to be relaized.

[dipstik]

yeah im talking about the elite war hero pvp relic. i updated the OP to make that more clear.

[danosaurusrex]

yeah im talking about the elite war hero pvp relic. i updated the OP to make that more clear.

 

OK, thanks. This fellow?

[dipstik]

Theres as shield one too

[Whizle]

Theres as shield one too

 

I think the question was, "which is better?" ...or are the numbers the same for the defense and shield relics?

[dipstik]

it doesnt matter as long as you stack you gear right, but vanguards like the shield one.

[yllesius]

The crafted vs UW implants/earring are a slightly more interesting decision. Going from UW to crafted you lose 54 End and 66 Str, but gain 26 Shield and 10 Defense. That's a fair amount of mitigation, however, I would still go with UW. If deciding between 54 End vs 36 Mitigation I would go with mitigation on Guardian, but the 66 Str tips the scale in UW's favor.

 

Curious about this as well. What you suggest for Shadow tanks, dipstik, is giving up 198 Willpower and 162 Endurance for 108 mitigation. Considering that we are already around 2,3-2,4k mitigation budget, would you still dump *so* much threat and HP?

 

I mean I am not able to give a solid % of additional mitigation you achieve with 108 mitigation at this DR, but according to your formula, you gain 11.664 hps with the additional 162 Endurance (and the number will be even slightly higher when including the talents + buffs percentage). Btw, is Battle Readiness included in that self-healing formula?

you get 0.072 hps per point of endurance

 

I would appreciate the mitigation % achieved at the 2.3k budget when dumping 162 Endurance for 108 mitigation. Threat loss is simply there.

Edited May 22, 2013 by yllesius
additional info

[yllesius]

I mean we still measure mitigation in %. Shouldn't we rather take every single hard mode boss fight into account, look at the *actual* DtPS, subtract the effective HPS and go from there? List the % mitigated by the tank on each boss, what was mitigated completely, what was shielded? Shadows can Resilience through some nasty sh*t which may *hugely* affect the results (Titan's grenade, Operation Chief's terminate etc.) and so will CDs of other tanks.

 

That would reflect the reality far more than just average numbers and it would also help new people way more than just theory.

 

Which btw doesn't really help when Thrasher hits you for 14k three times in a row because you were unlucky

(oh yes that spikiness, Kitru, huh..)

[AinuKvasir]

@dipstik This past weekend I put together an optimization site whose numbers seem to differ fairly substantially from yours. I was hoping you could take a look and see if we can figure out if there is a problem with my calculations or assumptions.

 

Site: https://www.tor-tank.com

Source to the Optimizer code: https://github.com/swtor-apps/swtor-tanking (the index.js file contains the raw data, and the src/swtor_tanking_optimizer.cc file has the actual calculation machinery)

[dipstik]

i have provided the slope and intercept of mitigation as a function of stat budget in the OP.

 

total d s a mit slope/intercept

2000 395 790 815 0.3356 -0.000039

2100 448 790 862 0.3313 0.41383

2200 502 790 908 0.3272

2300 555 790 955 0.3232 3.99002E-05

2400 608 790 1002 0.3193 3.85993E-05

2500 661 790 1049 0.3156 3.73456E-05

2600 714 790 1096 0.3120 3.61371E-05

 

the difference between 2300 and 2200 is 0.004. at 4500 dps that gives 18 more mitigation. 162 endurance gives 11.7 hps.

 

if there is a bonus healing term in the self heals or combat technique, this might make underworld better, but im not sure that bonus healing is anywhere in those values.

 

also, trying to make a thread that speaks to every tank combination and every boss and when to use every cd is outside of the scope of this thread. i think the pre mitigation damage types and amounts have been extremely helpful for determining which assumptions should go into these equations...

 

 

and as i have stated many times, i do not take long cooldowns, like battle readiness, into account for these formula.

 

as far as that web app: im not sure. kbn has a much higher tolerance for parsing through code than me, so i suggest you try there. the biggest difference is the same between KBN and I, which is the shield rating assumption. using your value of shield, which is set to the minimum of .364*budget, and optimizing for defense and absorb i get your values, for shadow anyway.

Edited May 22, 2013 by dipstik

[AinuKvasir]

Thanks for the reply dipstik. I looked again through your post and I can't see where you list your bounds on shield as a percent of budget. Should it not be .364? Also, in modelling relics, I have been using a swing time of 1.2s. In your post, I think your list S&V's average swing time as 0.87s. Am I interpreting that correctly?

[dipstik]

i set shield to 790 (kbn programs a bound on shield i dont), but in AMR profiles where shield is different form this value i calc out what is best and try to get gear as close as possible to optimal.

 

the swing timer set as 0.87 was from kbn analysis.

 

you might want to report the unbound results so peopel can see what you should do if it was possible. since people can use 2 elite war hero relics, their stat budget will be high, and due to your assumption on shield they will be using shield that they should not be using.

Edited May 22, 2013 by dipstik

[fire-breath]

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%.

 

Maybe I way off, but shouldn't the heatblast absorb be (rounded up) 18%?

 

I assume the following:

- every 1,5 seconds you either refresh or gain a stack of powerscreen

- giving you an effective cd of 9 seconds on heatblast

- Bioware calculates absorb at least in amounts of 0,5 seconds

 

If above is true then I get the following values after activating hb:

25% - 0 seconds

25% - 0,5 seconds

25% - 1 second

26% - 1,5 second (1 stack of powerscreen)

26% - 2 sec

26% - 2,5 sec

27% - 3 sec (2nd stack of ps)

 

And so on (you got the idea (and if I'm correct you assume the same trend))

 

This turns into:

(3*7+25*3+26*3+27*3+28*3)/19= 17,68%

 

I think the difference is that you assume an effective cd of 12 seconds, forgetting the cd-reduction of refreshing a stack.

 

I'm sorry if I used more words then needed. I wanted to be sure that the math I use for this is the correct one.

Edited May 26, 2013 by fire-breath

[dipstik]

i thought the cd was 15 seconds and it was brought down to 12

 

you are saying it is 12 and brought down to 9?

[fire-breath]

Nope, 15 to 9.

 

At 3 stacks the stacks will get refreshed and the cd diminishingbonus should peel off the other seconds. Unless its bugged it should be 15 to 9.

 

Or I'm wrong, that could be to.

 

EDIT:

Its even weirder. You gain a bonus second each time when you either refresh or gain a powerscreenstack. In an 'optimal' situation where you get hit my many adds (ergo, you shield a lot) then the cd is 9 seconds as shown in the ytube link below:

- cd starting on 0:08 - off cd at 0:18 (effective cd of 10 seconds)

- cd starting on 0:20 - off cd at 0:29 (effective cd of 9 seconds)

- cd starting on 0:31 - off cd at 0:41 (effective cd of 10 seconds)

- cd starting on 0:44 - off cd at 0:52 (effective cd of 8 seconds)

- cd starting on 0:53 - off cd at 1:02 (effective cd of 9 seconds)

and so on ....

 

Youtube link http://youtu.be/n0tvOWhT7NI

 

There is one big but .... I was molested in this vid by 3 mobs, each containing about 3-5 enemies. In boss situations you mostly will have 1 boss shooting at you.

Sinse my tank isnt my main I havent done many opses yet, so I never came around to vid one.

 

I will try it lateron against a solo add thats strong enough for me to survive attacks other then hammershot for more then a few seconds

 

One hint I got by attacking a golden is that it gave me 12 seconds cd (without me attacking it besides the energyblast).

Edited May 26, 2013 by fire-breath

[Yngow]

Nope, 15 to 9.

 

At 3 stacks the stacks will get refreshed and the cd diminishingbonus should peel off the other seconds. Unless its bugged it should be 15 to 9.

 

Or I'm wrong, that could be to.

 

EDIT:

Its even weirder. You gain a bonus second each time when you either refresh or gain a powerscreenstack. In an 'optimal' situation where you get hit my many adds (ergo, you shield a lot) then the cd is 9 seconds as shown in the ytube link below:

- cd starting on 0:08 - off cd at 0:18 (effective cd of 10 seconds)

- cd starting on 0:20 - off cd at 0:29 (effective cd of 9 seconds)

- cd starting on 0:31 - off cd at 0:41 (effective cd of 10 seconds)

- cd starting on 0:44 - off cd at 0:52 (effective cd of 8 seconds)

- cd starting on 0:53 - off cd at 1:02 (effective cd of 9 seconds)

and so on ....

 

Youtube link http://youtu.be/n0tvOWhT7NI

 

There is one big but .... I was molested in this vid by 3 mobs, each containing about 3-5 enemies. In boss situations you mostly will have 1 boss shooting at you.

Sinse my tank isnt my main I havent done many opses yet, so I never came around to vid one.

 

I will try it lateron against a solo add thats strong enough for me to survive attacks other then hammershot for more then a few seconds

 

One hint I got by attacking a golden is that it gave me 12 seconds cd (without me attacking it besides the energyblast).

 

The least you can get is 12 seconds, and that is from 2 rocket punches and 1 rail shot, since both will always trigger your ion gas cylinder and based on their cooldowns (9s and 15s not to mention the proc that refreshes rocket punch), that is why I never use rocket punch or rail shot if heat blast is available.

You get heat blast every time you shield and attack or activate your ion gas cylinder, but no more than once every 1.5s

[Kitru]

The least you can get is 12 seconds, and that is from 2 rocket punches and 1 rail shot, since both will always trigger your ion gas cylinder and based on their cooldowns (9s and 15s not to mention the proc that refreshes rocket punch), that is why I never use rocket punch or rail shot if heat blast is available.

You get heat blast every time you shield and attack or activate your ion gas cylinder, but no more than once every 1.5s

 

Mathematically, the least you can get is 9 seconds: assuming you get 1 proc *every* 1.5 seconds with no downtime (whether from getting lucky with an attack or a shield attempt), you'll have it at 9 seconds (15/2.5x = 6 GCDs; 6 * 1.5 = 9 seconds).

 

Assume a boss swing timer of 1 attack/GCD and BiS itemization (42% Shield chance), you have a 50% chance per GCD to reduce the CD by 1 second via shielding, and are guaranteed one proc every 15 seconds from HiB, one proc every ~6 seconds from Stockstrike (Static shield with 42% Shield chance and 19% defense chance equates to 34.02% chance to get a shield on an individual M/R attack; with 50% proc chance, the chance per attack is equal to 17.01%; since the math gets fuzzy with the interaction between multiple ICDs, I just wrote a sim; it was simpler), with a 15% chance with any attack other than those two (Hammer Shot is *slightly* better, but it would need to be weighted to determine the effect of the increased chance), all of which cannot occur simultaneously. This means that the average proc rate on a Power Screen proc is ~.45 per second (1/6 + (1-1/6)*1/15 + (1 - (1/6 + 1/15)) * (.1701 + (1-.1701)*.15)). Add that to the standard CD reduction you get per second anyways, and the use rate on Energy Blast is going to be ~10.3 seconds (15 / (1 + .45)).

 

Thus, with a 10.3 second interval and a 6 second duration, you get ~58% uptime on 25% Absorb for ~14.6% average Absorb over time from Energy Blast, and ~1.73% Absorb from Power Screen (((10.3 - 2.175) + (10.3 - 2.175 * 2) + (10.3 - 2.175 * 3)) / 10.3; each stack has its uptime and average contribution calculated separately so that the sum total uptime can then be divided by the Energy Blast interval to determine the average contribution).

Edited May 27, 2013 by Kitru
Stupid GCD<->second conversions >.<

[WillLongstick]

Mathematically, the least you can get is 6 seconds: assuming you get 1 proc *every* 1.5 seconds with no downtime (whether from getting lucky with an attack or a shield attempt), you'll have it at 6 seconds (15/2.5x = 6)..

 

Units check. 15 seconds divided by 2.5 seconds per GCD gives you an output of 6 GCDs for your quickest Heat Blast proc. 6 GCDS * 1.5 seconds per GCD = 9 seconds.

[Kitru]

Units check. 15 seconds divided by 2.5 seconds per GCD gives you an output of 6 GCDs for your quickest Heat Blast proc. 6 GCDS * 1.5 seconds per GCD = 9 seconds.

 

This is what I get for alternating between swapping between GCD based models and second based models. Now I gots to check over all that math again to make sure I didn't screw up any other GCD>second or second>GCD conversions. >.<

[Yngow]

what I meant with "least you can get" is, assuming you are in a scenario where you are not shielding any attacks and not getting lucky with ion cylinder procs, so the only thing reducing the cd is the rocket punch and rail shot. In other words what I meant was the higher cd timer possible, sorry for misleading

[dipstik]

Let me know. I need to update stuff for new gear anyway.