BrM with a Side of AskMrRobot

[SurrealNight]

Also, dual-wield, for both WW and BrM it suggests Str weapons for a good chunk of the BiS weapons. Even if I go in and change the weight of Str to 0. WTF!?

Yeah I don't think you would want to do that, giving up almost ~500 Agil wouldn't be worth it since BrM's aren't tied to weapon damage in any way for mitigation.

Other tanks might be able to get by with an Agil weapon if it's a big upgrade but it doesn't work out for us.

[Zoopercat]

TechGnosis - the only STR weapons I see are the ones with the epic gem slot (509), and those are listed #4 and lower in the list. But the math works out that they are better than some of the 502 weapons, even with the STR. Is that also what you are seeing? The BiS that I see are the Claws of Shek'zeer.

Also, Mr. Robot does allow for optimization up to a certain point, then it goes to the next stat. You can do this for cappable stats that matter, like Stamina for tanks, Haste breakpoints, Hit, etc.

[SurrealNight]

TechGnosis - the only STR weapons I see are the ones with the epic gem slot (509), and those are listed #4 and lower in the list. But the math works out that they are better than some of the 502 weapons, even with the STR. Is that also what you are seeing?

http://www.askmrrobot.com/wow/gear/usa/stormrage/rikimu

Elegion and Scimitar of Seven Stars are showing in my list, neither have a epic gem slot (Scimitar is showing with a +320 hit gem). Scimitar has Mastery too so it's secondary stats aren't even well itemized. 496 Claw's are showing as BiS but two of the weapons before that don't make much sense to me.

[Ragethorn]

Also, Mr. Robot does allow for optimization up to a certain point, then it goes to the next stat. You can do this for cappable stats that matter, like Stamina for tanks, Haste breakpoints, Hit, etc.

How, exactly, do you set it up for Haste breakpoints? I must be missing it somewhere.

[The Madgod]

TechGnosis - the only STR weapons I see are the ones with the epic gem slot (509), and those are listed #4 and lower in the list. But the math works out that they are better than some of the 502 weapons, even with the STR. Is that also what you are seeing? The BiS that I see are the Claws of Shek'zeer.

Also, Mr. Robot does allow for optimization up to a certain point, then it goes to the next stat. You can do this for cappable stats that matter, like Stamina for tanks, Haste breakpoints, Hit, etc.

I think that's very misleading. The only reason that those str weapons are better is because you can put your agility sha-touched gem into them.

As well, I really do want to stress that stamina shouldn't natively be so high in your stat weights... It's extremely confusing to a lot of people when it REALLY isn't that important. Stamina is only good to a certain point, and I've yet to have health trouble when barely gemming and enchanting for it. Hell basically every brewmaster here hasn't really found it necessary... one did, but it was contested by another who was doing the exact same encounter, so really it was more about the raid makeup.

[Zoopercat]

It all comes down to math. Here's how they stack up. As you can see, the Claws blow all other weapons out of the water. But the math is what determines the ranking:



EQUIPPED: Gara'Kal (489) (Ranked #8 on Mr. Robot)
491 Agi * 1 = 491
737 stam * 1.34 = 987.58
347 Haste * 0.8 = 277.6
295 Mastery * 0.4 = 118
Reforge 118 Mastery to 118 hit = (118 hit * .85) - (118 mastery * 0.4) = 100.3 - 47.2 = 53.1
TOTAL: 1927.28

Elegion (502) (Ranked #11 on AMR):
554 Str * 0 = 0
832 Stam * 1.34 = 1114.88
386 Hit * .85 = 328.1
342 Exp * .85 = 290.70
Total: 1733.68 (worse than the weapon equipped, which is why Mr. Robot puts it in slot #11)

Scimitar of 7 stars (509) (ranked #6 in AMR):
512 Str * 0 = 0
887 Stam * 1.34 = 1188.58
360 Hit * 0.85 = 306
345 Mastery * .4 = 138
Reforge 138 Mastery into Expertise = (138 Expertise * 0.85) - (138 mastery * 0.4) = 117.3 - 55.2 = 62.1
240 Stam gem * 1.34 = 321.6
60 Mastery socket * 0.4 = 24
Total: 2040.28 (better than the equipped weapon, even with STR)


Claws (509) (Rank #1)
592 Agi * 1 = 592
887 Stam * 1.34 = 1188.58
309 Exp * 0.85 = 262.65
444 Crit * .7 = 310.8
550 Agi gem * 1 = 550
Total: 2924.03 (blows all other weapons out of the water)

Regarding the gem option - hey, it matters, it makes a difference!! Also, once we put the stam cap back in (sorry we bugged that out), the weapons might rank different. Mr. Robot will correctly score stats a zero when they go over the cap.

[SurrealNight]

It all comes down to math. Here's how they stack up. As you can see, the Claws blow all other weapons out of the water. But the math is what determines the ranking:

Derp I forgot I have my AMR set to WW spec but I have heroic loot turned off as well =P I understand why it's suggesting Kilrak just below 496 Claws but the odds of Kilrak dropping and no STR class needing it + me being willing to put such an expensive gem in it are next to none anyways.

As to what Madgod said yeah the Stam thing is pretty crazy it pushed me over to the WW spec to get what I was after. Also, as mentioned I don't see any UI options for capping Stam or Haste?

[Zoopercat]

Surreal - there should be an option to cap stam, but we bugged it in a recent update. We'll have that fixed soon, sorry about that.

Regarding the Kilrak, and a STR class not needing it - totally understand. You are right about that for sure. But, we don't like to filter out gear based on that, because if we take it out, other people have arguments to see it.

[Ragethorn]

What about the haste breakpoints you mentioned earlier?

And yes, I saw you address this in the previous post. I wanted to bring your attention back to it because you really should edit that claim out.

[Yellowfive]

Hey -- I do most of the work currently on the actual optimizer code for Mr. Robot.

1. We'll add the stamina cap for brewmasters, it should have been there but wasn't turned on for some reason.

2. As for a haste cap option for brewmasters... can someone point me to the best/most specific discussion of the topic, a really solid argument that you will see better performance when achieving specific thresholds of haste? Unless I'm REALLY convinced that it's a "thing" and that it matters, I am loathe to add it. The tanking optimization is already the most processor-intensive due to all the DR calculations, so adding more caps has a significant impact. That said, if it really is a thing, we will certainly add it.

3. The whole stamina discussion... can someone point me to some good arguments that gearing for stamina is a waste on a brewmaster -- something that does not rely on anecdotes? Our decision to increase the weight on stamina for all tanks was not arbitrary. There is a very good discussion on sacredduty.net that is not specific to any particular tank class and describes a philosophy similar to ours. That blog post links to an older article from cataclysm, but that is still relevant today: what do i do after reaching the block cap Though the "block cap" is no longer a relevant topic, the discussion of stamina in that blog post is still very relevant.

To summarize all of the stamina talk extremely briefly: if you are doing a boss fight where increasing your stamina doesn't matter... then the boss is not stressing your survivability. Whether you gear for stamina or avoidance or whatever will not matter. You could gem for spirit and do just fine. But if you ever find yourself in a situation where the boss is actually hard... you are probably going to want that extra stamina. The 2nd article linked has a good discussion of how stamina can help your healers conserve mana, and how you don't need to cross very specific "thresholds" of total health for stamina to be good.

4. Strength weapons in the list for brewmasters: yeah, there are some strength weapons in the list The ranking is purely mathematical. If you don't want to use a strength weapon... don't use one!

[Ragethorn]

2. As for a haste cap option for brewmasters... can someone point me to the best/most specific discussion of the topic, a really solid argument that you will see better performance when achieving specific thresholds of haste? Unless I'm REALLY convinced that it's a "thing" and that it matters, I am loathe to add it. The tanking optimization is already the most processor-intensive due to all the DR calculations, so adding more caps has a significant impact. That said, if it really is a thing, we will certainly add it.

and this is why I will continue using wowreforge.com

[SurrealNight]

Running out of time for an in depth post or to read the article so will have to wait until tonight/tomorrow for that. For now I can start off by stating that the main reason stamina is so ignored by most BrM's is that Agil/Haste/Crit/Mastery basically have a near infinite ability to improve our CTC/mitigation so it's difficult to justify pulling resources away from them if you aren't getting killed due to to small of a health pull.

Haste - Energy to fuel our Chi generation that fuels Shuffle/Guard with excess able to go into L30 healing talents. "Haste Caps" are relative to creating enough Chi to utilize these abilities with near 100% uptime. After that point extra Chi's value has questionable value. Haste also improves Gift of the Ox sphere generation and Elusive Brew uptime (though slower than crit)

Crit - More Elusive Brew stacks means more uptime of 20-25% dodge bonus. Stacking Crit to EB being at anywhere near 100% uptime is impossible but something you "try" for.

Mastery - Increases the amount of damage we stagger away per hit so is effectively our version of Block. Synergies with Haste in that you need enough Chi to clear the dot this creates. This stat has no cap or DR or limit technically until you reach 100% which is impossible.

So when you look at all of the above and realize that you have an infinite pool to dump rating into that doesn't have diminishing returns it's hard to justify stacking stamina unless the fight or your gear level vs content requires it. Now factor in that stacking haste/crit is a DPS boost which means shorter fight, less chance of enrage timers, ect and it becomes even less attractive.

[The Madgod]

I don't understand Ragethorn's post -- you quoted me saying "I'll add a haste cap if it's a thing, otherwise not, and I have a very valid reason for that (performance)." OK... moving on to SurrealNight's actually helpful reply, which was more what I was looking for:

So Haste and Mastery seem like the interesting stats to discuss from a survival standpoint here. In general, going for avoidance via crit via elusive brew isn't considered that big a deal for surviving burst damage due to its highly random nature, so we'll just put that aside for now. It is certainly valuable, but I would consider it less valuable compared to haste, mastery, and stamina. Do you all think that's a valid statement? Or does crit have some other synergy with chi generation or some other mechanic?

Actually in certain applications Elusive Brew is one of the best ways to mitigate burst (see: Sha of Fear's thrash), but that's mostly a nitpick. For the most part, you are correct in saying that haste and mastery are the big "EH" stats, though crit is considered better than mastery for most encounters due to its benefits to multiple aspects of our class' ability to handle damage, and not just one.

Mastery is interesting for brewmasters... I don't have direct experience with a brewmaster, but unless I am misunderstanding how it works, it seems that a brewmaster always staggers incoming boss melee swings... so it's kind of like having a 100% chance to block, and mastery makes it bigger. Clearing the DoT is all good... but I think that spreading out a significant chunk of big hits over 10 seconds is valuable in and of itself. So I'll look into this, but it seems that mastery could be a good candidate for a stat to stack before stamina for a brewmaster, in terms of surviving bursts.

In terms of surviving burst, haste is still superior due to Blackout Kick costing resources. Mastery won't be nice to stack for a while, until we can get to a point where we're very comfortable with our energy and chi production, since mastery actually boosts the usefulness of haste by requiring more PB's.

Haste is a little trickier... I unfortunately don't know enough about the chi generation mechanics to evaluate this one off the top of my head. I'm sure that someone here has figured out the amount of haste that you need to get various things though -- 100% up-time on shuffle would be an important value for reducing spikes in damage with 100% reliability. One could perhaps argue that reaching this threshold of haste is more important for burst damage survival than getting more stamina, but it's pretty hard to run the numbers on that.

So some specific questions:
1. What is the amount of haste that you would need to get 100% up-time on shuffle?
2. Once you have 100% up-time on shuffle, how much does the value of haste really drop? There are several other chi-consuming abilities that you could use more frequently...
3. Where does hit/exp fall into the scheme of things? If you miss a chi-generator, I am assuming that it does not generate chi? I'm sure that the math is lying around here, but if we're talking about chi generation to get shuffle up-time, we should probably compare haste to hit/exp, since they'll serve a similar purpose.

1) Two minutes of combat necessitates 48 chi be used to get a max uptime for shuffle and guard. You can have 15 keg smashes during that time giving you 30 chi for a total of 600 energy required for that much, and then the rest of the 18 chi would cost 720, for a grand total of 1320 energy necessary at minimum (without Power Strikes. PS provides 6 "free" chi every 2 minutes assuming perfect play, so with PS it'd be about, at minimum, 1120, if my math is right)... so without PS the minimum energy regen you need is 11 er/s, and with PS it's 9.33 er's... however really what creates the haste "cap" is how many purifying brews you need during that period of time.

2) haste won't drop in value until you get to a point where you don't really need any more PB's, which is fairly relative.

3) hit and expertise help smooth out chi generation and keep it less wonky. However if we're talking pure math, techncially speaking you will get higher chi generation if you go into haste instead. Most people tend to try and get to at least 7.5% hit and expertise. Whether they go past 7.5% expertise is usually due to personal preference or a desire to deal more damage.

[Yellowfive]

So... my post just double-posted, then it got completely eaten when I tried to fix it =/ So here's try 3:

@Ragethorn Your post is useless and makes no sense, I stated that if there's a good reason for a haste cap we'll add it, but would prefer not to for performance reasons if it really isn't that big a deal. If it is a big deal, we'll add one. Moving on to useful posts....

SurrealNight and Madgod bring up a few interesting points.

I think the most interesting thing is that Stagger is a "guaranteed" mechanic. It smooths your incoming damage (but relies on chi to reduce it). That said, simply smoothing your damage has value in itself. This could put mastery as a candidate for higher priority than stamina (though of course it only works on physical... but how many bosses are there that don't do significant physical damage? Most of them do).

Haste is also interesting due to Shuffle up-time, which can become a permanent increase to a permanent damage-smoothing mechanic.

Crit is cool... but I would argue that it has a lower value than Haste, Mastery, and Stamina (from a burst damage survival perspective) because it only helps increase up-time on a dodge buff. Now it might be worth more than dodge or parry, or perhaps even hit or expertise... that I can't really say without more research.


So that said... I would like some more feedback if you all are willing to give it. Here's some specific questions:

1. What is the specific value of haste that is needed to reach 100% up-time on Shuffle? And related: if I increase my haste by say... 5%, how much of an increase in shuffle up-time am I looking at?

2. After you reach 100% up-time on shuffle... how much does haste's value really diminish? Or in other words, there are other chi-consuming abilities that you will be able to use more frequently. Does the increased frequency of those have considerable value, or not? And if it's not clear... yes, I'm looking for an excuse not to put in a haste cap But if it ends up being necessary because haste drops off significantly, then so be it! We'll write some code to handle it.

3. Where does hit/exp fall in this equation? I would assume it has a significant impact on chi generation as well.


And a few specific comments:

Brewmasters are significantly different from other tanks because our style of mitigation demands a focus on efficient resource generation and consumption. Being able to bump up your shuffle uptime will be much better for you than essentially proping yourself up with gear.

I don't think brewmasters are special snowflakes in this regard. All tanking classes rely heavily on generating and consuming a resource to mitigate a large % of incoming damage. That is why hit/exp/haste are so popular on paladins, or hit/exp on warriors, etc. Stacking stamina isn't "propping yourself up with gear" any more than stacking haste for shuffle up-time is "propping yourself up with gear." What we're trying to figure out is what's the best way to prop yourself up with gear?


As for the weapon rankings... I think that the current ranking is not misleading or confusing -- it is mathematically sound, and it illustrates that agility weapons are superior. It takes a heroic strength weapon to be about on par with an LFR agility weapon. I think that makes the case pretty clear that you should be using agility weapons.

[The Madgod]

SurrealNight and Madgod bring up a few interesting points.

I think the most interesting thing is that Stagger is a "guaranteed" mechanic. It smooths your incoming damage (but relies on chi to reduce it). That said, simply smoothing your damage has value in itself. This could put mastery as a candidate for higher priority than stamina (though of course it only works on physical... but how many bosses are there that don't do significant physical damage? Most of them do).

Haste is also interesting due to Shuffle up-time, which can become a permanent increase to a permanent damage-smoothing mechanic.

Mastery is still not something that you'd want to stack. EH is very important to tanks, but it isn't the end-all of them. Crit is better because it does more with less. You only need 600 crit rating to get 1% more crit. You need 1200 mastery to get 1% more shuffle. Because it's such an expensive stat, it's really only nice to stack when you're undergeared for an encounter, and even then, you need to stack haste along with it in order to be able to purify the extra damage.

Crit is cool... but I would argue that it has a lower value than Haste, Mastery, and Stamina (from a burst damage survival perspective) because it only helps increase up-time on a dodge buff. Now it might be worth more than dodge or parry, or perhaps even hit or expertise... that I can't really say without more research.

You also forget the benefit it has for the raid and for your self healing.

So that said... I would like some more feedback if you all are willing to give it. Here's some specific questions:

1. What is the specific value of haste that is needed to reach 100% up-time on Shuffle? And related: if I increase my haste by say... 5%, how much of an increase in shuffle up-time am I looking at?

2. After you reach 100% up-time on shuffle... how much does haste's value really diminish? Or in other words, there are other chi-consuming abilities that you will be able to use more frequently. Does the increased frequency of those have considerable value, or not? And if it's not clear... yes, I'm looking for an excuse not to put in a haste cap But if it ends up being necessary because haste drops off significantly, then so be it! We'll write some code to handle it.

3. Where does hit/exp fall in this equation? I would assume it has a significant impact on chi generation as well.

see my earlier post which was originally in response to these :P

I don't think brewmasters are special snowflakes in this regard. All tanking classes rely heavily on generating and consuming a resource to mitigate a large % of incoming damage. That is why hit/exp/haste are so popular on paladins, or hit/exp on warriors, etc. Stacking stamina isn't "propping yourself up with gear" any more than stacking haste for shuffle up-time is "propping yourself up with gear." What we're trying to figure out is what's the best way to prop yourself up with gear?

Not to the extent brewmasters do. The amount of native mitigation we have is vastly inferior to other tanks and thus we rely much more so on our active abilities. As well, stacking haste for more shuffle up-time is kind of a gray area in terms of "proping yourself up." Technically speaking with native haste you have enough to in theory keep both shuffle and guard at max uptime, but you still need to PB, so you can argue that haste allows for more PB's while keeping your shuffle and guard up at the same time, as well as argue that haste allows for a higher shuffle uptime while still allowing you to purify damage at acceptable intervals. However at acceptable levels of haste, really, your getting better at your class is what's going to make your shuffle uptime become better, not more haste.

As for the weapon rankings... I think that the current ranking is not misleading or confusing -- it is mathematically sound, and it illustrates that agility weapons are superior. It takes a heroic strength weapon to be about on par with an LFR agility weapon. I think that makes the case pretty clear that you should be using agility weapons.

I still stand by what I said. Just showing the outcome of the math is, in my opinion, not sufficient for a good source that gives people a better understanding of their class, not just what they need to do.

[Caeth]

Sorry one more....do you use stam flasks/food? I'd have to bring my own food for the agi buff then i guess.

[Zoopercat]

Caeth, we are using Stamina food and flasks, with the default weights. However, the Flask is being chosen based on stat weights, so if you change the weights, different flasks will be suggested.

The food might be hard-coded because you're given your 'best stat' from the feasts. I'll have to look up what your class is getting to double check that Stamina is right.

[The Madgod]

Sorry one more....do you use stam flasks/food? I'd have to bring my own food for the agi buff then i guess.

Agi flask or armor and haste / crit elixirs, I usually using the stam food, though.

he food might be hard-coded because you're given your 'best stat' from the feasts. I'll have to look up what your class is getting to double check that Stamina is right.

All tanks are given stamina from feasts.

[kaiadam]

Some tanking thoughts that you might wish to consider, AMR people - most likely counter to many suggestions here. Then again I've tanked and I've killed heroic Sha of Fear so what do I know, right?

1.) Total overall damage intake is nigh-irrelevant. In general, tanks do not die because healers run oom or that they constantly bleed hp because their damage intake is higher than chain flash healing restores. Tanks die because they took over their maximum EHP in damage over a short enough period where they were not healed - ie. spiked. Any absorbs/damage removals/overall avoidance mechanics are in general, more or less pointless regarding survival.

2.) Too many people emphasize "damage taken" as the stat they should be trying to minimize. This directly detracts from total EHP. While overall avoidance, more chi for purifies, etc. will reduce the frequency of taking spikes to the face, all of those are irrelevant when Sha dread thrashes you and connects six times in a row in the face of your 75% avoidance. Oh it's happened. Multiple times.

3.) On the other hand, tanks don't really die in current content. Not unless you're pushing the edges of progression - ever since cataclysm, the tanking/healing paradigm has changed to the point where bosses are NOT hitting you for most of your hp in one autoattack, and oftentimes significantly less. Because of this spike reduction, avoidance/reduction setups are viable.

4.) Tank damage is pretty important these days where tanks are able to compete with the dps in overall damage. Specific fights like Will of the Emperors or Sha, where tanks do the lion's share of the damage on the main boss while dps are concerned with taking care of adds, tank damage makes a huge difference in terms of overall lesser damage taken by the raid. Or you can point towards the stupid paladin tank with 1.3 mil dps on heroic windlord who outhealed every healer on the fight.

5.) I quite honestly have no idea what is behind the haste stacking movement for brewmasters. With 0 haste and chi brew talent, a brewmaster has around 27 chi/min (assuming hit/exp caps), enough for shuffle to be up the entire time, two guards, and three purifies. Mastery and stam give more overall survivability. Crit and agility gives more dps. Haste gives you more chi which you can spend on ... more purifies? more heals? /shrug.

6.) There really needs to be two styles of tanking optimization on AMR. a.) I am dying. Maximizes EHP via mastery, stam while retaining hit caps for constant and predictable chi flow. b.) I'm not dying, so I want to look good on the meters - emphasize haste here for lowest damage taken and healing done, or crit/agi for damage done. Both are valid for progression, depending on the fight. You pick choice b.) on amber shaper, for example, but a.) if you're always getting trounced by sherkzeer adds.

7.) hit/exp is basically essential and something that should be capped for any serious progression tanking. Raiding is all about bringing order to the chaos of the fight and having inherent randomness in the most important raid member's actions is just silly. For what it's worth, the stat values I go by are ordered as such: hit/exp cap > crit > mastery > haste.

[TechGnosis]

I think that's very misleading. The only reason that those str weapons are better is because you can put your agility sha-touched gem into them.

Ok, I can KIND of understand why H.Kilrak is above everything but the claws (due to the sha-touched gem, which is worth 550 agi), but why is Elegion and Scimitar of the Seven Stars so high as well? No Sha-Touched Socket, heck Elegion doesn't even HAVE a socket at all.