footloop

I did a simple test of this earlier today. The "before diminishing returns" only applies to the "x.xx% dodge from y dodge rating" that you see upon mouseover. The actual dodge percent shown on your character sheet seems to take DR into account.

 

Shadocat

does this hold true for DK's ?

 

Grayson Carlyle

TTL math was done extensively in the ratings at 80 thread. The end result is that gemming for defense > all other options except at extremely low values of dodge (levels so low you can't possible be that low wearing any gear at all). And it only gets better as the undiminished block % becomes a larger factor in the total TTL for each point of defense.

 

Whitetooth

Block is not considered at all in my TTL calculations, defense is better simply because is gives more avoidance under the same amount of item value.

For example:
2000 Parry rating = 22.32869363% Parry
2000 Dodge rating = 33.16162431% Dodge
2000 Agi (War/DK) = 21.50812438% Dodge + 4000 Armor
2000 Agi (Pal) = 27.59166181% Dodge + 4000 Armor
2000 Defense = 14.26064579% Dodge + 12.49202826% Parry + 5.6% Crit Avoidance + 14.26064579% Miss(Assuming same DR as Dodge) = 46.61331984%

Hotdogee@Ner'zhul US <Bahamut>
Author of RatingBuster

 

Kreen

Because of the way diminishing returns works, it's fairly intuitive that having that much of any type of rating will yield those results. Defense effectively mitigates the effect of the DR better as it spreads them over more stats. For instance, if you took 1000 parry rating and 1000 dodge rating and compared them to 2000 defense rating, I'm sure they would be closer (defense would probably still win given there is no miss gem, but you see where I'm going with this).

I think it would be more useful to compare them at similarly attainable incremental levels of each rating (i.e. if a warrior has x% parry, y% dodge and z% miss from gear how much TTL does adding 1 dodge, parry or defense gem give you at different boss swing speeds and hit magnitudes).

I'll see if i can put something like this together soon.

 

Riot

I'll update my OP into some kind of cohesive summary at some point, probably when my new motherboard arrives and I'm not on a shitty laptop.

Overslept, / So tired. / If late, / Get fired. / Why bother? / Why the pain? / Just go home / Do it again. - The Commuter's Lament
Norman B. Colp

 

◊ Ugato

Well, I'm trying to get an avoidance spreadsheet (very basic, plug in the rating numbers manually, and it spits out the raw and adjusted avoidance figures; even gives you a diminishing returns percentage for where you are currently. The problem I'm running into is that my avoidance is religiously coming in just a little higher than what armory shows.

Just looking for someone to take a second glance at my formulas, and make sure it's not something dumb. I even involve base numbers to come up with a final figure. 5 dodge/5 parry is pretty standard, though, as well as 5 more dodge/parry from talents. I'm confident it's somewhere in the diminishing returns formula, since the discrepancy grows along with the ratings.

Defense rating -> defense: X/4.9185

Defense -> Dodge/Parry: X*0.04

Dodge Rating -> Dodge: X/39.34799

Parry Rating -> Parry: X/48.18499

Agility -> Dodge: X/.0136



This is the diminishing returns formula given, and that's why I'm confused that my formula might be wrong. c is a constant and k is a constant.

According to TankSpot - where the formula comes from - k is .956 for warriors. c is 88.129021 for Warrior/Dodge, and 47.003525 for Warrior/Parry.

This makes 1/c 0.011347 for Warrior/Dodge and 0.021275 for Warrior/Parry.

So, the only input is your raw avoidance (A), which gives you 1/Ad. Dividing the second part of the formula into 1 should give you Ad (perhaps this is my mistaken assumption?). Regardless, here are the formulas as they appear in my spreadsheet (I'll replace cell numbers with 'A').

Dodge after DR: =1/(0.011347+(0.956/A))

Parry after DR: =1/(0.021275+(0.956/A))

 

Xerophyte

Yes, the base idea is correct.


There is one typo in your posted formula

Agility -> Dodge: X/.0136

should be

Agility -> Dodge: X * 0.0136

If your spreadsheet has the same then that'd account for getting higher than expected avoidance from agility by quite a lot.

 

Riot

Okay done. Quick, dirty and simple. If anyone wants to suggest something, suggest away.

Overslept, / So tired. / If late, / Get fired. / Why bother? / Why the pain? / Just go home / Do it again. - The Commuter's Lament
Norman B. Colp

 

Xerophyte

The current wording makes it sound like swapping to parry would give some small increase but isn't worth the effort, however for the vast majority of cases you'll never have sufficient dodge for the DR to make parry rating more efficient. Just the 5.6% undiminished parry you get from the def to crit immune is enough to push the break even point to around 23% undiminished dodge. I can't think of any gear and buff setup that reaches that mark right now, much less one that does so without adding parry chance, so replacing tank gear with dodge rating to an equal amount of parry rating is going to be a loss in avoidance.

 

Suesse

A few thoughts.
The tanking classes (I think) are currently balanced around paladins never taking an unblocked hit, but warriors "should" with the trade off that they have things like critical block with average out in their favor. When warriors reach no-unblocked-hits, they gain a big advantage, will they need to rebalance when this happens? From the warrior spreadsheet, I can see that this is already possible, but with otherwise non-optimal gear. Eventually, from later raid dungeons, I suspect this will be possible with normal gear.

What about when blizzard wants to add an "add" element to a boss fight with a massive number of melee adds? How do they balance this? In future raids, the warrior and paladin may be able to block every attack with a minimum of 4k block value. Let's say you want to have 20 adds each of which hit for 2k. The warrior and paladin are immune to this damage, so you have to jack it up to 5k damage. Solo tanking the warrior and paladin have a worst case of taking 20k damage when all swings happen at once. Non-shield tanks can take 100k damage instantly. Certainly you can design fights around this problem, but it really does restrict your options. I will be very upset if they start implementing mass numbers of mobs which disarm my shield or knock me down, for example, but this would be one fix.

100% block with high block values also restricts the design of rogue-type bosses if you want balance with non-shield classes. Let's assume 4k block value with 50k health on tanks. Two auto attacks 20k and 10k (offhand) plus one 20k special attack every two seconds. It seems like shield-tanks have a big advantage here (12k off their worst-case). However, with this case, one might assume that the flaw is assuming both warriors and druids will have approximately the same health -- it may be that if warriors have 50k that druids have 60-80k. I'm also ignoring armor differences. I'm not sure how different it is now, if the warrior/paladin take double armor pots (works better in short 4-ish minute fights) and gets many inspiration procs.

 

Branar

I don't think warriors gain "a big advantage" over paladins. The advantage paladins have is that it requires them substantially less block rating to reach a point where they never taken an unblocked hit; the itemization points spent on block rating to get a warrior to that point can translate into additional block value, stamina, or any other stat for paladins that increases mitigation in some other way.

Obviously if you just go with "warrior and paladin gear will both be itemized to add ~80% block rating", then yes, warriors have a huge advantage. But the advantage there for warriors is that the gear is itemized well, compared to paladins having wasted itemization points. It'd be like if druids had to use parry rating gear.

I suppose your "think of the adds" argument as well as your rogue-type boss argument are technically valid, but they seem remarkably rare in-game in reality. My personal experience is that tanking adds on fights rarely puts any tank at risk of deaths. The healing on Maexxna adds, for example, is an utterly trivial part of the fight no matter who is tanking. Ditto Faerlina adds. Adds are - generally - not a test of tank survival, but a test of tank control.

As well, I think rogue-type bosses are not typically a tank-gibbing risk. Any fight where you can block a quarter or more of the value of a given hit *before* hitting shield block (which is what your 4k block value against 20k hits looks like) is certainly going to be a trivial tank survival case, in my opinion...we're talking about a boss that hits about as hard as Loatheb vs. entry level blues and heroic epics). The tank-gibbing bosses seem to combine fast attack speeds with very high damage values - Sartharion+2-3 drakes, Patchwerk, etc.

Frankly, my perception with blocking in general is that it shines the most in the situations that are easiest to tank, as Amorpheus suggests. But really just about any tank can handle those encounters anyway, with or without block.

 

Amorpheus

It's not so much about handling, but when it starts making a big difference. I have a block suit now that raid buffed only leaves a few percent as normal hits. Block value is around 1500. With some upgrades it is entirely possible that I tank an entire continent worth of mobs. It's already working against attackers that are slightly below 80. Mostly this seems like a gimmick (can't solo heroics because it goes to hell against stuns or spell damage) but it's entirely thinkable that it might allow our raid to ignore some aspects of fights where you usually have to aoe now and then.

The problem with block is that it's a hard number. If the enemy hits lower than what's blocked, it's just as good as avoidance. If the hit is significantly higher the effect is that much smaller. Although I'm not sure how it should be balanced if block would modify a percentage of the received hit, that might be something worth thinking about to make it more universal... as it is right now my main equipment balances avoidance and offensive stats after getting as much stamina as possible. And I'm sitting barely above 20% block there because most of the time going for the block version in a slot does sacrifice more important stats. For example, [Lavanthor's Talisman] versus [Figurine - Monarch Crab]. Lots of block on that item, but the stamina trinket raises my health by as much as a block removes. Health is always there, block is a chance against physical hits.

"You are better than I am," Inigo admitted.
"So it seems. But if that is true, then why are you smiling?"
"Because," Inigo answered, "I know something you don't know."
"And what is that?" asked the man in black.
"I'm not left-handed."

 

Suesse

I think this discussion relates back to the last expansion. Recall how brutal unnerfed R&J or Prince could be in Karazhan or Tidewalker in SSC when the star aligned with parry haste and fast attacks eat all the charges of shield block resulting in huge burst involving crushing blows. Back then some warriors built sets which made them passive crush immune, but the sacrifice was huge (I'm not sure if this was possible before they got some SSC gear or not though). At the time I think most people thought this was strange and not worthwhile. However, by the time guilds were finishing up T6 content or starting Sunwell, most tanks were passive crush immune in their normal progression gear without trying to be (at least with raid buffs / debuffs up).

By the time warriors are geared in 3.1 and 3.2 raid gear, I think the sacrifices required for 100% block/dodge/miss/parry will be small or non-existant. The actual gain from not ever taking a unmitigated hit will depend mostly on the encounter: how fast are the hits and how much of the burst damage is magical. I do believe that the gain will be significantly less than in the previous expansion for two reasons: the removal of crushing blows and the fact that all burst damage bosses seem to have their parry haste flag turned off.

 

turbo_012

Hi, excuse me because I'm not really a tank.

My main tank was talking about how he saw a post of another tank that reached 40% parry and generates alot of threat. My main tank said that was stupid because of DR and thus parry wasn't as important in WofLK.

What do you guys think of this?

 

Tamoa

DR mean that dodge, parry and block all diminish in effectiveness (over a long period of time) relative to armour and stamina, but in absolute terms they are still very important.

 

turbo_012

What do you mean "over a long period of time"?

Do you mean during a boss fight?

 

Branar

First, block is unaffected by Diminishing Returns.

Second, DR has nothing to do with time. You can read the initial post of this thread for a decent description of the effects of DR, but in brief: the more dodge/parry/miss rating you have, the less each additional point of dodge/parry/miss rating is worth. As your dodge goes up, it takes more and more dodge rating to get 1% to dodge.

I don't believe that some tank had 40% parry - he's lying or perhaps combining his parry and dodge chances. The cap for parry percentage is ~47% (for the tanking classes, anyway), which would require a theoretically infinite amount of parry rating to reach - there's just no way anyone has 40% parry at current levels of gear, and frankly I find it unlikely that people will reach it in this expansion.

 

Tamoa

Sorry, I wasn't clear - I put 'long' in there to indicate over the months of raiding as gear improves. Should have been clearer.

 

◊ Birdemani

Linking this here due to the relevance of the initial post. Whitetooth has done some testing and may have found the Miss Cap:

http://elitistjerks.com/1053037-post335.html

Relwin: Besides, the BB is not some ivory tower of WoW knowledge, it's just less stupid here than elsewhere.
DeeNogger: Not less stupid, better stupid. The BB takes stupid very seriously. Now if you will excuse me, I have to go misspell the word fire.