I believe that he is concluding, based on his calculations, and the benefits of Agi, that it is better, point for point than any other stat, but not at least twice better. Since with the way gems are in MoP, with secondary stat gems having twice as much stats as primary stat gems, gemming secondary stats will be better than gemming Agi.

Basically, Agi > all, but 320 Exp (or whatever secondary stat) > 160 Agi. Thats how the new gems are set up.

Further clarified, Agi gives:
dodge -> obvious personal benefit
AP -> benefits statue as well as Guard and self heals
Crit -> benefits statue as well as Elusive Brew

Remember that the spreadsheet's 2nd page will allow you to disable the statue's contribution if you only want to consider personal survivability and ignore the raid.

It increases dodge

It increases AP which means more heals from expel Harm, Gift of the Ox, etc

It increases AP which affects size of Guard, and the guard from your statute

it increases crit, which increases frequency EB stacks

All of these are increased at a lesser extend with agility than with secondary stats, but teh combination is of all the increases its whats important.

i will repeat what ven mentioned/.. since gems provide twice as many secondary stat point than primary.. secondary stat gem cuts, profession bonus etc, will be more valuable than agility

@venya
The difference in damage between DW and 2h should hardly be noticible at all. This is one of the things where Blizz really did a good job. If needed I will break that down, but as I am posting from my mobile it would be to complicated. As long as the weapons have the same iLVL there should be nearly no difference.

For my own calcs I always consider 50% of the healing from ZS, EH and GotO to be overheal. We are in a raid where healers might try to constantly top you off. A precasted greater heal could heal you full a splitsecond after the big blow from the boss. It all depends on your healers.

For an average tank that doen't use GotOx (or not appropriately ie: overhealing instead of during damage spikes), and letting aside the statue, i disabled both from the spreashit and it gives :

agi >> armor >= haste/parry/crit/dodge >> hit/exp/mastery >> ap

But in this configuration, as you say, Monks not using efficently GotOx require more attention from the healers than any other tanks. So Monk performance in comparison with other tanks and their stat priority heavily depends on your ability to take avantage of GotOx, which, i think, will require a great deal of 'skill' (and encounter knowledge/training).
Sounds like an interresting challenge :-)

That's fairly true with some other tanks as well. The difference in survivability of a DK that uses death strike timed around damage spikes rather than spamming it asap is pretty noticeable. Further, the balance of mastery vs other avoidance is highly tied to how well you use that mastery for DKs.

So is this going to turn into exp hard cap having a higher skill level but more optimal play vs the lazy mans approach to tanking where you say hard cap be damned and jab a missed keg smash and throw your now excess points into more haste/crit/parry/dodge.

e:nvm

Not sure if I missed it in this thread but in terms of reforging what is the top priority?

Obviously Hit then Exp SoftCap then what?

What about gemming in red/blue/yellow

I know I am asking about the basics but the spreadsheet is pretty straight foward and I am easily confused by the simpler aspects.

Last 10 posts is basically a discussion about it.

I suggest you download the spreadsheet, type in your stats, type 0/1 in appropriate boxes about your usage of certain abilities and then click "OK + Graph" button = your stat weights in the middle.

With hit and expertise soft cap it should be: expertise (all the way to the hard cap) -> haste -> crit ->parry -> dodge -> mastery

Agility is NOT 2x better then expertise or haste, so I suggest gemming for those instead of pure agi. But those weights change if you for example don't care about Ox statue or don't pick up gift of the ox heals (it would lower expertise value) etc.

Important update of the spreadsheet :

http://www.filedropper.com/brewmasterspreadsheetcopy

Bug fixes

- Colossus and River Song could proc on missed Jabs, whereas they can only proc on hits, dodges and parries
- minor missed Jab issue (really minor, it affects the results in their most remote decimals ^^)

Faster calculation

I have made many modifications concerning what is calculated and what is stored as variables. The program used to be written in a very inefficient way. Now it should be significantly better, although there is still much to do to achieve perfection on that matter ; yet I don’t think I’ll go any further, currently I’m satisfied with it. But this is just computer science talk : what matters is that the “OK” and “OK + Graph” buttons should calculate the results more rapidly than before.

New feature : AOE

There’s a new set of options on the second sheet :



It’s easy to explain how to use them (and I hope it will be easy to understand), but the results need much more discussion to be fully understood.

How to use these options

First of all, of you want to make AOE calculations, the "Boss DPS" cell on the first sheet should represent the combined DPS of the mobs attacking you. You will probably have to decrease "Boss Swing" ; but don't put something inferior to 1, because this value is used for Swift Reflexes, and they have a 1 sec icd.

The “Targets” cell is the number of targets. If you enter 1, the spreadsheet will behave as it used to be, it will be a single target fight. If you put 2, the only difference is that KS will hit both targets (which has consequences on damage and enchant/GotOx procs). And if you put 3 or more, you can begin doing some AOE (you might want to decrease the “Target level” cell ; usually you don’t aoe bosses).

But you’re not AOEing yet. You need to specify your use of Spinning Crane Kick. The “SCK use” cell represents the ratio between your cast SCK and your successful Jabs. Putting 0.7 means that you cast 7 SCK for 3 Jabs. Putting 1 means you spend all your energy on SCK (and KS of course, the rate at which you use KS is still given in the usual cell), and putting 0 means you never use SCK and you Jab instead. You can use SCK by putting something superior to 0 even if you have less than 3 targets. You’ll probably have more enchant/GotOx procs given that SCK hits your targets a lot of times, the program knows about that. But your SCK won’t generate Chi with less than 3 targets, so the uptime of Shuffle will be very low.

I’ve included this “SCK use” cell because its not always easy to use SCK at its maximum potential. You may need to rapidly generate Chi with Jab because you need to purify your Stagger quickly, or more generally you might have other things to do than channeling SCK in the upcoming couple of seconds, so you prefer using Jab. Putting 1 in SCK use and 1/8 for KS maybe represents something that is impossible to do in game. And remember that the calculation is based on the optimal Shuffle uptime, which is already very abstract, it’s not realistic to introduce even more impossibilities by simulating perfection for SCK and KS.

By the way, you must have noticed that you generate an enormous quantity of GotOx when you use SCK. Please don’t be silly ; don’t put 1 in the GotOx cell if you’re AOEing. In game, you’ll never pick all the orbs (picking half of them is already difficult), and some of those you pick will do overhealing. I recommend putting a fairly low number in the GotOx cell if you use SCK. And I have a similar comment concerning Colossus : its base internal cooldown is of 1 sec, but if you’re SCKing in a group of mobs, you might get procs faster than their combined swing timer. That’s why I have introduced an artificial additional icd on Colossus, which you can modify. By default it’s 0.2, it means that you want a total of 1.2 sec icd to make sure you don’t have overlapping procs.

“Calculate optimal SCK use”

Let’s make a list of advantages and disadvantages of SCK (I might skip a couple of them) :

Advantages :

-- damage increase
-- always awards one Chi if it hits 3 targets, whereas Jab can miss which is a pure loss of 8 energy, or even more if it misses several times
-- big increase of the number of hits, which is very beneficial for enchant/GotOx procs

Disadvantages :

-- disables autoattacks which decreases their overall damage and affects Elusive Brew.
-- a detail about Jab and Colossus procs that I don’t need to explain

Depending on your enchant, and on what you put in the GotOx and Statue cells, the increases in GotOx and Statue implied by SCK might not cope with the loss of Elusive Brew. You will then have a balance point for the SCK/Jab ratio, telling you how many Jabs you must make in an aoe fight to have an optimal rotation. That’s what the “Calculate optimal SCK use” option does. If you put 1 in that cell, then when you run the spreadsheet with the “OK” button, the program will calculate the best Shuffle uptime / SCK use combination, write the SCK use result in the corresponding cell, and base the calculations on this optimal value. But beware : this calculation might take a while, possibly a dozen seconds, Excel could look bugged during that time. That’s mostly because of this option that I had to work on the calculation rapidity.

Yet you shouldn’t be using this option very often. The balance point will exist only for very low values of GotOx and Statue. For example, with the default stat values of the spreadsheet, if you put 0.05 in GotOx and 0.1 in the Statue, you should get a balance point. But it will very soon disappear if you increase one of these two, even if you don’t increase it much. Most of the time, the optimal SCK use will simply be 100%. You really need low values for GotOx and Statue to get a balance point between 0% and 100%. But at least this cell proves that you should not hesitate to cast a SCK, the loss of Elusive Brew is far less important that the gain in GotOx, Statue and Colossus.

Breath of Fire

Quite simply, the “BoF / s” cell works exactly like the “KS / s”, “EH / s” and “GU / s” cells. Just next to it is the calculation of the time between two BoFs (I called it “delay”). The default value for BoF/s is 0, so the delay cannot be calculated and there’s an error in the “delay” cell ; mathematically, there is an infinite delay between two BoFs.

The two cells underneath indicate the maximum use of BoF enabling you to keep a 100% Shuffle uptime. If you put something superior to that value, part of the graph on the first sheet could be wrong. Say if you can generate 10 Chi over 30 sec, you can’t use Guard once, BoF twice and Shuffle three times, because you would thus spend 12 Chi. But normally, the Shuffle uptime calculated by the spreadsheet sould be correct ; only the part of the graph that deals with higher uptimes should be false and represent impossible uptimes given the use of BoF.

Unless you use BoF so often that you don’t even have enough chi for 0% uptime. Namely if you use more than all your Chi on BoF. You’ll notice that very quickly, you might get a negative number of Purifying Brew on the first sheet, and some other strange things will happen. Well, what you need to remember is that if you don’t want to take risks with the validity of the results, don’t use BoF too often. And if you don’t want to take any risk at all and be sure that all the calculation represents something that’s doable in game given your Chi generation, don’t put a frequency in the BoF cell that is superior to the maximum frequency indicated in the cell underneath.

Such a problem can also happen if you’re using SCK on less than 3 targets. You’ll have a very poor Chi generation, and the program might end up using more Chi than it could. To make long stories short : don’t use the spreadsheet to simulate something you can’t do in the game. The Bof section is here to show that using Chi on BoF (which is like nerfing Chi generation) results into a big decrease in the optimal Shuffle uptime rather than a decrease in the use of Purifying Brew ; and that’s an important result.

I'll try and add conditions on the BoF use so that you don't have to bother with the "maybe the graph is wrong" issue. I'll try to make such verifications automatic.

Remarks concerning the results

SCK recent changes

Some of you might not have noticed the last changes on SCK. It now has a base duration of 2.25 sec, its damage has been increased by 10%, and haste decreases the channeling duration, together with the time between its ticks. Hence a consequence that can be both guessed and observed on the spreadsheet :

Haste

Haste is an excellent stat for AOE situations, whether it be from a dps point of view or a damage reduction point of view. It even becomes better than hit/exp in many cases (the different cases being especially the different values you put in GotOx and Statue).

Colossus

Colossus was already our best enchant for single target fights. I don’t think I need to explain why it becomes even better in an AOE situation when you use SCK.


Ending remark

Quite a few people on forums seem to be still valuating hit/exp as our best stat. It is absolutely true that on paper it is a great stat. But remember that you need to take into account the Statue and GotOx for it to be the case. Expertise is quite poor for pure damage reduction, it is only good for overall tanking. So yes, hit/exp are good, but just make sure you know what you’re doing when you stack it.