Remove Advertisements

Quantifying Gear Stats

Get help with your character's gear

Moderators: Fridmarr, Worldie, Aergis, Sabindeus, majiben, lythac, Digren

Re: Quantifying Gear Stats

Postby theckhd » Thu Jan 06, 2011 4:06 pm

I agree. A quick-reference guide for damage reduction potential and combat table coverage would be incredibly handy. For example:

Code: Select all
Item    -  DR%   CTC%
Item #1 - 1.2% - 3.4%
Item #2 - 1.4% - 3.0%
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7739
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Quantifying Gear Stats

Postby Digren » Thu Jan 06, 2011 7:21 pm

theckhd wrote:If you ignore diminishing returns (A'=0), this gives you (1-Cm/(2.25*k*Ca)) = 0.5284, or about 53%. Using a more realistic value of A'=0.075, we get Ca'=215.3 and (1-Cm'/Ca')=0.63, or 63%.

Let me assume a hypothetical item with 100 dodge rating and some amount of hit rating. According to your result, if the item has about 63 or less hit rating, better combat table coverage is attained by reforging the dodge rating. However, if the item has 63 or more hit rating, better combat table coverage is attained by reforging the hit rating. (All for a realistic amount of diminishing returns given our current gear options.)

That means for a hypothetical item with 100 dodge rating and 63 hit rating, reforging either value to mastery rating should yield about the same increase in combat table coverage. In other words, (+40 mastery rating -40 dodge rating) equals (+25 mastery rating -25 hit rating) in terms of combat table coverage.

Assuming the hit rating it unwanted, +40MR-40DR=+25MR, or 15MR=40DR, or 1DR=0.375MR. A person who gears to maximize combat table coverage should value avoidance rating as just 37.5% the value of mastery rating.

In this scenario, if 1MRE=1V, then 1ARE=0.375V. This greatly reduces the value of avoidance rating, which is just 37.5% as good as mastery rating. And note that this is based on our current levels of diminishing returns. As our overall avoidance goes up, the incremental value of avoidance relative to mastery will go down, pushing this number lower.

Other the other hand, Theck wrote:
Or in other words, the break-even point occurs when the amount of threat rating on the item is 7.5% of the avoidance rating on the item. For an item with less threat rating, you would be better off reforging the avoidance rating to mastery. Since this should not be the case for nearly all (all?) items available right now, you will in general get a larger reduction in damage intake by reforging the threat rating to mastery.

For minimizing damage taken, the breakeven item has 100 dodge rating but just 7.5 hit rating. In this case, +40MR-40DR=+3MR, or 37MR=40DR, or 1DR=0.925MR.

In this scenario, if 1MRE=1V, then 1ARE=0.925V. Avoidance rating is 92.5% as good as mastery rating. This too is based on typical diminishing returns, and will drop as total avoidance increases.

------

It was my personal goal to gear to maximize combat table coverage. However, in order to do so, I would need to greatly devalue avoidance rating from where my "gut" feel said it should be. On the other hand, gearing to minimize damage taken values avoidance higher than I felt, and would require me to revalue gear at every tier. Most likely, neither extremes are best. In one case, you take more but smoother damage. In the other, you take less but spiky damage. The best choice is probably somewhere in the middle.
User avatar
Digren
Moderator
 
Posts: 2139
Joined: Sat Jul 26, 2008 2:41 pm

Re: Quantifying Gear Stats

Postby Digren » Thu Jan 06, 2011 7:30 pm

theckhd wrote:I agree. A quick-reference guide for damage reduction potential and combat table coverage would be incredibly handy. For example:

Code: Select all
Item    -  DR%   CTC%
Item #1 - 1.2% - 3.4%
Item #2 - 1.4% - 3.0%

The logistical problem is that this requires each item to be evaluated with two possible gem and reforging strategies - one designed to maximize damage reduction, and the other designed to maximize combat table coverage. There's quite a bit more work involved in creating such a guide, and it really only seems useful if tanks need to create two separate sets of gear (one for each scenario) to gear for different encounters.

If most (all?) encounters favor a set blended for damage reduction and damage smoothing, then the effort to maximize each piece for each extreme is unnecessary and perhaps unwanted, as a third gem and reforging strategy might be wise for the middle ground. The swing in valuation of avoidance rating is huge - from 37.5% of mastery rating all the way to 92.5% of mastery rating - between the two scenarios. I don't know yet that helping people gear for either extreme is the best choice.

(Yes, we could just give out data and let people make their own decisions. That tends to lead to people gearing for avoidance in WotLK and threat in Cata. I'd rather help guide people towards making the correct decisions. If the correct choice is a blended set, then I want to recommend a blended set.)
User avatar
Digren
Moderator
 
Posts: 2139
Joined: Sat Jul 26, 2008 2:41 pm

Re: Quantifying Gear Stats

Postby econ21 » Fri Jan 07, 2011 2:23 am

Digren wrote:The swing in valuation of avoidance rating is huge - from 37.5% of mastery rating all the way to 92.5% of mastery rating - between the two scenarios.


That's a very illuminating point. I wonder, is anyone currently gearing for Combat Table coverage? I would do it if I could get to 102.4%, but that's impossible with the current content. Maybe it makes sense if healers have enough mana and you are just smoothing out damage. At the moment, I keep an eye on my unhittable macro but am not trying to maximise it. What I would tend to do if we did have a gear list with the two numbers Theckhd suggested is pick the item with the highest damage reduction, unless another one with a similar DR % had much better combat table coverage (i.e. mastery + 1 avoidance > dodge + parry).

With the gem normalisation, my inclination would be to just assume a balanced set - 20 mastery/30 stamina in blue, prismatic and yellow, 20parry/30 stamina in red. In most cases, gemming wouldn't affect the relativities between items (e.g. most 346 chests have 2 gems). And people wouldn't use a gear guide to exactly compute combat table coverage - the macro is for that - just to see which item looks like it gives you more than another.

PS: I have a vague memory of a BC era gear guide that included contribution to combat table coverage - does anyone else? Or maybe I just annotated that myself.
econ21
 
Posts: 1376
Joined: Tue Apr 01, 2008 8:53 am

Re: Quantifying Gear Stats

Postby theckhd » Fri Jan 07, 2011 7:11 am

Digren wrote:
theckhd wrote:I agree. A quick-reference guide for damage reduction potential and combat table coverage would be incredibly handy.[/qutoe]
The logistical problem is that this requires each item to be evaluated with two possible gem and reforging strategies - one designed to maximize damage reduction, and the other designed to maximize combat table coverage. There's quite a bit more work involved in creating such a guide, and it really only seems useful if tanks need to create two separate sets of gear (one for each scenario) to gear for different encounters.


Which is why a quick-reference guide where someone else has already done the work would be handy! :P

More seriously though, if you look at the gearing patterns of some of the top-tier raiding tanks, you'll see that they don't all follow one strategy unilaterally. There are variations from player to player, and even from item to item within a single player's gearing. Many were reforging hit and expertise to mastery even in cases where combat table coverage would be better by reforging an avoidance stat, despite reforging and choosing gear to maximize combat table coverage on other items.

Which I guess is good news for your project, because it means that in essence, they're aiming for an average, well-rounded increase in both factors. That means that there should be an average equivalency value somewhere between the two extremes that you identified a few posts ago, and that's the target you're aiming for.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7739
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Quantifying Gear Stats

Postby inthedrops » Fri Jan 07, 2011 8:05 am

theckhd wrote:
Digren wrote:
theckhd wrote:I agree. A quick-reference guide for damage reduction potential and combat table coverage would be incredibly handy.[/qutoe]
The logistical problem is that this requires each item to be evaluated with two possible gem and reforging strategies - one designed to maximize damage reduction, and the other designed to maximize combat table coverage. There's quite a bit more work involved in creating such a guide, and it really only seems useful if tanks need to create two separate sets of gear (one for each scenario) to gear for different encounters.


Which is why a quick-reference guide where someone else has already done the work would be handy! :P

More seriously though, if you look at the gearing patterns of some of the top-tier raiding tanks, you'll see that they don't all follow one strategy unilaterally. There are variations from player to player, and even from item to item within a single player's gearing. Many were reforging hit and expertise to mastery even in cases where combat table coverage would be better by reforging an avoidance stat, despite reforging and choosing gear to maximize combat table coverage on other items.

Which I guess is good news for your project, because it means that in essence, they're aiming for an average, well-rounded increase in both factors. That means that there should be an average equivalency value somewhere between the two extremes that you identified a few posts ago, and that's the target you're aiming for.


And this is exactly why I want to know what the tradeoffs would be between different pieces of gear, in a quick lookup manner. It's why I don't want a single flat number but separate weightings. For example, if I was trying to keep a target of 140k health and maximize actual avoidance, and I had two pieces that allows me to keep my stam where I want it, seeing that:


PieceA: 1.1% mit, 0.4% avoid.
PieceB: 0.9% mit, 0.42% avoid.

I would happily go ahead and choose PieceA even though it has less avoidance. I'll take the 0.2% boost in mitigation at the loss of 0.02% avoidance.

The trick here is that I would need this tool to account for me specific gemming strategies, and/or each method of gemming a piece of gear. In other words, I would need a gear calculator.

I used to use TankPoints for that, maybe I should see what it's up to these days.
inthedrops
Maintankadonor
 
Posts: 1281
Joined: Mon Oct 29, 2007 9:19 am

Re: Quantifying Gear Stats

Postby econ21 » Fri Jan 07, 2011 8:12 am

inthedrops wrote: The trick here is that I would need this tool to account for me specific gemming strategies, and/or each method of gemming a piece of gear. In other words, I would need a gear calculator.


I am hoping we can use RAWR for that, when it is done. I used to use it to check I would be uncrittable under various gear/gem/enchant permutations. Doing it manually, even using a gear list with some handy calculations, would be cumbersome.
econ21
 
Posts: 1376
Joined: Tue Apr 01, 2008 8:53 am

Re: Quantifying Gear Stats

Postby inthedrops » Fri Jan 07, 2011 8:14 am

It'll take a lot to convince me to use Rawr. That program was crap when it supported paladin tanks in WotLK. I don't want the program to tell me what's better, that's my decision to make. I want the program to just give me the numbers and let me decide.
inthedrops
Maintankadonor
 
Posts: 1281
Joined: Mon Oct 29, 2007 9:19 am

Re: Quantifying Gear Stats

Postby Digren » Fri Jan 07, 2011 11:23 am

econ21 wrote:
Digren wrote:The swing in valuation of avoidance rating is huge - from 37.5% of mastery rating all the way to 92.5% of mastery rating - between the two scenarios.


That's a very illuminating point. I wonder, is anyone currently gearing for Combat Table coverage? I would do it if I could get to 102.4%, but that's impossible with the current content. Maybe it makes sense if healers have enough mana and you are just smoothing out damage. At the moment, I keep an eye on my unhittable macro but am not trying to maximise it. What I would tend to do if we did have a gear list with the two numbers Theckhd suggested is pick the item with the highest damage reduction, unless another one with a similar DR % had much better combat table coverage (i.e. mastery + 1 avoidance > dodge + parry).

With the gem normalisation, my inclination would be to just assume a balanced set - 20 mastery/30 stamina in blue, prismatic and yellow, 20parry/30 stamina in red. In most cases, gemming wouldn't affect the relativities between items (e.g. most 346 chests have 2 gems). And people wouldn't use a gear guide to exactly compute combat table coverage - the macro is for that - just to see which item looks like it gives you more than another.

PS: I have a vague memory of a BC era gear guide that included contribution to combat table coverage - does anyone else? Or maybe I just annotated that myself.


I was gearing and guiding for combat table coverage, by recommending the "60% rule" (aka the "55 percent rule" before the math was finished) describing when to reforge which stat on avoidance/threat gear. Effective last night I am changing my strategy to always reforge the threat stat.

Likewise, after the discussion here I have slightly devalued stamina in comparison to mastery. Thus, effective immediately I will also recommend mastery/stamina gems in all blue, yellow, and prismatic sockets, except for JCers who can use a mix of their special stamina and mastery gems. (I recommend two mastery and one stamina of those gems.) Sure pure mastery in yellow and pure stamina in blue have the exact same effect, but it's easier to just buy or hire a JCer to cut a bunch of one gem, then use it most everywhere.

I'm still unsure on my red gem slot recommendation, though, as 20 par/20 mas is threatening to overtake 20 par/30 sta. In either case, if the socket bonus sucks, 20 mas/30 sta goes there.
User avatar
Digren
Moderator
 
Posts: 2139
Joined: Sat Jul 26, 2008 2:41 pm

Re: Quantifying Gear Stats

Postby Digren » Fri Jan 07, 2011 11:35 am

I'm working my way through my gear guide, reordering the recommendations based on the results of this discussion. I'm not going to publish blended values V, as I don't want to convey the aura of objectivity that V can imply.

I'm sourcing this information from a spreadsheet, though, and when I'm done I can post the spreadsheet in its entirety. It's not as automated as some of you would hope - you have to manually account for gem and reforge strategies right now - but it contains all the numbers to let you tweak and tune gear, weigh stats, and then generate your own best in slot list.
User avatar
Digren
Moderator
 
Posts: 2139
Joined: Sat Jul 26, 2008 2:41 pm

Re: Quantifying Gear Stats

Postby inthedrops » Fri Jan 07, 2011 11:35 am

Digren wrote:I'm still unsure on my red gem slot recommendation, though, as 20 par/20 mas is threatening to overtake 20 par/30 sta. In either case, if the socket bonus sucks, 20 mas/30 sta goes there.


One simple suggestion, completely dependent on a bajillion factors I realize, is that if you have two solid stamina trinkets to go ahead and go 20 par/20 mas, otherwise 20 mas/30 sta might be warranted depending on your gear level.

The premise being, if you've got two stamina trinkets equipped, there shouldn't be as much need to gem for it, even with the weaker trinkets.

The inverse would apply with two mastery trinkets as well.
Last edited by inthedrops on Fri Jan 07, 2011 11:41 am, edited 1 time in total.
inthedrops
Maintankadonor
 
Posts: 1281
Joined: Mon Oct 29, 2007 9:19 am

Re: Quantifying Gear Stats

Postby yappo » Fri Jan 07, 2011 11:39 am

Digren wrote:Likewise, after the discussion here I have slightly devalued stamina in comparison to mastery. Thus, effective immediately I will also recommend mastery/stamina gems in all blue, yellow, and prismatic sockets, except for JCers who can use a mix of their special stamina and mastery gems. (I recommend two mastery and one stamina of those gems.) Sure pure mastery in yellow and pure stamina in blue have the exact same effect, but it's easier to just buy or hire a JCer to cut a bunch of one gem, then use it most everywhere.

I'm still unsure on my red gem slot recommendation, though, as 20 par/20 mas is threatening to overtake 20 par/30 sta. In either case, if the socket bonus sucks, 20 mas/30 sta goes there.



Seems mostly sound. However, given input from people, who, in difference from me, have valid information from raids, we should probably try to find the health-bar needed. Pushing for it and after that going for full avoidance/mastery is likely to be more efficient than slamming a stam/mastery gem into every available slot with no concern taken to the tank's actual health.
yappo
 
Posts: 1107
Joined: Wed Sep 17, 2008 4:15 pm

Re: Quantifying Gear Stats

Postby Digren » Fri Jan 07, 2011 11:40 am

I have to give credit to Blizzard. They took away a bunch of stats and simplified gear design and yet made gear choice a lot more complicated in Cata. The ability to reforge alone means every piece is as many as three or four pieces based on the different viable reforge options. Throw in different viable gem options thanks to usable socket bonuses and more than one stat to stack and evaluating any one piece of gear can require looking at it ten different ways to see how it works best.
User avatar
Digren
Moderator
 
Posts: 2139
Joined: Sat Jul 26, 2008 2:41 pm

Re: Quantifying Gear Stats

Postby Digren » Fri Jan 07, 2011 11:42 am

yappo wrote:Seems mostly sound. However, given input from people, who, in difference from me, have valid information from raids, we should probably try to find the health-bar needed. Pushing for it and after that going for full avoidance/mastery is likely to be more efficient than slamming a stam/mastery gem into every available slot with no concern taken to the tank's actual health.

In the gear guide I want to recommend a baseline stamina for each tier. So far I have 115k for a stage one tank, 130k for a stage two tank, and 150k for a stage three tank. If my gem recommendations make it difficult to reach those baselines, then yeah, I agree I'll have to adjust the recommendations.

It's difficult to play with that without a working character profiler. Wowhead's still isn't there yet.
User avatar
Digren
Moderator
 
Posts: 2139
Joined: Sat Jul 26, 2008 2:41 pm

Re: Quantifying Gear Stats

Postby yappo » Fri Jan 07, 2011 6:57 pm

Digren wrote:
yappo wrote:Seems mostly sound. However, given input from people, who, in difference from me, have valid information from raids, we should probably try to find the health-bar needed. Pushing for it and after that going for full avoidance/mastery is likely to be more efficient than slamming a stam/mastery gem into every available slot with no concern taken to the tank's actual health.

In the gear guide I want to recommend a baseline stamina for each tier. So far I have 115k for a stage one tank, 130k for a stage two tank, and 150k for a stage three tank. If my gem recommendations make it difficult to reach those baselines, then yeah, I agree I'll have to adjust the recommendations.

It's difficult to play with that without a working character profiler. Wowhead's still isn't there yet.


Seems Blizz is doing a really good work this time :D
yappo
 
Posts: 1107
Joined: Wed Sep 17, 2008 4:15 pm

PreviousNext

Return to Gear Discussions and Advice

Who is online

Users browsing this forum: Google [Bot] and 0 guests


Remove Advertisements

Who is online

In total there is 1 user online :: 1 registered, 0 hidden and 0 guests (based on users active over the past 5 minutes)
Most users ever online was 380 on Tue Oct 14, 2008 6:28 pm

Users browsing this forum: Google [Bot] and 0 guests