PDA

View Full Version : Why is parry rating bad?



swelt
06-08-2009, 05:05 AM
Before DR, parry rating returns less avoidance point for point than dodge rating.

Dodge 39.3~
Parry 49.2~
source (http://elitistjerks.com/f31/t29453-combat_ratings_level_80_a/)

This typically ends up with items with parry rating ending up looking worse than those with dodge rating. So "parry rating is bad".

But why? Blizzard haven't arbitrarily made up these or any other of their numbers, they have models for this kind of stuff. The answer is Parry Haste. "When you parry an attack, the remaining time on your current swing is reduced by 40% of your weapon speed, unless this would result in a reduction to less than 20% of your swing time remaining" WoWiki (http://www.wowwiki.com/Formulas:Parry) Links to this old thread: http://www.tankspot.com/forums/f14/37427-effective-weapon-speed-due-parry.html

Aside: I have a completely unsubstantiated personal theory that the roots of this are all in the animation. They want player models to look like they parry, so they have to adjust their next hit to fit that in, or something like that...

My real points are:
- If Blizzard thinks parry hasting is worth about 25% extra cost in the rating system are we undervaluing it? By looking soley at raw avoidance, are we doing parry rating a disservice?
- If not, why aren't we bitching about it to Blizzard to fix it? They adjusted the armor pen/haste stats because they were over budget for their returns, why not parry?

Satorri
06-08-2009, 05:48 AM
The reason people will tell you to stack dodge over parry is simple, but two-fold:
The community sees dodge and parry as the same goal, survival, avoidance, not getting hit. To that end Parry is in every way inferior. It has a less generous exchange rate as you referenced, it also diminishes to a bit more than half the value (dodge caps at 88% parry at 47%). To that end it's almost never possible for Parry to get you better survival value than Dodge, especially since tank gear has a rounded baseline on both stats.

Here's the other side though. While most people don't think of it in terms of a threat/survival hybrid stat (like Expertise), the value of parry haste is also a little more complicated, and limited for 2 of the 3 tanks who can parry. Warriors and Paladins usually use very fast weapons, though there was a spell there where there was a slower option. A fast weapon presents a smaller window and a smaller gain in swing speed per parry. The net effect *can* still be a buff to threat, but it can be hard to justify, and many people aren't big on half-stats, they want something pure and trackable.

Death Knights are designed to get more parry because it works very well with their gear style. Parry haste on a 2h with a 3.xx swing speed can noticeably increase your swing rate, though I have been meaning to actually do some measured tests and some math to go with it. That said, DKs, thanks to their class special, will have significantly higher parry rating than other classes, so they get a bump along the curve. That said they also get a few sources of flat % buffing to parry that doesn't affect diminishing returns, so they can sport better parry %'s more easily.

This may be the right time for math, but I won't have time to get to it this morning. =) I'll post back when I have time.

Vanthus
06-08-2009, 06:16 AM
Before DR, parry rating returns less avoidance point for point than dodge rating.



But why? Blizzard haven't arbitrarily made up these or any other of their numbers, they have models for this kind of stuff. The answer is Parry Haste. "When you parry an attack, the remaining time on your current swing is reduced by 40% of your weapon speed, unless this would result in a reduction to less than 20% of your swing time remaining" WoWiki (http://www.wowwiki.com/Formulas:Parry) Links to this old thread: http://www.tankspot.com/forums/f14/37427-effective-weapon-speed-due-parry.html

My real points are:
- If Blizzard thinks parry hasting is worth about 25% extra cost in the rating system are we undervaluing it? By looking soley at raw avoidance, are we doing parry rating a disservice?
- If not, why aren't we bitching about it to Blizzard to fix it? They adjusted the armor pen/haste stats because they were over budget for their returns, why not parry?

I've always been curious about this too, and look at parry as possibly an avoidance and threat stat at the same time. Kind of like a hybrid (purple, green, orange) gem, not for min/maxing, but a nice addition if you can afford it.
The number one issue is that it is not the best avoidance stat, and I would never use a gem slot for parry rating.
That being said, at some point there is merit for getting a higher threat stat at the cost of some avoidance. We all do that with hit rating and expertise already. Would be interesting to see really how much of an option this might be as I am sure it is overlooked by almost everyone.

My tanking spec is already different from just about anyone else, because I have found that haste has a lot more value to my DPS (especially rune striking) than other paths (such as longer diseases, etc).

swelt
06-08-2009, 07:17 AM
A very interesting parallel is with Expertise. We generally view expertise as a good threat stat, because it makes your attacks more likely to land. However there have been some good arguments put about for expertise as a damage avoidance/mitigation stat because it reduces the chance of *boss* parry haste.

VerticalEvent
06-08-2009, 07:32 AM
There's another difference between Dodge and Parry
You can dodge an attack from behind
You can only parry attacks ahead of you.

Which makes Dodge superior in a lot of trash encounters (Spider quarter in Naxx comes to mind) where you are flooded with mobs from all sides.

There's also parry hasting the boss. Unless your expertise hard capped, there's a chance that parrying an attack will make the next attack from the boss come faster.

Personally, I think, taking into account parry hasting, that Dodge and Parry should scale at roughly the same pace (maybe 4 parry = 3 dodge instead of the 4 parry = 2 dodge we have now). The fact we can't parry from behind should make up for the haste we gain from it.

Stephanius
06-08-2009, 07:39 AM
The nominal higher cost for parry avoidance is justified by it's threat component. But... parry gets hit much harder than dodge by Wrath's 100% avoidance countermeasure aka 'diminishing returns'. Because of this, parry's effective avoidance value to item budget cost ratio is bad compared to dodge.

Since you already get a certain amount of parry rating from direct gear properties, you are very likely to be already in the range where the diminishing returns on parry gems and enchants are a waste of gem and enchant slots.

Add on top that threat is not really a Wrath tanking issue and you get to the following:
TLDR: parry < dodge; don't enchant or gem for it.

Tharr
06-08-2009, 08:54 AM
There's another difference between Dodge and Parry
You can dodge an attack from behind
You can only parry attacks ahead of you.

No no no no no. Players can NEVER dodge attacks from behind. ONLY npc's can do that.

Just wanted to add that.

Tarigar
06-08-2009, 09:09 AM
TLDR: parry < dodge; don't enchant or gem for it.

I believe there is one exception to the rule. Bladeward or the DK equivalent. I don't go for gear with parry usually I would rather have dodge/expertise/hit/stamina. So I made up for the lack of parry by enchanting for bladewarding. It bumps me up over 19% with one proc.

As always my armory is in my sig.

Lizana
06-08-2009, 11:30 AM
I believe there is one exception to the rule. Bladeward or the DK equivalent.

DK's runes on their weapons are diffrent.. but you still should not enchant bladeward due to the crap proc rate

Superspy23
06-08-2009, 01:05 PM
So basically we need to quantify the average threat increase for white attacks and HS per 1% of parry to actually gauge its effect properly. Then compare that to the threat increase from the same stat budget given to hit and expertise.

It could be that parry gems would be a viable way to increase threat after you're hit and exp caped. Of course we could also come to realize that it really is poorly alloted budget wise for its threat component as well as how terribly it shows to be for avoidance; and even adding together the net gain still yields sub par performance.

For instance if parry rating is worth half the same amount of dodge and is also worth half the same amount of hit then it legitimately evens out and is a fair stat allocation.

Lizana
06-08-2009, 01:07 PM
I have yet to meet a tank that is expertise capped, hit capped and at the defensive mins...

Muffin Man
06-08-2009, 01:10 PM
Someone did actually calculate the threat benefit of parry. But that was back in TBC where 1) no DR's so you could actually get pretty high parry wise and 2) tank melee swings did crap damage.

Kerg
06-08-2009, 01:11 PM
Yeah, I would like to see an updated comparison of Parry vs. Expertise in terms of pure threat. I think that is the only way we can make an informed argument to Blizz that Parry Rating is overvalued on gear. Any takers? :)

Lizana
06-08-2009, 01:16 PM
By the same logic that due to the parry haste it is a threat increase, it also increased the chance that the boss will parry you during the encounter.

Superspy23
06-08-2009, 01:16 PM
I have yet to meet a tank that is expertise capped, hit capped and at the defensive mins...

Let me introduce you to one Zuggi (http://www.wowarmory.com/character-sheet.xml?r=Azshara&n=Zuggi). Admittedly he's only soft caped for exp and he gets the rest of his crit immunity from his resilience helm enchant. Its not unreasonable to get these stats and be survivable as well.

Here's another one Domichi (http://www.wowarmory.com/character-sheet.xml?r=Azshara&n=Domichi).

Tarigar
06-08-2009, 01:19 PM
DK's runes on their weapons are diffrent.. but you still should not enchant bladeward due to the crap proc rate

I am running 20-25% uptime on bladeward on average. To me that is considerable if you compare it to a darkmoon card that runs about the same.

Lizana
06-08-2009, 01:24 PM
Let me introduce you to one Zuggi (http://www.wowarmory.com/character-sheet.xml?r=Azshara&n=Zuggi). Admittedly he's only soft caped for exp and he gets the rest of his crit immunity from his resilience helm enchant. Its not unreasonable to get these stats and be survivable as well.

Here's another one Domichi (http://www.wowarmory.com/character-sheet.xml?r=Azshara&n=Domichi).

Soft cap is not capped...

Lizana
06-08-2009, 01:34 PM
I am running 20-25% uptime on bladeward on average. To me that is considerable if you compare it to a darkmoon card that runs about the same.

You may have 1 stack of it up 25% of the time, but look at the number of parries you actual get from it. From you last log you posted with you MTing for almost 4 hours... you had 38 parries while you had blade ward up... 38 for 4 hours isnt that much.

2 on Razorscale
1 on ignis
2 on your X002 kill
1 on iron council
1 on your kolo kill


7 parries gained on 5 fights. And actually we don't really know if those parries were gained at all.. Because you base parry chance could have caused a parry when you had the buff up in the first place, causing the extra parry rating to be useless.

squats
06-08-2009, 01:36 PM
blizzard has put a heavy DR on parry because it not only serves as a survival stat, but a DPS/TPS increase. high amount of pary will increase your TPS also, due to the swing haste you get from parying.

f14flier7
06-08-2009, 01:37 PM
I am able to get "hard" expertise capped in my Threat set (I lose minimal HP compared to my EH set about 200HP out of 33k HP and a little avoidance). I too have wondered Why Parry is looked down upon SOOOO heavily if I am "hard" expertise capped?

Tarigar
06-08-2009, 01:39 PM
You may have 1 stack of it up 25% of the time, but look at the number of parries you actual get from it. From you last log you posted with you MTing for almost 4 hours... you had 38 parries while you had blade ward up... 38 for 4 hours isnt that much.

2 on Razorscale
1 on ignis
2 on your X002 kill
1 on iron council
1 on your kolo kill


7 parries gained on 5 fights. And actually we don't really know if those parries were gained at all.. Because you base parry chance could have caused a parry when you had the buff up in the first place, causing the extra parry rating to be useless.

Ok XT hits like a truck so out of those 2 parries that is ~40k dmg. Yes it could have been my base parry that caused it but you could say the same about mongoose. So do you believe that blood draining would have healed me for 40k in that XT fight?

Lizana
06-08-2009, 01:40 PM
You have 64 expertise skill and are hit capped at the same time, while still having 540 def? I would love to see your gear setup

Dubzil
06-08-2009, 01:43 PM
I was exp and hit capped w/ 550~ def pre-ulduar, have changed the gear setup since for survival reasons.

squats
06-08-2009, 01:46 PM
I am able to get "hard" expertise capped in my Threat set (I lose minimal HP compared to my EH set about 200HP out of 33k HP and a little avoidance). I too have wondered Why Parry is looked down upon SOOOO heavily if I am "hard" expertise capped?
you are almost hit capped, however you arnt anywhere near expertise capped.

Lizana
06-08-2009, 01:51 PM
Ok XT hits like a truck so out of those 2 parries that is ~40k dmg. Yes it could have been my base parry that caused it but you could say the same about mongoose. So do you believe that blood draining would have healed me for 40k in that XT fight?

No, but unless your healers are re actively healing ( and in ulduar they wont be) you were healed for that 40k anyways.. so all you did was give your healers more overhealing.

From my wws i show blood reserve healed me for almost 6k when i needed a heal the most. Can you say the same for you parrys?

Superspy23
06-08-2009, 01:53 PM
you are almost hit capped, however you arnt anywhere near expertise capped.

He was referring to his threat set. The set you're looking at is probably not his threat set.

Also don't forget that after you remove dodge from the table expertise becomes much devalued. Something to consider when we are able to evaluate the threat value of parry haste vs. exp pre or post "soft" cap.

Lizana
06-08-2009, 01:54 PM
He was referring to his threat set. The set you're looking at is probably not his usual set.

Also don't forget that after you remove dodge from the table expertise becomes much devalued. Something to consider when we are able to evaluate the threat value of parry haste vs. exp pre or post "soft" cap.


Your right it becomes devalued to the exact same amount of effective hit as hit does.. and hits still a huge threat stat...

Tarigar
06-08-2009, 01:55 PM
No, but unless your healers are re actively healing ( and in ulduar they wont be) you were healed for that 40k anyways.. so all you did was give your healers more overhealing.

From my wws i show blood reserve healed me for almost 6k when i needed a heal the most. Can you say the same for you parrys?

To say when those parries were timed no.. Those parries could have come at the time I was needing a heal.. Right now we are both boiling down to hypotheticals. It could have done this it could have done that. But the truth is those two parries saved me 40k health while your blood draining gave you 6k health.

The only thing we know is your blood draining proc'd once and my bladeward caused a parry twice. Since that is the design behind bladeward similar to the Wrath card in TBC. It was designed to force a crit.

Lizana
06-08-2009, 02:22 PM
Okay.. just used your own combat log for your X002 fight. Your two parries caused 35793 in overhealing to you. So you effectively wasted your healers mana... So your effective gain.. 5k hp "saved" from blade ward... were as blood drain healed 6k... i think 6 > 5

Muffin Man
06-08-2009, 02:23 PM
I am able to get "hard" expertise capped in my Threat set (I lose minimal HP compared to my EH set about 200HP out of 33k HP and a little avoidance). I too have wondered Why Parry is looked down upon SOOOO heavily if I am "hard" expertise capped?

Just looking at your armory, I'd like to see this set up as well. Since 200 HP is ~20 stamina, and you can someone trade that for over 200 expertise rating?

I would have assumed the first step of changing to expertise capping was swapping one of your trinkets to Mark of Norgannon... but that would drop you at least 1000 HP

Tarigar
06-08-2009, 02:33 PM
Ok, What you are saying then is we shouldn't stack dodge or Parry because our healers will be overhealing.

Also with receiving 20k blows from XT and 41k hp raid buffed.

Lets do some math
for me to proc blood draining I would have to drop below 14350 and it will never be right there so lets say I hit 13k to be fair and that is 31.7% of my health. So with the 6k heal from blood draining it would bump me up to 19k. XT hits me again with no heal incoming which does happen I am still dead.

I still really don't see the benefit from wanting something to proc when i am low health. When i would rather have it keep me from getting there anyways. Hence where avoidance comes into play.

Dubzil
06-08-2009, 02:35 PM
Ok, What you are saying then is we shouldn't stack dodge or Parry because our healers will be overhealing.

Also with receiving 20k blows from XT and 41k hp raid buffed.

Lets do some math
for me to proc blood draining I would have to drop below 14350 and it will never be right there so lets say I hit 13k to be fair and that is 31.7% of my health. So with the 6k heal from blood draining it would bump me up to 19k. XT hits me again with no heal incoming which does happen I am still dead.

I still really don't see the benefit from wanting something to proc when i am low health. When i would rather have it keep me from getting there anyways. Hence where avoidance comes into play.


But what if you have HoTs on you, don't get a heal, but you get 2k heals from HoTs, then ur still alive and that heal the healers were trying to get off on you before you died goes off shortly after putting you to max hp

Tarigar
06-08-2009, 02:41 PM
what if

Isn't that the problem with both of these enchants? What if I was at 6k.. given your scenario I am still dead.. What if I took 3 25k hits back to back... Dead

We are really just trying to convince why the other why one is better.

I feel more comfortable with the what if.. that my 20-25% chance to have my parry increased will help increase my survivability. Where someone with Blood Draining feels more comfortable with the what if.. that 6k heal will keep me alive long enough to get another heal.

Dubzil
06-08-2009, 02:43 PM
ok, well take the what if out, and just say you have a resto druid, you WILL have the HoTs on and it WILL save you. I haven't looked at blade warding recently, but was under the impression it granted 4? 5?% parry.

Lizana
06-08-2009, 02:45 PM
Ok, What you are saying then is we shouldn't stack dodge or Parry because our healers will be overhealing.

Also with receiving 20k blows from XT and 41k hp raid buffed.

Lets do some math
for me to proc blood draining I would have to drop below 14350 and it will never be right there so lets say I hit 13k to be fair and that is 31.7% of my health. So with the 6k heal from blood draining it would bump me up to 19k. XT hits me again with no heal incoming which does happen I am still dead.

I still really don't see the benefit from wanting something to proc when i am low health. When i would rather have it keep me from getting there anyways. Hence where avoidance comes into play.


You should never expect a heal not to come in... It is the healers job to heal. you should always expect for there to be a heal incoming. Your job is to live long enough for that heal to get there. The blood drain heal isnt supposed to save you by itself.. but instead it gives the healers a buffer. As a good example of healers. Useing your own damage numbers and the heal numbers from the log.

your max hp 41k. You get hit to that 13k. Blood drain heals you for 2k, Flash of light heals you for 6k, 20 k hit lands you live. Holy light hits you for 15k and another flash lands and brings you above 20k again. Thats just 2 healers healing you. (yes and one of your pallies was FOL spaming)

Now take that blood drain heal out, now you are at 13 k, you get that 6k heal your at 19k, the holy light has .01 seconds left till it heals you boss hits you and your dead...

And yes avoidance is nice, but an avoidance proc imo isnt becuase your hopeing something will proc that will give you the chance something else will proc...

Its like me telling you roll a dice, if you get a 1 you get a chance to roll another dice to see if you can roll a 1, and only if you do that twice in a row will you be able to get a prize..

Muffin Man
06-08-2009, 02:46 PM
To everyone arguing bladewarding: http://www.tankspot.com/forums/f97/48831-blade-warding-analysis-what-proc-makes-better.html

Here are some good old fashioned numbers.

Dubzil
06-08-2009, 02:56 PM
Ty Muffin Man

So from the post we have a proc chance of: 4.36%
Which will grant you 3.9% parry

omm no thanks.

additionally for threat:
"we see that the expected damage (and consequently, added threat) will be equivalent if Blade-Warding has a proc chance of approximately: 27.2112%" <- this is versus accuracy

swelt
06-08-2009, 02:59 PM
I think there are probably enough threads about blood draining vs blade ward on the forum that we don't need to turn this one into the same debate?

Could we instead try to establish how much (if any of consequence) threat is gained from parry rating?

Lizana
06-08-2009, 03:00 PM
I think there are probably enough threads about blood draining vs blade ward on the forum that we don't need to turn this one into the same debate?

Could we instead try to establish how much (if any of consequence) threat is gained from parry rating?

Depends when you parry in relation to your swing timer and the speed of your weapon, the dps of the weapon.

Tarigar
06-08-2009, 03:08 PM
And yes avoidance is nice, but an avoidance proc imo isnt becuase your hopeing something will proc that will give you the chance something else will proc...

Wasn't that what Mongoose was all about the proc on the weapon increasing your dodge?

Lizana
06-08-2009, 03:12 PM
Wasn't that what Mongoose was all about the proc on the weapon increasing your dodge?

Agi increases armor, dodge, and crit chance, also it provided a consistent haste increase. And was not affected by DR at the time it was popular.

Muffin Man
06-08-2009, 03:12 PM
Mongoose had a *great* uptime though, and very little competition (for survivability anyways).

Tarigar
06-08-2009, 03:18 PM
It is gonna take a lot to convince me otherwise but then I am one of those stubborn jackass tanks.

Dubzil
06-08-2009, 03:19 PM
a lot more than ~4% proc rate that grants ~4% parry ? seems like no-brainer to me

Tarigar
06-08-2009, 03:21 PM
What is the heal percentage per fight on Blood Draining in comparison total?

Tarigar
06-08-2009, 03:25 PM
Also if you read the auryiaya report it proc'd 9 out of my 44 parries. 20% proc rate on that fight :). So somefights it will be higher than others.

Vanthus
06-08-2009, 03:25 PM
What is the heal percentage per fight on Blood Draining in comparison total?

This thread is not about Blood Draining or any weapon enchant proc's. It is about Parry rating, and how that translates to haste. How one goes about getting more parry rating or parry % is completly un-related to the topic at hand.

Dubzil
06-08-2009, 03:27 PM
well depends on the fight really, I tested it just on the dummy for proc rate, and i refreshed my 5 stack 14 times before it didn't proc and dropped off, so it will be 2.2k heal on any fight that you are constantly attack in a boss. On a fight like Yogg P1, it will usually proc with a 1-2 stack since you don't get that many hits in on the adds before they die, but still heal ~800.

From Wednesday night's raid: World of Logs - Real Time Raid Analysis (http://www.worldoflogs.com/reports/7wgZfJIz53o2SFUv/details/21/?s=0&e=14749)

I have: Healing done: 72508 / 52 hits / Avg 1394.4 / .6% of all heals taken

over the course of 8 bosses

f14flier7
06-09-2009, 04:41 AM
Just looking at your armory, I'd like to see this set up as well. Since 200 HP is ~20 stamina, and you can someone trade that for over 200 expertise rating?

I would have assumed the first step of changing to expertise capping was swapping one of your trinkets to Mark of Norgannon... but that would drop you at least 1000 HP

Gah, you're right... I forgot to put another zero on the end of that...

And yes, Armory has my EH set on, I would have to regem and such to get back to my Naxx "Threat Set"

Kerg
06-09-2009, 08:03 AM
To all the Bladeward vs. Blood Draining people.... quit shitting in this thread about Parry Rating. Take it to another thread. Now...

How much threat does Parry Rating give you? What we need is for someone to come up with a model that would give us an idea. Because right now, most people don't even consider it a threat stat at all. But apparently Blizzard does, and significantly. Therein lies the problem.

Superspy23
06-09-2009, 11:15 AM
To all the Bladeward vs. Blood Draining people.... quit shitting in this thread about Parry Rating. Take it to another thread. Now...

How much threat does Parry Rating give you? What we need is for someone to come up with a model that would give us an idea. Because right now, most people don't even consider it a threat stat at all. But apparently Blizzard does, and significantly. Therein lies the problem.

I agree, lets get back on topic.

Things we know:

Parry haste will only affect white damage and HS as well as increase rage production due to more white hits.

Additionally the old model is outdated and needs to be reworked.

Parry haste is based on time remaining to your swing cycle and will give different haste values based on when in that cycle you achieved the parry. Thus it will be effected by the RNG and give an inconsistent increase in haste/threat.

What we can model is the range of numbers from the varied haste increases and then formulate the average. Then we can take this average and apply it to your parry percentage. Therefor giving us the non enchant related numbers that this thread is actually about.


PS. My other post must have been deleted or something. I blame the government... or my tongue n cheek response may have been a little over the top. Anyhow this response has a little more to it than the last one in regards to the thread topic.

Martie
06-09-2009, 11:33 AM
Correct me if I'm wrong here - but here is how I've read how much swing speed reduction parry gives.

A parry reduces your swing timer by 40% if it happens during the first 40% of your swing timer.
A parry reduces your swing timer by 0% if it happens during the last 20% of your swing timer.
A parry reduces your swing timer to 20% if it happens between the first 40% and the last 20% of your swing timer.

So the first 40% is 40% faster. (0.6 * weapon speed)
The second 40% is, on average, 20% faster. (0.8 * weapon speed)
The last 20% isn't any faster. ( 1 * weapon speed)

So, on average, your parry will make your hit be 100* (1-(0.4 * 0.6 + 0.4 * 0.8 + 0.2 * 1) = 24% faster.

So 1% parry is worth 0.24% white swing haste.

So, take whatever threat you get from your white damage + your heroic strikes and multiply it by 1.0024 to know your threat gains from 1% parry.
Take your threat generation from white swings and multiply it by 1.0024 to get your rage gains from 1% parry.
If you use abilities that can only be used after a parry, those can be used 1% more often too.

Seems pretty easy to me.

Superspy23
06-09-2009, 12:18 PM
Correct me if I'm wrong here - but here is how I've read how much swing speed reduction parry gives.

A parry reduces your swing timer by 40% if it happens during the first 40% of your swing timer.
A parry reduces your swing timer by 0% if it happens during the last 20% of your swing timer.
A parry reduces your swing timer to 20% if it happens between the first 40% and the last 20% of your swing timer.

So the first 40% is 40% faster. (0.6 * weapon speed)
The second 40% is, on average, 20% faster. (0.8 * weapon speed)
The last 20% isn't any faster. ( 1 * weapon speed)

So, on average, your parry will make your hit be 100* (1-(0.4 * 0.6 + 0.4 * 0.8 + 0.2 * 1) = 24% faster.

So 1% parry is worth 0.24% white swing haste.

So, take whatever threat you get from your white damage + your heroic strikes and multiply it by 1.0024 to know your threat gains from 1% parry.
Take your threat generation from white swings and multiply it by 1.0024 to get your rage gains from 1% parry.
If you use abilities that can only be used after a parry, those can be used 1% more often too.

Seems pretty easy to me.


So each percentage of parry is only worth roughly a quarter percent of haste overall? So it seems that most tanks will be getting 4% to 4.8% more threat from white attacks and HS overall.

Unless I'm missing something it doesn't seem that great. Needing roughly 4% parry just to get 1% more haste doesn't seem to balance out the deficient rating conversion relative to dodge. I'll have to compare the parry rating to haste value compared to haste's rating to value when I have time unless someone else wants to do that comparison.

Lizana
06-09-2009, 12:25 PM
So each percentage of parry is only worth roughly a quarter percent of haste overall? So it seems that most tanks will be getting 4% to 4.8% more threat from white attacks and HS overall.

Unless I'm missing something it doesn't seem that great. Needing roughly 4% parry just to get 1% more haste doesn't seem to balance out the deficient rating conversion relative to dodge. I'll have to compare the parry rating to haste value compared to haste's rating to value when I have time unless someone else wants to do that comparison.

Not everything has to be a balanced stat.

Superspy23
06-09-2009, 12:41 PM
Not everything has to be a balanced stat.

Of course. Thats a given. However... "Now we know, and knowing is half the battle. Go Joe!!!"

I would however like to hear some more opinions and input on the value of parry and parry haste threat. Specifically at what gear level does it make the poor parry rating budget less poor or even worth it.

Molohk
06-09-2009, 12:45 PM
So, on average, your parry will make your hit be 100* (1-(0.4 * 0.6 + 0.4 * 0.8 + 0.2 * 1) = 24% faster.

So 1% parry is worth 0.24% white swing haste.

1% parry doesn't mean you'll get 1% of your hits hasted, it means you will parry 1% of the incomming hits. So the equivalent haste is proportional to the relation between the attacker's swing speed and your own.

If the boss hits twice as fast as you (imagine a 1.5 swing speed for the boss and a 3.0 swing speed for yourself) With 1% parry, you are likely to haste 2% of your swings, or a little less than 2% to be precise because of the chance that you may parry twice before your next swing, effectively "wasting" parry hastings when back-to-back parries occur. Note the chance to have back-to-back parries increases with your total chance to parry, so it could act as diminishing returns for parry haste.

I don't have time to do the actual math on back-to-back parry probability, but of the top of my head: with a 2:1 speed ratio, out if you swing 100 times and the boss swings 200 times, there's a 100/39800 or 1/398 chance of 2 back-to-back parries occuring between 2 of your swings. It's not very significative at low parry %s but this probability should increase exponentialy as you increase your parry chance.

Satorri
06-09-2009, 02:25 PM
More swings for a warrior means more haste.
More swings for a pally means more chances for seal stacks/procs.
More swings for a bear means more chances for SD procs.
More swings for a DK means easier used RS's, and the slower weapons get a more significant swing speed increase on absolute value (though arguing % would require math).

Whether or not it is parried is a valid question. Blizz did design it with the idea that it would be balanced, they didn't say, "hey, screw parry! Let's put it on all their gear but not make it worth having!"

If a stat is not balanced but it is getting no chatter from players or Blizz, then chances are there are aspects you are just missing. Not understanding or being able to account for how it works goes a long way to people not wanting to take it though.

Superspy23
06-09-2009, 02:43 PM
1% parry doesn't mean you'll get 1% of your hits hasted, it means you will parry 1% of the incomming hits. So the equivalent haste is proportional to the relation between the attacker's swing speed and your own.

If the boss hits twice as fast as you (imagine a 1.5 swing speed for the boss and a 3.0 swing speed for yourself) With 1% parry, you are likely to haste 2% of your swings, or a little less than 2% to be precise because of the chance that you may parry twice before your next swing, effectively "wasting" parry hastings when back-to-back parries occur. Note the chance to have back-to-back parries increases with your total chance to parry, so it could act as diminishing returns for parry haste.

I don't have time to do the actual math on back-to-back parry probability, but of the top of my head: with a 2:1 speed ratio, out if you swing 100 times and the boss swings 200 times, there's a 100/39800 or 1/398 chance of 2 back-to-back parries occuring between 2 of your swings. It's not very significative at low parry %s but this probability should increase exponentialy as you increase your parry chance.

I think you're reading it wrong. Nowhere does it say that 1% parry gives 1% haste. What it says is that each parry hasted attack averaged haste is 24%. Furthermore its multiplied by 100 to make that equate into the roll system so its relevant to the percentage of parry you can see on your character sheet. Therefor 1% parry approximately equals 1/4th of one percent (0.24%). Its plainly written in the portion you quoted.
1% parry is worth 0.24% white swing haste.
Notice that 0.24% is absolutely not the 1% you seem to have read.

As to your reference to boss swing speeds, that is unchanged due to this being a percentage and therefore scales with the number of hits you receive and yields the same percentage of hasted attacks.

Superspy23
06-09-2009, 02:58 PM
the slower weapons get a more significant swing speed increase on absolute value (though arguing % would require math)

How can a slow swing speed weapon gain more from parry haste than a fast swing weapon? Its not as if the speed of the weapon has any relationship to the amount of time spent in the the first 40%, second 40% or last 20% of the swing.

IE. (speeds chosen for ease of comparison)

A 3.0 speed weapon (in a 3 second period) would spend 1.2 seconds in the 40% haste zone, 1.2 seconds in the 20% haste zone, and 0.6 seconds in the 0% haste zone.

A 1.5 speed weapon (in a 3 second period) would spend 1.2 seconds in the 40% haste zone, 1.2 seconds in the 20% haste zone, and 0.6 seconds in the 0% haste zone.

Am I not understanding something about your statement? Why the declaration of inequality or superiority?

Muffin Man
06-09-2009, 03:03 PM
Satorri is correct, if you count absolute time taken off per parry haste, a slow weapon will have more taken off... but percentage wise it's the same.

However, this distinction does matter because the fewer swings you have, the more of them are likely to be parry hasted.

This is because your benefit from parry haste is a function of your parry % and *the boss's* swing speed.

So, you're more likely to get your swings hasted if your swing timer is equal or slower than the boss's swing timer (see Molohk's post above for more details).

Martie
06-09-2009, 03:18 PM
1% parry doesn't mean you'll get 1% of your hits hasted, it means you will parry 1% of the incomming hits. So the equivalent haste is proportional to the relation between the attacker's swing speed and your own.
<snips good explanation>
Yer right, that's an oversight on my part.

So if the mob hits faster then you, you'll get a little more haste, if it hits slower then you, you'll get a little less.
As far as memory serves, most bosses hit at 2-3 sec timers, so DKs should get a little more haste from parry while warriors and paladins get a little less.

Satorri
06-10-2009, 03:06 PM
I did a work up on this a long time ago in the Demystifying the Myths of Dual Wield Tanking thread. It doesn't work out to be quite so clean and (depending on spec, rotation, etc) your expertise has a bigger effect on how much parry liability you have than class does.

Molohk
06-10-2009, 03:24 PM
How can a slow swing speed weapon gain more from parry haste than a fast swing weapon? Its not as if the speed of the weapon has any relationship to the amount of time spent in the the first 40%, second 40% or last 20% of the swing.

IE. (speeds chosen for ease of comparison)

A 3.0 speed weapon (in a 3 second period) would spend 1.2 seconds in the 40% haste zone, 1.2 seconds in the 20% haste zone, and 0.6 seconds in the 0% haste zone.

A 1.5 speed weapon (in a 3 second period) would spend 1.2 seconds in the 40% haste zone, 1.2 seconds in the 20% haste zone, and 0.6 seconds in the 0% haste zone.

Am I not understanding something about your statement? Why the declaration of inequality or superiority?

That's because the number of your swings that will get parry-hasted is the same if you have a slow weapon or fast weapon (with the small variation I mentioned, from back-to-back parries in between swings). If you have 20% parry chance and the boss is hitting you every 2 seconds, you are likely to get one parry haste every 10 seconds regardless of your own weapon's speed, which translates into 6 parry-hasted swings per minute. If you gained an average of 24% haste in each of the parry-hasted swings, during the whole minute (6 parry-hasted swings) you gained a total of 1.44 white swings (again, regardless of your own weapon speed), and obviously if you have a slower that 1.44 swing hits harder than if you have a faster weapon speed. This is what I was explaining: the slower your weapon the larger % of your own hits that gets parry-hasted, therefore you gain more benefit from the haste.

In short: you gain more damage from parry-haste with a slower weapon, and you also gain more damage from parry-haste the faster the boss is hitting you, because it's the proportion of both speeds that matters.

Superspy23
06-10-2009, 03:27 PM
I did a work up on this a long time ago in the Demystifying the Myths of Dual Wield Tanking thread. It doesn't work out to be quite so clean and (depending on spec, rotation, etc) your expertise has a bigger effect on how much parry liability you have than class does.

I read that back then. I thought that was about boss parries resulting in a parry gib because of hitting the boss more frequently. This thread is about player parries and the resulting threat increase from our parry haste, not parry liability.

Superspy23
06-10-2009, 03:33 PM
Here I is answering your question.

I don't know why I was having such a hard time wrapping my mind around that. Thanks. I think I understand now.

Satorri
06-11-2009, 04:42 PM
Yes, Super, the implications are the same though. Our increase in threat correlates with the increase in damage taken from being parried. Kind of a no-duh statement, but...

hitristan
06-14-2009, 11:43 PM
Parry also works on spells, it's called deflect.

Let's say you're tanking KT, you see him start to cast frost bolt, you have 1k hp left, and you have nothing else do do but wait and see if your healers can get to you in time.

A 100% dodge chance would not help you in this situation at all.

However, if you deflected this spell (deflection is based on parry rating), you would live.

Parry is an excellent addition to spell mitigation.

DK's were designed to be the best caster tanks, and in addition to death grip, their high parry chance allows them to trump most other tanking classes in caster tanking.

Kazeyonoma
06-15-2009, 09:11 AM
Can you link confirmation that deflect is based on parry rating cuz I've never read that anywhere.

Superspy23
06-15-2009, 10:53 AM
Can you link confirmation that deflect is based on parry rating cuz I've never read that anywhere.

Deflect - WoWWiki - Your guide to the World of Warcraft (http://www.wowwiki.com/Deflect)

This link has a little about deflect and mentions that parry gems increase the chance. Apparently deflects aren't usually shown by the UI and there's a macro to get the deflects to show. However when you read about deflect in the combat page it says "Deflect -This is an outcome which has been found by looking to GlobalStrings and CombatFeedback lua files, in Blizzard's UI (FrameXML), which might be added later by Blizzard to possible results for auto-attacks and physical abilities." Which implies that the mechanic isn't in game. However the Deflect page seems to have a screen shot showing a deflect implying that it is in game.

Before reading this deflect thing that hitristan posted I was totally unaware of this mechanic. I thought I was well versed on these type of mechanics. I had to look it up. This could be a huge thing in terms of the avoidance value of parry. I'm interested in learning more. Imagine the change on encounters like Hodir and other magic heavy fights.

So how much will parry give in terms of granting deflection?

Tarigar
06-15-2009, 10:59 AM
Deflect - WoWWiki - Your guide to the World of Warcraft (http://www.wowwiki.com/Deflect)

This link has a little about deflect and mentions that parry gems increase the chance. Apparently deflects aren't usually shown by the UI and there's a macro to get the deflects to show. However when you read about deflect in the combat page it says "Deflect -This is an outcome which has been found by looking to GlobalStrings and CombatFeedback lua files, in Blizzard's UI (FrameXML), which might be added later by Blizzard to possible results for auto-attacks and physical abilities." Which implies that the mechanic isn't in game. However the Deflect page seems to have a screen shot showing a deflect implying that it is in game.

Before reading this deflect thing that hitristan posted I was totally unaware of this mechanic. I thought I was well versed on these type of mechanics. I had to look it up. This could be a huge thing in terms of the avoidance value of parry. I'm interested in learning more. Imagine the change on encounters like Hodir and other magic heavy fights.

So how much will parry give in terms of granting deflection?

I have seen melee abilities glance. I have never seen a spell deflect.

I have dodged during frozen blows before but what I dodged I didn't stop to double check as it may have been his melee swing. I have had spells miss though but deflect is a new one.

Are we sure this isn't theory based. Does anyone have any reports to show a deflection?

Molohk
06-15-2009, 11:22 AM
It's funny that the only evidence of deflection posted in WoWWiki comes from the original alpha build. It sounds to me like a mechanic that was removed possibly before the original WoW launch. Maybe that's what inspired the Spell Deflection - Spell - World of Warcraft (http://www.wowhead.com/?spell=49497) talent in DKs, since it's not at all unusual to see devs resurrect ideas this way.

Superspy23
06-15-2009, 11:54 AM
It's funny that the only evidence of deflection posted in WoWWiki comes from the original alpha build. It sounds to me like a mechanic that was removed possibly before the original WoW launch. Maybe that's what inspired the Spell Deflection - Spell - World of Warcraft (http://www.wowhead.com/?spell=49497) talent in DKs, since it's not at all unusual to see devs resurrect ideas this way.

OK so some form of deflection is in game but possibly only for DK's with this one specific talent. Are we sure there's no other form of it in game? I'm thinking probably not. How can we miss such a mechanic for so long... I suppose it's possible.

jere
06-15-2009, 01:27 PM
In the various WWS parses I have taken in the past, my deflection % of Hammer of the Righteous closely follows my parry % of melee attacks, with various levels of expertise.

Lizana
06-15-2009, 01:29 PM
Just because a boss does it doesnt mean players can...

Martie
06-15-2009, 01:47 PM
Just because a boss does it doesnt mean players can...
This brings something to mind you all should know.
"Quod licet Iovi non licet bovi"

It's a lovely saying, usable in a ton of fun situations.
(And guaranteed to confuse.)

Molohk
06-15-2009, 02:03 PM
As Lizana said, deflects do exist, it is one of the types of SPELL_MISS in WOW's API. But that doesn't mean players can deflect. I've done a good ammount of log analysis, and I've never seen a player deflect. Especially with the parry rates that tanks now have, considering deflection is supposed to be related to parry chance, I'm greatly inclined to thinking players cannot deflect.

For obvious reasons, it is harder (if not impossible) to demonstrate that something doesn't exists, so if anyone has any evidence of player deflections actually ever occuring, please speak up. Until said evidence arises, it is safe to assume that players cannot deflect (excluding the DK talent).

Satorri
06-15-2009, 03:48 PM
I've watched this mechanic. Players do not parry spells without the DK talent Spell Deflection (which still doesn't show up as a "deflect").

The places where you will see deflect is with ranged/'melee' abilities and the Hunter CD Deterrence (which can deflect spells).


The reason DK's used to be the best anti-caster/magic tanks is because we had the biggest innate magic damage reduction. The only reason we are now is Anti-magic Shell. It is possible to make a significant anti-magic build, but otherwise I'd put us right on par with Warriors and Pallies (Unholy can get a 6% spell dmg reduction to match both those classes, Frost can get Acclimation which is a little harder to track the value of, but works quite well, and Blood's Spell Deflection is a nice value, though it doesn't amount to much better than the 6% magic damage reduction in the grand average and is a little less consistant, ewww RNG). Anti-magic shell is delicious, but that has also been nerfed to keep it from being too much of an advantage or a way to cheat mechanics (I used to love that as Unholy I could be immune for 5 sec).