[WW] 5.2 -Enter the Fist - Page 10

gahddo · 01/14/13, 3:38 PM

We just recently got the full extent of the CB:TP bug into simc, it correctly reduces chi tracking during CB:TP uptime. Our values on haste do seem extremely high and I'm looking into how the sim is playing currently to see if its realistic. The sim is currently running with an avg energy level of about 40, which just seems off to me.

tastysnack · 01/14/13, 3:47 PM

Originally Posted by Crevan

I might be misunderstanding your point here, but 3000 mastery rating per 1% is what it should be. Six hundred mastery rating is generally one "point" of mastery, and it gives 0.2% increase per stack of bottled fury, so 3000 rating will provide 5 times that, or 1%.

Is it? Perhaps I need to remath - all of my math and research showed that 429 rating = 1% of mastery.

Saltycracker · 01/14/13, 6:36 PM

Originally Posted by tastysnack

Is it? Perhaps I need to remath - all of my math and research showed that 429 rating = 1% of mastery.

If you log onto the PTR, your mastery % is very low in comparison to live. I think they changed the coefficient.

+0.2% / +1% =|= 429/3000

tastysnack · 01/14/13, 8:36 PM

Originally Posted by Saltycracker

If you log onto the PTR, your mastery % is very low in comparison to live. I think they changed the coefficient.

+0.2% / +1% =|= 429/3000

Got it - I tried logging into PTR to see if it had changed, but couldn't connect to it.

I will update the guide accordingly.

gahddo · 01/15/13, 2:01 AM

In the process of an overhaul on the windwalker priority list, should have some interesting results that change how a lot of people think about the rotation.

Saltycracker · 01/15/13, 2:57 AM

Updated my spreadsheet:

https://docs.google.com/spreadsheet/...ENHVmSHc#gid=6

Notes:

-Ability priority: RSK > FOF > BOK > TP
-Haste to GCD Cap (~35% haste) -> Mastery ~ Crit >> Haste.
-More haste increases the value of mastery until BOK cap; haste double dips in this manner and exponentially improves damage.
-Additional mastery increases damage at a linear rate. Simply 1% increase in mastery directly translates to 1%*TEB/4 in damage due to the monk generating the same number of TEB stacks. Each stack just does more damage.
-Crit is crit. It makes all your damage hit harder, but does not interact with any other secondary stat. There is a crit floor, but we can't give up enough crit rating or agi to get to it. Just a nice to have stat now, but much much better than the old mastery any day.
-Ascension still better than power strikes until you are past BOK cap (45% haste)

Just wondering how to treat jab going forward. It's just barely a significant source of damage, but enough to change your rotation around it? Is haste capping even a worry since Jab might as well be a "Push me to get Chi" button instead of an attack. Also, those pesky CB:TP procs screwing up our TEB generation. Just a thought. Maybe a job for simcraft.

I'll work in the T15 bonuses soon, they are pretty easy to math in. Looks like with the 2pc and 4pc, you can start stacking mastery much earlier (~29% haste) as the two secondary stats become equivalent in value. Neato.

gahddo · 01/15/13, 3:52 AM

Originally Posted by Saltycracker

Updated my spreadsheet:

https://docs.google.com/spreadsheet/...ENHVmSHc#gid=6

Notes:

-Ability priority: RSK > FOF > BOK > TP
-Haste to GCD Cap (~35% haste) -> Mastery -> Mastery ---------------------------> Mastery -> Crit. Seriously, mastery is worth more than crit at every feasible level of mastery/haste.
-More haste makes mastery worth more until cap; haste double dips in this manner and exponentially improves damage.
-Additional mastery increases damage at a linear rate. Simply 1% increase in mastery directly translates to 1%*TEB/4 in damage due to the monk generating the same number of TEB stacks. Each stack just does more damage.
-Crit is crit. It makes all your damage hit harder, but does not interact with any other secondary stat. There is a crit floor, but we can't give up enough crit rating or agi to get to it. Just a nice to have stat now, but much much better than the old mastery any day.
-Ascension still better than power strikes until you are past BOK cap (45% haste)

Just wondering how to treat jab going forward. It's just barely a significant source of damage, but enough to change your rotation around it? Is haste capping even a worry since Jab might as well be a "Push me to get Chi" button instead of an attack. Also, those pesky CB:TP procs screwing up our TEB generation. Just a thought. Maybe a job for simcraft.

I'll work in the T15 bonuses soon, they are pretty easy to math in. Looks like with the 2pc and 4pc, you can start stacking mastery much earlier (~29% haste) as the two secondary stats become equivalent in value. Neato.

Massive overhaul of the windwalker action list, and the sim now reflects the current state of the CB:TP bug on the ptr. I'm running stat weights with the new updated action list to see whats changed from simc's PoV, but what you came up with sounds about right for what I was seeing before the changes I've made. I'm interested in seeing how the updated list will effect the values of haste primarily. The new action list can be found over at:
Diff of Monk_Windwalker_1h_T14H.simc r15385 - simulationcraft - World of Warcraft DPS Simulator - Google Project Hosting

Basic ptr rotation:
-TP on CB:TP procs (not reflected in the live list because TEB isn't as valuable on live as on the ptr.)
-BoK on CB:BoK procs if you're about to cap energy
-BoK if you are less than 2 chi from cap and about to cap energy
-RSK
-FoF if EB isn't up, energy won't cap during channel (currently represented by a hardcoded 4 but subject to change during blood lust), and TP buff won't drop
-Chi wave if you won't cap energy in the global (not reflected in live action list because of obvious reasons...)
-BoK on CB:BoK procs
-Jab, <=2 chi without ascension, <=3 chi with ascension
-BoK if you will have enough energy to jab for the next RSK, or chi is at cap (just thought about this and probably will end up changing it to >=4 for ascension, as you'd have enough to RSK afterwards either way, just can't test right now because stat weights are running)

Feel free to scratch your heads and cry at how this stuff wasn't implemented a long time ago :/.

urotas · 01/15/13, 4:43 AM

If the bug still exists when 5.2 goes live, we might also consider running "/cancelaura Combo Breaker: Tiger Palm" macros on Blackout Kick, RSK and FoF. You can do that right now, but it may not be worth it depending on your gear. The loss of 1/4th of a TEB charge now is worth much less than 1/3rd of a charge will be with the new mastery.

The largest loss comes from using FoF with a CB: TP active, as you lose half of a TEB charge now, or 2/3rds of a charge on the PTR.

gahddo · 01/15/13, 4:45 AM

Originally Posted by urotas

If the bug still exists when 5.2 goes live, we might also consider running "/cancelaura Combo Breaker: Tiger Palm" macros on Blackout Kick, RSK and FoF. You can do that right now, but it may not be worth it depending on your gear. The loss of 1/4th of a TEB charge now is worth much less than 1/3rd of a charge will be with the new mastery.

The largest loss comes from using FoF with a CB: TP active, as you lose half of a TEB charge now, or 2/3rds of a charge on the PTR.

The way the sim plays out you don't have a very high uptime on CB:TP to begin with, I highly doubt cancel aura will be required but might skew weights towards higher haste levels which could increase dps. I just have no idea how I'd go about doing that in the sim.

gahddo · 01/15/13, 5:09 AM

After further testing I've found that the sim prefers haste up to around 9500 (guestimate for between 9000-10000) at which point the wait time drops below 2 seconds and further haste only goes into auto attack damage. After that mastery > crit >>>>> haste is the clear outcome. Agility is worth more than twice haste/mastery at all levels, no yellow gems.

-edit- it appears it may be a gain to gem haste to achieve the breakpoint, as it isn't reachable in current gear without doing so.

opinel · 01/15/13, 6:26 AM

Originally Posted by gahddo

After further testing I've found that the sim prefers haste up to around 9500 (guestimate for between 9000-10000) at which point the wait time drops below 2 seconds and further haste only goes into auto attack damage. After that mastery > crit >>>>> haste is the clear outcome. Agility is worth more than twice haste/mastery at all levels, no yellow gems.

-edit- it appears it may be a gain to gem haste to achieve the breakpoint, as it isn't reachable in current gear without doing so.

And this is with Ascension ?
How is it possible not to cap on energy with such huge amounts of haste ?
I'm already almost never energy-locked with as little as 4K haste and Ascension, I can't imagine how I'd feel with 9-10K. And this is without even taking the 2pc into account.

Eyedore · 01/15/13, 7:41 AM

Right now with 1500 mastery on gear you have a 21.7% combobreaker chance (raid buffed), that will be a fixed 12% in patch 5.2.
Our T14 2p lets us do more FoF which can lead to less energy usage.
Our T14 4p gives 50 energy per minute if EB is used on cd.

On the other hand T15 2p gives 15% chance on 15 energy on a Jab/EH

All in all we will be capping energy less fast and thus we can use more haste

Venyasure · 01/15/13, 9:15 AM

Originally Posted by Saltycracker

-Additional mastery increases damage at a linear rate. Simply 1% increase in mastery directly translates to 1%*TEB/4 in damage due to the monk generating the same number of TEB stacks. Each stack just does more damage.

I didn't delve very deep into your spreadsheet, but i think this is a flaw in the crit/mastery comparison. You don't use homogeneous formulas for crit scaling and mastery scaling. You write that mastery increases damage in a linear fashion, but you don't use the same "type of increase" for crit. What you're doing is :

Mastery scaling = Damage(1% mastery) / Damage (zero mastery)
Crit scaling = Damage(current_crit + 1%) / Damage(current_crit)

You can of course say that mastery increases damage linearly, but then you should do the same for crit and make 1% crit being a 1% damage increase, instead of a 0.76-ish if you have a "base" 32% crit. Thus you should either calculate :

Mastery scaling = Damage(1% mastery) / Damage (zero mastery)
Crit scaling = Damage(1% crit) / Damage(zero crit)
Modified spreadsheet : http://www.filedropper.com/rotundwin...difiedabsolute

OR

Mastery scaling = Damage(current_mastery + 1%) / Damage(current_mastery)
Crit scaling = Damage(current_crit + 1%) / Damage(current_crit)
Modified spreadsheet : http://www.filedropper.com/rotundwin...difiedrelative

(the Excel files were uploaded from my French Excel document, i hope the formulas will be translated)

The second option (which is probably the more relevant) is mainly about replacing your formula in V3 :
vlookup(1+round(V27,1)/100,A:Y,16)/4*.01
by :
(1+((V25+1)/100)*(vlookup((1+(round(V27;1)/100));A:Y;16)/4))/(1+((V25)/100)*(vlookup((1+(round(V27;1)/100));A:Y;16)/4))-1

Both ways of calculating scaling will give crit>mastery in your spreadsheet, unless you have a lot of haste (about 25%+). What you did in the original spreadsheet was a handicap for crit, because you said :

- Going from 30% crit to 31% crit is an increase of 1.31/1.3 - 1 = 0.77%
- Having a 4% overall damage increase from 1% mastery is an increase of 1.04/1 - 1 = 4%

Instead of either :

- Having a 1% overall damage increase from 1% crit is an increase of 1.01/1 - 1 = 1%
- Having a 4% overall damage increase from 1% mastery is an increase of 1.04/1 - 1 = 4%

OR

- Going from 30% crit to 31% crit is an increase of 1.31/1.3 - 1 = 0.77%
- Going from a 16% damage increase with 4% mastery to 20% damage increase with 5% mastery is an increase of 1.2/1.16 - 1 = 3.4%

Am i correct, or do i misunderstand your spreadsheet ?

Saltycracker · 01/15/13, 10:56 AM

Thanks for looking at it Venyasure. Maybe it's a probelm with how I am initially calculating crit damage:

Damage*2*crit% + Damage*1*(1-crit%) = Avg Damage (Davg) (Simplified example)

Davg = 2Dc +D(1-c)
Davg = 2Dc + D -Dc
Davg = Dc +D

Davg = D(c+1)

So if you add 1% crit to the formula and calculate Davg as:

Davg(c+.01)/Davg(c)

IncDavg% = D(c + .01 + 1) / D(c +1)

IncDavg% = (c + 1.01)/(c + 1)

If I use that formula, or the one I have currently I get the same number. It's still early, let me think about your post a little bit more. I know in your analysis you include glancing blows and such, but I'm just trying to keep the scope small and not let the math get too hairy.

Also, I don't believe that damage output scales linearly with crit chance. It scales logarithmically i.e. it's worth more at lower levels than at higher levels.

Re: Mastery calculations

To keep math simple, I'm assuming a player blankets their rotation with TEB, meaning they are using it on CD right when the previous buff duration expires. To do this you need to be using it 4 times per minute.

(TEB/min)/(4CD/min) = TEB/CD

Each stack of TEB provides mastery% of damage increase. So the total damage increase will depends on your TEB/CD ( which is a function of your chi spent i.e. haste) and your mastery %

Davg(h,m) = D(1+m%*TPCD)

IncDavg = D(1+(m+1)%*TPCD)/D(1+m%*TPCD)

IncDavg = (1+(m+1)%*TPCD) / (1+m%*TPCD)

That's where I made my mistake. I'll adjust my spreadsheet. I wasn't undervaluing crit, I was overvaluing mastery. They look to be about even in value at my nominal level of haste.

immikey · 01/15/13, 12:13 PM

Well on PTR 5.2 you get .0016667% crit per point of crit, whereas you get .00033493% mastery per point of mastery. Just using some basis of napkin math, it looks like even if 1% of mastery is worth 4x as much as 1% of crit you would still receive more total damage from stacking crit after comfortable levels of haste.

For mastery to be a better stat than crit it must bring about 4.976 times as much damage than crit by percentage.