Please upgrade your browser for the best possible experience.

Chrome Firefox Internet Explorer
×

Mean Mitigation


Jethsidi's Avatar


Jethsidi
02.17.2015 , 10:11 AM | #21
Quote: Originally Posted by Methoxa View Post
You are aware that your Juggernaut defstats are totally different to KBN's numbers?
Very much aware. Current stat pool differential he is about 500 Defense rating higher than I am. However at 1094 I have a Defense Chance of 25%. Which on Defense-heavy fight has me deflecting/parrying 47-53% of incoming damage. I found that pushing the percentage higher resulted in a negligible difference in how much I actually defended in these fights (on both my Juggernaut and Assassin). However, putting those 500 points in Shield drastically reduced the amount of damage I take on Shield/Absorb fights. One could make the argument that my Shield is too high and that I need to put some into Absorb, but I like my 50% chance and 44% Absorb seems to be adequate for Juggernauts.

Don't get me wrong, KBN has solid math, and his breakdown will no doubt give you the most bang for your buck in terms of raw percentages given the selected stat budget (meaning very little DR across all 3 stats), but that is not how operations bosses usually work. Each tank is different and therefore you will need to push into the DR on certain stats to achieve mean mitigation (the whole theory this is based on), otherwise you would need to be switching which tank you used based on the fight because they are tailored to different things. Juggernauts push in the Shield DR, Assassins go into Absorb, and my current PT configuration is pushing into Defense, but I will be re-assessing that after chatting with justinplainview (probably not going to drop all of it, but I may take it down a couple hundred points).

P.S. I prefer to have 1 set that I can use in all fight situations, but if you are one to have different gear sets for different fights you wouldn't be looking at KBN's stats either.
Don't ever learn how to tank.
They won't let you do anything else once they know you are good at it.

KeyboardNinja's Avatar


KeyboardNinja
02.17.2015 , 12:55 PM | #22
Quote: Originally Posted by Jethsidi View Post
Don't get me wrong, KBN has solid math, and his breakdown will no doubt give you the most bang for your buck in terms of raw percentages given the selected stat budget (meaning very little DR across all 3 stats), but that is not how operations bosses usually work. Each tank is different and therefore you will need to push into the DR on certain stats to achieve mean mitigation (the whole theory this is based on), otherwise you would need to be switching which tank you used based on the fight because they are tailored to different things. Juggernauts push in the Shield DR, Assassins go into Absorb, and my current PT configuration is pushing into Defense, but I will be re-assessing that after chatting with justinplainview (probably not going to drop all of it, but I may take it down a couple hundred points).
I think you're misunderstanding how my math works. I'm not optimizing for "minimum effect from diminishing returns"; I'm optimizing for "minimum damage taken". Diminishing returns are a factor in every serious calculation of this sort, but they are far from the dominant factor. For example, if you have a boss where the only non-internal/elemental damage force/tech (e.g. Soa is the classic example of this, but more recent close calls are Brontes and Corrupter Zero), then it doesn't matter what the diminishing returns on shield/absorb vs defense are, defense is still useless! My numbers consider the actual damage ratios in each boss fight and optimize for minimizing total damage taken, respecting tank-specific bonuses (e.g. Blade Barricade, Dark Ward, etc) and diminishing returns.

Just looking at a log and saying "well, I guess 10 more points of defense doesn't increase my dodge rate very much" is insufficient. It's too small of a sample size (RNG is going to blow away any specific results from just 10 stat points), and it's failing to take into account the full tradeoff picture. Even if adding 10 points of defense doesn't make you dodge that much more often, how much better is your mitigation going to be adding those 10 points in absorb instead? (remember to scale by your shield chance!) This is why tank mitigation is interesting from a theory crafting standpoint, since the results are quite hard to see when you get down to the level of shuffling a few points around, but they're still very quantifiable in the math.

I'm not saying that you can't clear content with the stat distributions you list. Tank mitigation is a relatively small component of a raid group's total competency (a decent healing pair will make bad itemization nearly irrelevant, and a bad healing pair will render good itemization insufficient). What I'm saying is that your definition of "optimal" is too shallow, and your raid results will be strictly better following a different gearing strategy.

To be clear, I'm not claiming that my numbers are gospel. There are plenty of quantitative reasons to use different stat distributions, particularly in very specific situations or during progression on a certain boss. What my numbers are is optimally tuned with respect to a very quantifiable and measurable set of criteria, and this is ultimately what matters with tank gearing. As long as you can very precisely define what you want out of your gear (more precisely than "MOAR TANK"), then you're in good shape already. My problem with the OP is that it doesn't precisely define this.

Quote: Originally Posted by Jethsidi View Post
P.S. I prefer to have 1 set that I can use in all fight situations, but if you are one to have different gear sets for different fights you wouldn't be looking at KBN's stats either.
Plenty of people actually do look at my numbers when optimizing for specific fights. This is why I've traditionally had "low/mid/max-MR" profiles in addition to the average. Those profiles aren't up right now because of the forum intern mistake, but they're coming back when I restore the post.
Computer Programmer. Theory Crafter. Dilettante on The Ebon Hawk.
Tam (shadow tank) Tov-ren (commando healer) Aveo (retired sentinel) Nimri (ruffian scoundrel)
Averith (marksman sniper) Alish (lightning sorcerer) Aresham (vengeance jugg) Effek (pyro pt)

December 13, 2011 to January 30, 2017

Jethsidi's Avatar


Jethsidi
02.17.2015 , 01:39 PM | #23
@KBN There was a lot in there so I will try to write my replies in the order you listed things.

I think you're misunderstanding how my math works. I'm not optimizing for "minimum effect from diminishing returns"; I'm optimizing for "minimum damage taken".
- To clarify, I am not doubting that is how you optimize your math, simply stating when I have geared characters according to it the DR comment is what I noticed. I have no problem with your numbers, and I started all my tanks off with them and then shifted stats to better suit how I perceived I was taking damage.

For example, if you have a boss where the only non-internal/elemental damage force/tech (e.g. Soa is the classic example of this, but more recent close calls are Brontes and Corrupter Zero), then it doesn't matter what the diminishing returns on shield/absorb vs defense are, defense is still useless! My numbers consider the actual damage ratios in each boss fight and optimize for minimizing total damage taken, respecting tank-specific bonuses (e.g. Blade Barricade, Dark Ward, etc) and diminishing returns.
- I'm going to couple this with your statement about "10 additional points". If I read this correctly you are stating that if the fight is Shield based it doesn't matter how far into the DR you are it is still better than putting points in Defense. I completely agree. As we all know different fights have different damage profiles. This can be solved by having different gear sets optimized to different fights and switch between them. This is something I prefer not to do. I have my Juggernaut in a place where I am confident his gear can handle any fight. The principle of "more Defense will help you in this fight than Shield" based on damage type is what I was referring to when I mentioned that I did not see and increase in mitigation by going higher than 25% Defense chance. It literally didn't matter either way. You then mention adjusting values by 10 points, and again I agree. A 10 point stat difference is negligible, but we are talking in the hundreds of points stat difference. Five hundred point in Shield vs. 500 points in any other stat will be a noticeable difference, whether good or bad.

Tank mitigation is a relatively small component of a raid group's total competency (a decent healing pair will make bad itemization nearly irrelevant, and a bad healing pair will render good itemization insufficient).
- I doubt anyone is going to argue that point.

What I'm saying is that your definition of "optimal" is too shallow, and your raid results will be strictly better following a different gearing strategy.
- This is where you have missed the original point entirely. I never said these numbers were "optimal", this thread was created based on a theory where you can attain the same mitigation numbers across all three classes. As most PT's in the game will tell you, they don't run high defense. I do and I don't have any problems with it, but the PT numbers listed are simply there to show you can achieve the same percentages (roughly) as the Juggernaut/Assassin. As far as your definition of "better" raid results, I have to disagree. My average DTPS is between 1200 and 1500 on most fights, never going about 1600 unless I am completely messing something up (obvious exceptions include Master/Blaster and I am sure Revan). During fights where I am just "on" for some reason my numbers go down to just above 1k and if you throw in a Sorc healer that knows what they are doing they can be even lower. Gearing by your numbers would, granted, provide different results, but I maintain that "better" is a point of view at this level of discussion.

What my numbers are is optimally tuned with respect to a very quantifiable and measurable set of criteria, and this is ultimately what matters with tank gearing.
- Last update on your numbers that I saw was posted in 2013. If you have updated numbers based on the new operations' damage profiles I would very much like to see them.
Don't ever learn how to tank.
They won't let you do anything else once they know you are good at it.

KeyboardNinja's Avatar


KeyboardNinja
02.17.2015 , 02:21 PM | #24
Quote: Originally Posted by Jethsidi View Post
The principle of "more Defense will help you in this fight than Shield" based on damage type is what I was referring to when I mentioned that I did not see and increase in mitigation by going higher than 25% Defense chance. It literally didn't matter either way. You then mention adjusting values by 10 points, and again I agree. A 10 point stat difference is negligible, but we are talking in the hundreds of points stat difference. Five hundred point in Shield vs. 500 points in any other stat will be a noticeable difference, whether good or bad.
25% is an incredibly arbitrary line in the sand. What if I said 26%? 24%? What are your objections to those points instead? What if I gave you a stat budget of 8k? Would you still refuse to go above 25%?

How did you measure "mitigation" when you were looking for increases? Clearly not just character sheet, since your character sheet would have shown exactly how much mitigation you gain from defense vs dumping those points in shield or absorb. If you were looking at combat logs, how many did you examine? Remember, RNG washes this stuff out. Even a swing of 500 stat points (still only 12.5% of the total budget!) is going to be hard to pick out from the noise, especially if you're only looking at one or two logs.

Quote: Originally Posted by Jethsidi View Post
This is where you have missed the original point entirely. I never said these numbers were "optimal", this thread was created based on a theory where you can attain the same mitigation numbers across all three classes. As most PT's in the game will tell you, they don't run high defense. I do and I don't have any problems with it, but the PT numbers listed are simply there to show you can achieve the same percentages (roughly) as the Juggernaut/Assassin.
…why would you want to? You're wasting a lot of points to diminishing returns on defense, trying to make your percentages reach the same levels that a Jugg gets nearly for free with their baseline class bonuses. You're also failing to take advantage of your own bonuses (as a PT). Remember that stat ratings increase percentages additively with intrinsic bonuses, such as Power Screen or Heat Blast. This means that, the bigger the intrinsic bonus for a particular pool, the better that stat becomes. This is why Juggs stack so much defense: they have very large intrinsic defense bonuses.

Quote: Originally Posted by Jethsidi View Post
As far as your definition of "better" raid results, I have to disagree. My average DTPS is between 1200 and 1500 on most fights, never going about 1600 unless I am completely messing something up (obvious exceptions include Master/Blaster and I am sure Revan).
Have you tanked HM Sword Squadron Unit 1? Heck, even tanking Unit 2, I don't know if I've ever seen a log going below 1600 unless the raid group in question was outright ignoring the tank swap (thus, offloading damage onto Unit 1). Tanking Unit 1 and getting less than 1.7k DtPS would be insane. Even with objectively perfect positioning (only unavoidable damage taken), precisely optimized cooldown usage and some 198 gear, assassin tanks (who take the least damage on that fight) are getting about 1.7k.

When you tank HM Torque, do you manage the turrets correctly, or do you allow them to chew on the raid? A corollary to this would be to ask how much damage your melee DPS are taking on that fight, since correct turret tanking represents a swing of about 700 DtPS in a mDPS's log. Tanking HM Torque while correctly managing the turrets is going to push your DtPS up pretty close to 2k, regardless of class (though it should be lowest on a PT).

My point is not to criticize you as a player or cast any doubt on your expertise, but your numbers simply do not fall within the realm of plausibility based on what I've seen, both in my own logs and in the logs of many other tanks in other groups.

Basically the only thing I can think of that would drop your DtPS down to the levels you're listing is if you're a Jugg (Sonic Barrier) and your group is using a Sorc healer who keeps you bubbled on CD. This artificially deflates DtPS, since you're effectively getting healed without the original damage appearing in the log summary.

Quote: Originally Posted by Jethsidi View Post
Gearing by your numbers would, granted, provide different results, but I maintain that "better" is a point of view at this level of discussion.
Gearing by my numbers will reduce your DtPS to a minimum. If there is an alternative itemization which reduces DtPS by any amount below what my numbers generate, then something is wrong with my math.

You're right though that "better" is a point of view. My point is that you need to be very precise when you say "better". I don't see anywhere in the OP where you define what it is you're looking for exactly. If you can't quantify your goal, then you have no way to objectively pursue it.

Quote: Originally Posted by Jethsidi View Post
Last update on your numbers that I saw was posted in 2013. If you have updated numbers based on the new operations' damage profiles I would very much like to see them.
The post was deleted. I've updated the numbers four times since 2013, including once in the week following the release of 3.0. I haven't updated for the new operations yet (though I do have 10/10 HM logs, thanks to some members of the community) mostly because I've been absurdly busy IRL.
Computer Programmer. Theory Crafter. Dilettante on The Ebon Hawk.
Tam (shadow tank) Tov-ren (commando healer) Aveo (retired sentinel) Nimri (ruffian scoundrel)
Averith (marksman sniper) Alish (lightning sorcerer) Aresham (vengeance jugg) Effek (pyro pt)

December 13, 2011 to January 30, 2017

Jethsidi's Avatar


Jethsidi
02.17.2015 , 04:06 PM | #25
See if this quoting stuff works out, never tried it with multiple parts before.

Quote: Originally Posted by KeyboardNinja View Post
25% is an incredibly arbitrary line in the sand. What if I said 26%? 24%? What are your objections to those points instead? What if I gave you a stat budget of 8k? Would you still refuse to go above 25%?
Twenty-six or 24% have no objections, 25% does not mean 25.00000000000%. Current accuracy BiS is 760, but at the 198 level with 5 pieces and 3 augments you are sitting at 756 which is equal to 99.99%, which is effectually the same. Then you keep mentioning I need a "goal" in mind, 25% is a goal. If you go over or under it will have different effects, but as you are stating a couple percentage points is not earth-shattering. Plus, the percentages are listed to show how equal they can be, the stat budget is what truly matters.

What if stat budget of 8k? - Really? That's your argument? Obviously with different stat pools comes different math. Because with different stat pools BW/EA are going to change the DR values (Crit is a perfect example of this).

Quote: Originally Posted by KeyboardNinja View Post
…why would you want to? You're wasting a lot of points to diminishing returns on defense, trying to make your percentages reach the same levels that a Jugg gets nearly for free with their baseline class bonuses. You're also failing to take advantage of your own bonuses (as a PT). Remember that stat ratings increase percentages additively with intrinsic bonuses, such as Power Screen or Heat Blast. This means that, the bigger the intrinsic bonus for a particular pool, the better that stat becomes. This is why Juggs stack so much defense: they have very large intrinsic defense bonuses.
Why? Really? The first point I make is -

Quote: Originally Posted by Jethsidi View Post
Let me preface this by saying this guide is based on an average of mitigation across all 3 tanking classes. Meaning if you add up your armor, stats, and health all 3 classes should have roughly the same DTPS per fight (obvious variances not included), achieved by having similar stat percentages. If this is NOT a style of tanking that you believe is viable you can save your time, you will not agree with anything I am about to say.
You also fail to see the logic behind what is being shown. The point I am trying to make is that is it not about what tank you are currently playing, you should be able to achieve the same amount of survivability, before factoring in cooldowns, based on your stats alone. If you go through an entire tier of operations as an Assassin with your Defense rating in the 700's, Shield in the 1400's and Absorb in the 1700-1800's and your survivability is exactly where you want it you can then look at your percentages and try to achieve the same on the other classes, same goes for if you started on a Juggernaut or Powertech, although your stat distribution would be different based on specific DR's to that class (obvious difference is Assassin's have light armor and therefore take more base damage than the other 2, which is balanced by the fact they have roughly the same percentage difference higher in their Absorb rating). Or one could just say 'mean mitigation', it is simpler.

And unless I am reading the rest of it wrong (based on your use of 'intrinsic bonuses' and 'better'), I completely disagree with you. Let's take the PT - in your discipline you have bonuses that give you 4% Absorb, 4% Defense. and 2% Shield. You also have Heat Screens, Heat Blast, and Shield Enhancers - these are all flat percentages and none of them care how much of that stat you currently have, they will give you only the flat percentage listed. Could I go up on Shield/Absorb on my PT? yeah, but I could also defend 5% more of the time (on applicable fights). Given the shear amount of up-time I have on Heat Blast's 25% buff I feel my points are better spent elsewhere. Juggernauts stacking Defense - ok, cool, you can hit 30% (ish) with your buffs, but your Shield/Absorb ratings are going to get you annihilated in fights that have very little to defend.

Quote: Originally Posted by KeyboardNinja View Post
Gearing by my numbers will reduce your DtPS to a minimum. If there is an alternative itemization which reduces DtPS by any amount below what my numbers generate, then something is wrong with my math.
Did you not also say -

Quote: Originally Posted by KeyboardNinja View Post
To be clear, I'm not claiming that my numbers are gospel.
And yet now you are saying your numbers are the best possible numbers in the game to be strived for and achieved? And that the only better mitigation attainable would be if you made an error?

Quote: Originally Posted by KeyboardNinja View Post
The post was deleted. I've updated the numbers four times since 2013, including once in the week following the release of 3.0. I haven't updated for the new operations yet (though I do have 10/10 HM logs, thanks to some members of the community) mostly because I've been absurdly busy IRL.
Ok, then why does this post - http://www.swtor.com/community/showthread.php?t=790085 - by you 3 days ago list to the number the exact same stat differentials as this post - http://www.swtor.com/community/showthread.php?t=792434 - which quotes you from 2013? I fail to see how that is updated.
Don't ever learn how to tank.
They won't let you do anything else once they know you are good at it.

KeyboardNinja's Avatar


KeyboardNinja
02.17.2015 , 04:46 PM | #26
Quote: Originally Posted by Jethsidi View Post
What if stat budget of 8k? - Really? That's your argument? Obviously with different stat pools comes different math. Because with different stat pools BW/EA are going to change the DR values (Crit is a perfect example of this).
When 2.0 dropped, a reasonable stat budget was 2k. Now a reasonable stat budget is 4k. The DR values have not changed since then, only the level (55 to 60).

My argument was simply showing how arbitrary the "25%, no more" claim is. I mean, if you have some extremely specific justification for this (e.g. there's this one attack on this one boss that I want to dodge exactly 1 out of 4 times), then I respect that. Just saying "I don't feel like going higher is beneficial" is insufficient.

Quote: Originally Posted by Jethsidi View Post
You also fail to see the logic behind what is being shown. The point I am trying to make is that is it not about what tank you are currently playing, you should be able to achieve the same amount of survivability, before factoring in cooldowns, based on your stats alone.
I would love it if that were the case. It's not though. Powertechs take more damage than the other two tanks, even before factoring in cooldowns. Part of this is damage type (e.g. Revan absolutely murders PTs in the DtPS department), and part of it is simple class balance. Heck, Juggernauts are hilariously OP for any fight where the pre-mitigation DtPS is particularly low (e.g. off-tanking Underlurker) and will take significantly less DtPS than the other two tanks, due to the non-scaling nature of Sonic Barrier.

Tank balance is an approximation. A very, very close approximation, but an approximation nonetheless. Even above and beyond this, you can see very dramatic differences from fight to fight. Properly itemized Juggs are going to take less damage on Bulo than any other tank by a fairly wide margin, even completely discounting cooldowns. Powertechs will take the least on Torque, again discounting cooldowns. Properly itemized assassins are going to take the least damage on most fights, but it will be particularly noticeable on fights like Revan (so basically, Revan).

The goal should be optimize for the tank you have, rather than trying to set up the mean mitigation such that it's perfectly equivalent on every fight (since you can't get there!).

Quote: Originally Posted by Jethsidi View Post
And unless I am reading the rest of it wrong (based on your use of 'intrinsic bonuses' and 'better'), I completely disagree with you. Let's take the PT - in your discipline you have bonuses that give you 4% Absorb, 4% Defense. and 2% Shield. You also have Heat Screens, Heat Blast, and Shield Enhancers - these are all flat percentages and none of them care how much of that stat you currently have, they will give you only the flat percentage listed. Could I go up on Shield/Absorb on my PT? yeah, but I could also defend 5% more of the time (on applicable fights). Given the shear amount of up-time I have on Heat Blast's 25% buff I feel my points are better spent elsewhere. Juggernauts stacking Defense - ok, cool, you can hit 30% (ish) with your buffs, but your Shield/Absorb ratings are going to get you annihilated in fights that have very little to defend.
This isn't really a matter of opinion that either of us can "agree" or "disagree" on. We're talking about objective math here. Exact formulae:

1 - (1 - (0.1 + p)) (1 - 0.3) < 1 - (1 - 0.1) (1 - (0.3 + p))

It's pretty trivial to algebraically show that this inequality holds for all p > 0. What is being empirically demonstrated by the above is that a larger intrinsic bonus (e.g. PTs have more intrinsic shield/absorb than defense; Juggs are the inverse) implies that you are better off statting even more into that stat bucket.

You're very welcome to disagree that DtPS is the best thing to optimize for. Heck, even I disagree on that point, despite the fact that my somewhat boldly titled "ideal tank stats" optimizes for precisely and only DtPS. However, disagreeing on algebra itself isn't really an option. What you are claiming when you say that the points are "better spent elsewhere" is that you believe the above inequality does not hold, and this is an objectively false belief.

tl;dr: Additive bonuses are awesome (for tanks) and must be leveraged by any optimization methodology. Ignoring them in favor of thresholds and quotas is objectively sub-optimal by any standard.

Quote: Originally Posted by Jethsidi View Post
And yet now you are saying your numbers are the best possible numbers in the game to be strived for and achieved? And that the only better mitigation attainable would be if you made an error?
I'm drawing a distinction between "best" and "produces lowest DtPS". My numbers are not "gospel" in that they are not guaranteed to be the "best" way to gear your tank, whatever that means. My numbers are the numbers which produce the lowest DtPS though, barring errors in my calculations. I would absolutely welcome corrections to the math if and where there are errors, but the whole definition of my "ideal tank stats" problem space is to reduce DtPS to a minimum. If there is a different set of numbers for the same stat budget which objectively produces lower DtPS for the same damage distribution as my numbers, then something is wrong with my math, because my math specifically seeks out this minimum.

Are my numbers the "best in the game"? No. I don't know how you or your raid group define 'best'. Even I have a hard time defining 'best' where tank gear is concerned. Do my numbers produce the highest mean mitigation possible? Yes.

Quote: Originally Posted by Jethsidi View Post
Ok, then why does this post - http://www.swtor.com/community/showthread.php?t=790085 - by you 3 days ago list to the number the exact same stat differentials as this post - http://www.swtor.com/community/showthread.php?t=792434 - which quotes you from 2013? I fail to see how that is updated.
It's not quoting me from 2013. The author of that post copy/pasted from a web archive of my post that included the original post date. If you look at your own OP, you'll notice that the date shown by the forums is the date and time you originally posted, not the date and time of your last edit.

You also might take the time to notice the change log section in the very next paragraph, which lists Nightmare Dread Fortress, an operation that did not exist in 2013.

As a final note, the web archive used by the author of the post you linked did not have a fully up to date copy of my post. So he used the web archive to grab an old version (it looks like from last fall), and then he swapped out the tables with the ones I re-posted here. This is why the numbers are equal.
Computer Programmer. Theory Crafter. Dilettante on The Ebon Hawk.
Tam (shadow tank) Tov-ren (commando healer) Aveo (retired sentinel) Nimri (ruffian scoundrel)
Averith (marksman sniper) Alish (lightning sorcerer) Aresham (vengeance jugg) Effek (pyro pt)

December 13, 2011 to January 30, 2017

Jethsidi's Avatar


Jethsidi
02.17.2015 , 09:12 PM | #27
Cooldowns notwithstanding.
The game truly does not care which class of tank you bring to a fight. The only thing that matters is the damage and damage types going out versus the mitigation statistics that are trying to counter them. There is no "more out of stacking [any stat] intrinsically" on any tank. The stats you put in give you a raw percentage that is then added with flat percentages in your Discipline or the passives based on that class. The DR is exactly the same across each class.

Defense Chance = 30 * { 1 – [ 1 - ( 0.01 / 0.3 ) ]^[ ( DefenseRating / 60) / 1.2 ) ] }
Shield Chance = 50 * { 1 – [ 1 - ( 0.01 / 0.5 ) ]^[ ( ShieldRating / 60) / 0.78 ] }
Absorb Percentage = 50 * { 1 – [ 1 - ( 0.01 / 0.5 ) ]^[ ( AbsorptionRating / 60 ) / 0.65 ] }
*60 is the dividing factor in the exponent because it is current max level.

Regardless of class plug in your number and you will get the percentage it is contributing.

https://onedrive.live.com/redir?resi...hint=folder%2c

I did screenshots to show that the values are identical coming off of your raw stat pool. (Shield is the exact same value, Absorb and Defense are off by 2 stat point but you can see the results are the same regardless of class)

Next you have to look at what the classes offer:

Powertech
Discipline: Shield +2%, Dmg Red + 4%, Armor +15%, M/R Def +4%
Cylinder: Armor +60%, Dmg Red +5%, Shield +15%
Active Mitigation: Shield +3%, Absorb +3-28%

Assassin
Discipline: Armor +20%, Dmg Red +4%, M/R Def +6%, Absorb +4%
Stance: Armor +130%, Shield +15%, I/E Dmg Red +10%
Active Mitigation: Shield +15%, Absorb +1-8%, Dmg Red 1-4%

Juggernaut
Discipline: Armor +15%, Shield +4%, I/E Dmg Red – 5%
Stance: Armor +60%, Dmg Red +6%, Shield +15%, Defense +3%
Active Mitigation: Defense +5%, Dmg Red +3%, Absorb +3%, Sonic Barrier

If you have the logs your self or look at someone else's you can find the damage profiles for each fight, and, based on your class's inherent buffs + the stat equations find the best way to get your stats where the need to be. You can do this on a fight-to-fight basis or you can take an average across all fights for a single gear set.

The fact that the stat pool gives the same percentages and the class buffs are flat percentages means that if you find a percentage distribution that works for one tank, attaining the same percent levels on the other 2 can attain the exact same mitigation (to re-iterate: cooldowns notwithstanding). I also find it odd that you state PT's will always take more damage, considering in the screenshots you can see that their passive dmg red buffs are ahead of the Juggernaut in both respects, and ahead of the Assassin in terms of E/K damage.

The reason I go high on shield with PT/Jugg is that at 1200 rating (base in 198's) the Assassin is already at 55.22 without doing anything. So, based on the premise I am trying to get as close as I can to that, but 50% takes a bit much to get to so I stop there and start putting points into other stats. Because my Juggernaut and PT put so much into Shield, their absorption suffers and they end up in the mid-40th percentile. Which is balanced by the fact that they simply take less damage than the Assassin would with the same percentage. Which balances around to Assassin's stacking Absorb (which they have been doing for a very long time). After balancing Shield/Absorb the mean Defense percentages I was achieving with the remaining stat pool was roughly 25%, roughly - I honestly don't know why you are so hung up on that number - after seeing this I balanced it a bit further to make the percentages a bit more exact (i.e. the 25% and 50% respectively), can you really blame me for liking even numbers? Granted, it may not be as *perfectly* mitigating as 25.42% and 49.06%, but we have already touched base on the differences changes that small make.

That was the whole premise of Mean Mitigation. Because I have found that it *does* work (the only true exception to this is the Assassin's inherent +10% I/E damage reduction). The only variance in my damage profiles across my tanks is based on what cooldowns are good in which fights. Which is why I honestly prefer my Juggernaut, as they have at least one cooldown that works against every damage type.
Don't ever learn how to tank.
They won't let you do anything else once they know you are good at it.

Jethsidi's Avatar


Jethsidi
02.17.2015 , 10:15 PM | #28
But truly the end point is this: we have both stated that "optimal" and "best" are points of view based on what you are gearing towards. The OP was never meant as a challenge to your math statistics, it was simply as statement of this is how my tanks are geared and it works. The reason I posted is that aside from your math most of the other posts containing specifics on how to gear tanks are very debatable, especially if you read the ones concerning Endurance stacking. SW:toR is a game played by people (gold-spammers don't count) and people will make their own decisions on how to gear regardless of what you are I say. The purpose was to provide a different perspective on how gear can be optimized (and after all we have discussed I use the term loosely).

The end result of what works is always defined by the same thing: progression. If you are downing content, what you have need little changing (if any). If you are not downing content, then you need to look at if the reason is your own fault and seek ways to adjust your stats accordingly (if reasons had nothing to do with mechanical error). As you and I are both downing content in current tier and have NiM achievements from the previous tier I would put forth that neither of us is "wrong", we simply go about doing what we need to get done a different way.

Whether or not people actually use the stat differentials I posted is entirely inconsequential to me, I am simply trying to provide an alternate way of thinking about the stats. The best I am hoping for from this is, no offense to either of us, that people will see that there is not one way to go about doing things, and (even more hopefully) that some of them may even do their own math to find what works best for them instead of just taking what we have stated and following it. Granted the math shows that some things are ridiculous (1k Crit Rating, for example), but within the bounds of "I am not squishy" different tanks are going to have stats that fluctuate based on their play-style.

This is my final response to you personally, and I would like to thank you for the debate we have had. It is rare for me to actually be able to talk about these things this in-depth without people becoming dis-interested. I do hope in the future we may correspond on similarly thought-intensive topics.
Don't ever learn how to tank.
They won't let you do anything else once they know you are good at it.

Cidgarrillo's Avatar


Cidgarrillo
02.18.2015 , 07:21 PM | #29
If im not mistaken you are trying to get more or less the same % of Def,Shield and Abs in all your tanks.
You consider the passive bonus of each class as less points to spent to get the "standar" d-s-a percents. Right?
The thing is, and as KBN have stated, if we are maximizing gear for minimizing the damage taken, the passive bonus on the tanks is what makes the gear different for each of them.

I think i got what made you think that:
If we consider only the shield and absortion they obviously get better the closer the percents are:
*50% shield and 50% abortion is better DTPS than 75% shield and 25% absortion, even if they are at the same.
This is what made the PT and Assasins try to get the same but with different ratings.
This can generate a cognitive bias: the idea of an standarized optimal stats for the tanks.


But KBN put the mathematical proof of the idea of diferent optimal ratings and thus percents for each tank.


However, if my understanding of english is even worse than I believe please let me know. I want to understand what were you discussing.








Also, KBN, your stats as maximized for the average minimun damage taken per second?(the average table)
Wouldn't it be better to optimize for the lowest of the damage taken in the boss with the biggest damage taken between all the damages taken?

Jethsidi's Avatar


Jethsidi
02.18.2015 , 07:50 PM | #30
Quote: Originally Posted by Cidgarrillo View Post
If im not mistaken you are trying to get more or less the same % of Def,Shield and Abs in all your tanks.
You consider the passive bonus of each class as less points to spent to get the "standar" d-s-a percents. Right?
The thing is, and as KBN have stated, if we are maximizing gear for minimizing the damage taken, the passive bonus on the tanks is what makes the gear different for each of them.
More or less, Juggernaut and Powertech have the same percentages as the have the same armor rating and similar Damage Reducation quotients. Assassin's have roughly 10% more Absorb because it evens out the fact they take more base damage (from E/K, they take the least I/E damage of any tank).
I'm not sure the point you are trying to make about "what makes the gear different", as I completely agree that your stat differentials are going to vary based on the Class Bonuses. This is what has my Juggernaut at 1094 Defense Rating while my Assassin is at only 750 - because their passives give them different base percentages.

Quote: Originally Posted by Cidgarrillo View Post
I think i got what made you think that:
If we consider only the shield and absortion they obviously get better the closer the percents are:
*50% shield and 50% abortion is better DTPS than 75% shield and 25% absortion, even if they are at the same.
This is actually untrue, if you run the damage equations through algorithms that account for DtPS, Hit-to-Hit damage profiles, and even total damage taken over a fight you see that when your Absorb % reaches a certain point you actually get more mitigation out of stacking Shield. It's a balance between how much you are absorbing vs. how often you are actually shielding. To a certain degree it is more beneficial to be shielding more often for less than it is to be shielding for more less often (to a certain degree).

Quote: Originally Posted by Cidgarrillo View Post
But KBN put the mathematical proof of the idea of diferent optimal ratings and thus percents for each tank.
If you read through everything that we said to each other you will notice multiple times that we both mention that "optimal" is a point of view based on what you are gearing for. He and I go about what we do in a different way but we both accomplish the same goal - staying within average/acceptable DtPS profiles on fights and progressing through content.

I will re-iterate what I said in my final response to him: the point of this post was not "choose my numbers over anyone else's", it was simply a breakdown of how I run my stats and the statement that it works. It also works FOR ME, the purpose was to show a different perspective that can provide similar levels of mitigation but in a different way so that people would (hopefully) discern that there are multiple ways to accomplish the same thing and find what works best for them based on their given class and play-style.
Don't ever learn how to tank.
They won't let you do anything else once they know you are good at it.