Jump to content

Tanking: A Primer


Gankstah

Recommended Posts

  • Replies 105
  • Created
  • Last Reply

Top Posters In This Topic

I can also confirm that I am seeing the % increase. I have reported this as a bug and I suggest others do so as well. It looks like a decimal place was moved or something on accident. = ) I think it's supposed to be 60.15 not 61.5.

 

As far as the simplification of the mean mitigation calc, thank you very much. I will have that updated as soon as I get home (posting on my phone) monday. I knew eventually someone with an affinity for math would come along and clean up my sloppiness. I'm not very good at math. :)

 

Thanks again Tanis.

Edited by Gankstah
Link to comment
Share on other sites

As far as the simplification of the mean mitigation calc, thank you very much. I will have that updated as soon as I get home (posting on my phone) monday. I knew eventually someone with an affinity for math would come along and clean up my sloppiness. I'm not very good at math. :)

 

Thanks again Tanis.

 

No problem. By the way, I was thinking about the following. In the Georg Zoeller quote you posted on the previous page of this thread, he says that if shield chance and crit chance come into conflict, some shield chance can be pushed off of the table. That would presumably only happen when crit chance boosting cooldowns were blown in PvP though, if at all, but it should probably figure into Crit Immunity Theory somehow, right?

 

Anyway, if I understand it correctly, that Georg Zoeller quote would imply that the roll after the miss/avoidance check is based on a table which looks like one of these, depending on whether or not crit chance + shield chance > 1:

 

   crit      shielded    normal           crit      shielded
|--------|------------|----------|     |--------|------------|
0         X            X+B          1     0         X              1

 

Where X is the crit chance and B is the shield chance. If this is the case, that would mean that the base formula for crits in the guide is off because it assumes that the crit chance only applies after a check is made to determine that the attack wasn't shielded. If the check is made at the same time, as the dev post seems to indicate, then every attack which isn't avoided has probability X of being a crit despite the fact that an attack can't be both shielded and a crit. In other words, the formula involving crits should be:

 

1-(1-A)(1-D)(B*(1-C)+(1-B)+.5*X) = Mean Mitigation

 

instead of

 

1-(1-A)(1-D)(B*(1-C)+(1-B)(1+.5*X)) = Mean Mitigation

 

This will give a different answer than Baltazarr's formula because of the differing assumption. Also, I really ought to replace the .5 with a variable, such as:

 

1-(1-A)(1-D)(B*(1-C)+(1-B)+X*(Y-1)) = Mean Mitigation

 

where Y = the mob's crit damage multiplier (1.5 + whatever it gets from surge). It's also probably worth noting that B cannot exceed 1 - X, as I showed above.

 

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

 

It's really, really late at night right now, so please let me know if I misunderstood anything or made any mistakes. :D

Edited by tanis_
Link to comment
Share on other sites

I'm not exactly sure what it is you mean.

 

The crit theory section simply makes a probability check. What are the odds of both being hit and crit. Roll 1 * Roll 2. Since in TOR, unlike other MMO's, it's impossible to actually achieve crit immunity the only thing we can do is figure the probability and attempt to get that value as low (or high depending on your point of view) as possible.

 

Since crit trumps shield the only actual stat which affects this probability check that a Tank has influence over is avoidance (i.e. defense).

 

As far as how this has bearing on Balthazaar's calc, I think that discussion is best taken up with him since he would be the best person to discuss his frame of mind when coming up with that calc.

Link to comment
Share on other sites

I'm not exactly sure what it is you mean.

 

The crit theory section simply makes a probability check. What are the odds of both being hit and crit. Roll 1 * Roll 2. Since in TOR, unlike other MMO's, it's impossible to actually achieve crit immunity the only thing we can do is figure the probability and attempt to get that value as low (or high depending on your point of view) as possible.

 

Since crit trumps shield the only actual stat which affects this probability check that a Tank has influence over is avoidance (i.e. defense).

 

As far as how this has bearing on Balthazaar's calc, I think that discussion is best taken up with him since he would be the best person to discuss his frame of mind when coming up with that calc.

 

On the crit theory thing, I don't think what I said is terribly relevant after having slept. ;)

 

On the formulas for mitigation involving crits, I'm convinced I'm right there as long as it goes like this:

 

check 1: Hit is either avoided (miss, dodge, parry, etc) or it is not.

 

check 2: If not avoided, hit is either a crit, shielded or neither.

 

 

As long as that's how it works, what I said about it in my previous post should be correct. Baltazarr's version would be right if it went like this:

 

check 1: Hit is either avoided (miss, dodge, parry, etc) or it is not.

 

Check 2: If not avoided, hit is either shielded or it is not.

 

Check 3: If not avoided or shielded, hit is either a crit or it is not.

 

 

 

 

Edit: Obviously I could be wrong, as I frequently am. Hehe. I don't think so though. To any other math-oriented people, Baltazarr or whomever, I'd be obliged either way if you would confirm or disprove my position.

Edited by tanis_
Link to comment
Share on other sites

Based on tests using duels with other players, Taugrim has concluded that only weapon attacks can be shielded. (http://taugrim.com/2012/01/04/guide-to-bounty-hunter-powertech-trooper-vanguard-mechanics-and-pvp/)

 

Assuming this restriction holds true for PvE content, would you consider the effective contribution of shield chance and absorption to overall mitigation to be lower than your original formulas indicate?

Link to comment
Share on other sites

Based on tests using duels with other players, Taugrim has concluded that only weapon attacks can be shielded. (http://taugrim.com/2012/01/04/guide-to-bounty-hunter-powertech-trooper-vanguard-mechanics-and-pvp/)

 

Assuming this restriction holds true for PvE content, would you consider the effective contribution of shield chance and absorption to overall mitigation to be lower than your original formulas indicate?

 

I'm well aware of this situation. I recorded a few videos last night and sent links to BioWare illustrating this problem. However, I try to keep sensationalism out of this thread.

 

But believe me, BELIEVE ME, once I discovered this I was extremely perturbed. I find myself spending an absolute fortune on respec'ing costs. Being forced to swap between Immortal spec for PvE'ing and back to Rage when I want to PvP.

 

It's gotten so bad that I've had to roll another Jugg. One for PvP and one for PvE. My gripes with the defense issues tank spec's have with PvP has seriously jaded me towards this class and this game. I hoping that BioWare addresses it quickly. Because this is just completely unsatisfactory.

 

Investing 25+ points into a tree in order to efficiently soak damage only to have 90% of attacks BYPASS that damage soaking is completely unacceptable. With the current incarnation of the game Immortal is utterly useless in PvP. All forms of mitigation are bypassed.

 

Combine this with the fact that Deflecting/Parrying interrupts the GCD and you have one very unhappy Gankstah on your hands.

Edited by Gankstah
Link to comment
Share on other sites

I'm an older player and have never really been a true tank but I would like to try it. Unfortunately I have no idea what all this means, so I guess I'll just jump in and see how it goes.:eek:

 

Great post though!

Edited by Piteous
Link to comment
Share on other sites

I'm an older player and have never really been a true tank but I would like to try it. Unfortunately I have no idea what all this means, so I guess I'll just jump in and see how it goes.:eek:

 

Great post though!

 

Any questions you have about the content provided, feel free to ask. That's what we're here for.

 

Do you need to know any of this to play a tank? No. Could understanding the content help you in your quest? Yes. The information is intended to give players an idea of what's going on in the background of the game. But it is, by no stretch of the imagination, required to understand in order for you to play TOR.

 

Admittedly I've completely abandoned my Tanks in TOR. Until BioWare resolves the Defense GCD issue and the mitigation bypass issues Tanks have in PvP I will not be playing any of my Tanks. There's no point. I genuinely believe that I'm coming down with carpel tunnel because of the state of Tanks ATM. My hands and forearms actually hurt after playing for any period of time.

 

I need to take a break.

 

Despite my extremely jaded disposition I will endeavor to keep this thread up to date and attempt to answer questions to the best of my knowledge. For now, I'm going to take a break. I'll be checking in from time to time.

 

Any questions, concerns or updates should be sent to me through PM in order to get a prompt reply.

Edited by Gankstah
Link to comment
Share on other sites

I've got a problem with the given data. I've been crunching the numbers, and I've come to a weird conclusion.

 

Not counting seperate benefits from skills or talents and just using the provided formulae, it seems Defence trumps any of the stats no matter what the rating is.

 

The most beneficial allocation of stat points seems to be to put everything in defence, despite diminishing returns.

Link to comment
Share on other sites

I've updated the Primer.

 

  • The Mean Mitigation section has been cleaned up and the calc's simplified thanks to Tanis.
  • The Mods 4 Tanks section has been updated and cleaned up.
  • More refferences were inserted and format changed a little.

Link to comment
Share on other sites

I've got a problem with the given data. I've been crunching the numbers, and I've come to a weird conclusion.

 

Not counting seperate benefits from skills or talents and just using the provided formulae, it seems Defence trumps any of the stats no matter what the rating is.

 

The most beneficial allocation of stat points seems to be to put everything in defence, despite diminishing returns.

 

Yes and no. You have to bear in mind that Mean Mitigation is average mitigation. Some fights your mitigation will be higher. Some lower. There are several threads outside of the Primer and sites that can be found on the web going over "the best mitigation" combination. Typically they range in the neighborhood of 350-450 Defense, 300-350 Shield and 250-350 Absorption and they all defend their opinions quite adamantly.

 

To put it into perspective, would you rather invest 100 points to gain 1% defense or 100 points to gain 40% Absorption? Of course you would want to invest in Absorption. The idea is that even though Defense yields a higher Mean Mitigation it also yields a higher "spikey" mitigation pattern. In order to level that pattern out yet keep a high Mean Mitigation we sacrifice Defense to invest into Shield/Absorption. To help keep those "spikey" moments to a minimum while maintaining a good Defense%.

 

At what point you decide that enough is a enough is entirely up to you. I would never recommend that a person invest solely in Defense. And to be honest, that's not even a possibility anyway given how mods function in TOR. Ultimately what you decide to do will be the right answer. We're dealing with Mean Mitigation differences in the neighborhood of 1-3% which, to be honest. really isn't that big of a deal.

 

BioWare doesn't design their content with extreme min/maxers in mind. So long as you're in the neighborhood of your peers in hard/nightmare content you won't have any harder a time soaking damage than anyone else. Do what you do and don't sweat the small stuff.

 

Leave the min/maxing debates to those that are prone to those discussions and enjoy the game as you would enjoy it. You won't be locked out of content because you decided to allocate 50 points into Shield instead of Absorb. And if you are, why would you want to roll with that kind of crowd anyway? It's their loss. Not yours.

Link to comment
Share on other sites

You really should add a stipulation in the "mean mitigation" section about accuracy/damage debuffs. With debuffs, Assassins and Juggs swap places.

 

There are two addendums in the Primer which stipulate that not all bonuses are applied. They are:

 

  • Many abilities (particularly high CD abilities) are not factored in for a reason. There's a lot that goes into making a tanking class and DR/Shield/Avoidance is just one part.

  • These are situational modifiers from abilities that directly affect DR. I mention these separately because the effects are longer than the CD so they can be applied constantly but are situational as not ALL targets on the board will be under their influence at ALL times.

 

Furthermore the reader is encouraged to add or remove bonuses as they see fit to simulate different situations and/or encounters. I am not here to account for all possible situations and/or outcomes TOR may provide. I am not here to do the work FOR the reader. This isn't my job. I don't get paid for this. Nor am I required to cater to the whims of every poster who feels he or she is entitled to have her chosen archetype shown in a greater light. However situational that light may be.

 

I'm not saying you are doing so but we have had several other posters come in here and act as such. I am here solely for the purpose of providing the means by which the reader may do the work for his or her SELF.

 

If the reader chooses not to, that is not my fault. That is the reader's. Let's say I humor the posed request, and by the way several other posters have requested this as well so you are not alone, I would have to humor the BH's who have likewise requested to break down the math to illustrate how the Sin's Mean Mitigation drops while tanking mobs over a count of 2 thanks to Dark Ward's inferior handling of multiple mobs. Then I would have Sin's coming in here and bemoaning me for doing that as well.

 

This t.i.t. for tat mentality is exactly what I try to avoid.

 

While I understand and appreciate the enthusiasm forumites have for their preferred Tanking classes, the Primer is not a battle ground for enthusiasts to come in and wage war over who is the better Tank in what situation. Nor will I be goaded into painting one archetype in a better light than another. Because my words will be used as gospel from on high to "prove" someone's point in another thread. This has happened before and I'm not putting myself into that situation again.

 

Finally, I have stated before and I will state again that the Primer is not, nor shall it ever be, all encompassing. It is strictly a means to educate and promote the understanding of theory and game mechanics. That is all. There are a multitude of resources both on the official forums and on the web which go into very intricate detail the varying circumstances and mitigation differences between Tanking archetypes.

 

There's a reason it's called "Tanking: A Primer" and not "Tanking: A Guide".

 

I want to stress that I'm not angry at you or your request. I'm really not. I'm mostly typing this out so I can link this post the next time this or similar questions/requests are posed. There are a great many PM's I receive and posts that have been removed because of this exact situation. This is mostly for my benefit.

Edited by Gankstah
Link to comment
Share on other sites

The changes to the mean mitigation section look good, except that you're using the first formula I posted for crits, which is wrong.

 

In that formula, "(1-B)" is directly multiplied by "(1+.5*.25)" which is the crit chance and damage part. So you've got the chance to not shield an attack times crit stuff. This means that this formula works if and only if increasing your shield chance automatically decrease how much damage you take from crits, and we know that it doesn't.

 

The formula that's correct is 1-(1-A)(1-D)(B*(1-C)+1-B+X*(Y-1)) = Mean Mitigation

 

Where Y is the crit damage multiplier (ie 1.5 if the mob has 0 surge).

 

Note that in this formula, 1-B is not multiplied by the crit stuff (X*(Y-1)). If you are confident that crit damage from mobs is always 150% of non-crits (and it may be, I really have no idea), then you can replace the (Y-1) with .5.

 

For completeness, here's what the newest formula says when broken down.

 

* All attacks have to deal with (1-A) and (1-D), which means that all attacks can be shielded or mitigated, unless those variables are set to zero of course.

 

* Attacks which are not avoided are either shielded, a crit, or neither. B*(1-C) deals with shielded attacks, 1-B deals with normal attacks, and X*(Y-1) deals with crit attacks by adding the extra damage that occurs over and above a normal attack.

 

Or, you could break it down like this:

 

B*(1-C) = M = mean percent unmitigated damage of non-avoided, shielded attacks before DR

 

1-B = N = mean percent unmitigated damage of non-avoided, non-shielded attacks before DR and crits

 

X*(Y-1) = P = mean percent unmitigated damage of non-avoided crits before DR, minus N

 

M + N + P = Q = mean unmitigated percent damage of all non-avoided attacks before DR

 

(1-A)(1-D)*Q = R = mean unmitigated percent damage of all attacks after avoidance and DR

 

1 - R = 1-(1-A)(1-D)(B*(1-C)+1-B+X*(Y-1)) = mean mitigated percent damage from all attacks

Edited by tanis_
Link to comment
Share on other sites

this is to you Gankstah.

 

when you compare classes and the migitation, will you try and take a look at the 14/27 immortal/vengence tank spec?

 

you loose sonic barrier, and we can debate how good this really is! :D

you loose 4% shield chance.

you loose 4% internal/elemental damage reduction.

 

you gain 4% more hp, and intercede on your self (20% damage reduction for 6 sec when you use it on an ally - use it on a melee for the buff)

you gain unstoppable after you use charge (20% damage reduction for 4 sec, use it when the boss does heavy AoE move out charge back in)

you gain 4% flat damage reduction to both armor and internal/elemental

 

 

here's the build, just in case.

http://www.torhead.com/skill-calc#101dMG0uZhGMMrhdzMM.1

Link to comment
Share on other sites

Im truely amazed you put this together so quickly, with almost all gaps filled. Its somthing to make a post like this 2 years down the line, but to grind all this info together and put it in such a sure and mathmatically sound way is fantastic.

 

Not that im poking fingers or holes as this is obviously sound well thought out information, but im curious where you go the figures from so accurately. Personally I didnt know that ranged had a higher threat cap than melee in this game i thought it was a static 1.1-1.2 accross the board.

 

Gonna have to reflick over the crit immunity as it looks quite different to WoW TBC formulae, glad you crunched it out. Would be interested to see if all bosses have a set base stat of hit/crit, or of its varied not providing us with an efficient cap to reach.

 

Overall this has been a fantastic help to me, thanks.

Link to comment
Share on other sites

...im curious where you go the figures from so accurately. Personally I didnt know that ranged had a higher threat cap than melee in this game i thought it was a static 1.1-1.2 accross the board.

 

From datamining before they compressed the client for release.

 

Now all this information is held server side so it makes it more difficult to find. But not impossible. If they change anything though, it's relatively easy to test for so long as you know the calc's, which we do.

 

Unless of course they change the calc's lol. But if they did that we have worse things to worry about. See: SWG Pre-CU, CU and NGE.

Edited by Gankstah
Link to comment
Share on other sites

Also, earpieces can have augment slots from critical crafting.

 

And Implants and weapons.

 

I just haven't gotten around to fully updating that section. Last time was a pass through to change the dialogue and nomenclature. Will be taking care of that "soon" rest assured. Thanks for the reminder though! :D

Link to comment
Share on other sites

Thank you very much for the recent update about Threat! I am a relative noob and have been trying to learn the ropes of tanking. I read up here and there, and never realy knew the difference between threat and aggro and how threat works until now. Very concise and easy to understand. It has already made a big difference in my PvE parties.

 

So thanks again.

 

Sincerely,

 

Nooby McTank

Link to comment
Share on other sites

Thank you very much for the recent update about Threat! I am a relative noob and have been trying to learn the ropes of tanking. I read up here and there, and never realy knew the difference between threat and aggro and how threat works until now. Very concise and easy to understand. It has already made a big difference in my PvE parties.

 

So thanks again.

 

Sincerely,

 

Nooby McTank

 

I'm glad you took something away from it. That's what we're here for.

Link to comment
Share on other sites

×
×
  • Create New...