ShadowCraft for Mists of Pandaria - Page 10

Seliathan · 03/07/13, 7:55 PM

I think he already avereraged it out though, and so far we've also received about 2 to 3 TF items each boss, with the obvious exceptions of having zero TF on one boss, and a whopping 5 on another one.

It should therefore be possible to get fully thunderforged gear by the end of this raidtier, yet I don't think that we should use it to sim EP values either. Six itemlevel are quite alot when it comes to calculating EP as accurately as possible, and while these six itemlevel won't change the exact numbers much, they are for all intents and purposes "wrong" numbers for the majority of people using them. Additionally, those of us who are able to get their hands on a fully thunderforged gearset most likely don't use generalized EP values anyway.

As far as creating BiS lists goes, it really doesn't matter much. While the additional secondary stats from the thunderforged gear influence EP values, there won't be any difference between heroic-only and thunderforged-heroic lists concerning the base item type. I'd rather be interested in how important a perfectly rolled random enchantment bracer is going to be, but seeing how we are most likely going to overcap on both expertise and hit, I don't think they will have any influence on the rest of the gear slots.

Nalinaa · 03/07/13, 9:17 PM

Originally Posted by Schmoopy

RNG can be in and against your favor and a sample size of 1 isn't very sound for statistical analysis.

What I said was never meant to be fact, just what I observed, and what I observed was that thunderforged were appearing often enough that it was extremely plausible we could be in full thunderforged heroic by the end of the tier. But again like you said RNG could be not in your favor meaning you will never get 1 thunderforged heroic piece at all.

Kryptomaniac · 03/08/13, 12:08 AM

Originally Posted by Seliathan

I think he already avereraged it out though, and so far we've also received about 2 to 3 TF items each boss, with the obvious exceptions of having zero TF on one boss, and a whopping 5 on another one.

It should therefore be possible to get fully thunderforged gear by the end of this raidtier, yet I don't think that we should use it to sim EP values either. Six itemlevel are quite alot when it comes to calculating EP as accurately as possible, and while these six itemlevel won't change the exact numbers much, they are for all intents and purposes "wrong" numbers for the majority of people using them. Additionally, those of us who are able to get their hands on a fully thunderforged gearset most likely don't use generalized EP values anyway.

As far as creating BiS lists goes, it really doesn't matter much. While the additional secondary stats from the thunderforged gear influence EP values, there won't be any difference between heroic-only and thunderforged-heroic lists concerning the base item type. I'd rather be interested in how important a perfectly rolled random enchantment bracer is going to be, but seeing how we are most likely going to overcap on both expertise and hit, I don't think they will have any influence on the rest of the gear slots.

Speaking of bracers, I just got the regular thunderforged bracers of midnight comet, and the random enhancements seem kind of odd. Either I missed something, but I thought there were only certain combos you can get such as "of the zephyr".

I just got one with double rolled expertise, so it had 907 expertise as the only secondary stat.

When ever I try to reload shadowcraft from my armory with it equipped, the page breaks.

snowman2050 · 03/08/13, 3:40 AM

I have to agree with Pathal and state that it might be best to not include the TF pieces when calculating BiS / EP. We as a guild have killed 7 bosses and only seen 2 TF pieces drop, RNG can be a killer for some and a winner for others. It would also go without saying that a HC TF piece would replace its non TF equivelant.

Usunar · 03/08/13, 3:56 AM

I've been thinking about one other thing regarding all this thunderforged stuff, can it theoretically be better to aim for all heroic tf offset pieces since there is no tf tier items?I know its a very far with all of rng, and on the other hand we still dont have an accurate ep for T15 bonuses.

Hamsda · 03/08/13, 6:59 AM

I doubt that blizzard would keep a setbonus which is so weak that having 6 ilvl on 4 pieces of your gear more would be a dps increase.

The double rolled expertise bracers seem really weird to say the least. If they don't have a different name that could cause some trouble for simc, shadowcraft and the like.

Seliathan · 03/08/13, 8:09 AM

It really all depends on how strong the T15 bonuses are going to turn out, or at least how accurate the calculation is. In my opinion especially the 4piece bonus is in jeopardy if it turns out to be around 1-2k EP, since there are gloves, chest and legs available with an additional socket, and most of them provide better secondary stats aswell.

I am currently calculating raw stat differences for combat which I am going to post in the combat thread itself for further discussion, to keep this one on topic.

Regarding the bracer issue, we also had randomstat mail bracers drop that only had doublerolled mastery (858) and the additional socket. Maybe these are intended to only have a higher amount of a single secondary stat. It would make sense, since that way getting the "best" bracer possible isn't nearly as random as it was back in T11 with over double the possible roll-combinations.

Kryptomaniac · 03/08/13, 9:33 AM

Originally Posted by Seliathan

It really all depends on how strong the T15 bonuses are going to turn out, or at least how accurate the calculation is. In my opinion especially the 4piece bonus is in jeopardy if it turns out to be around 1-2k EP, since there are gloves, chest and legs available with an additional socket, and most of them provide better secondary stats aswell.

I am currently calculating raw stat differences for combat which I am going to post in the combat thread itself for further discussion, to keep this one on topic.

Regarding the bracer issue, we also had randomstat mail bracers drop that only had doublerolled mastery (858) and the additional socket. Maybe these are intended to only have a higher amount of a single secondary stat. It would make sense, since that way getting the "best" bracer possible isn't nearly as random as it was back in T11 with over double the possible roll-combinations.

There are different versions of the bracers, some with singular stats, some with double stats. None of them have item suffixes though. There was a random rogue whos name doesn't come to the top of my head who had mastery crit.

Its really weird how they did it.

Seliathan · 03/08/13, 9:42 AM

Originally Posted by Kryptomaniac

There are different versions of the bracers, some with singular stats, some with double stats. None of them have item suffixes though. There was a random rogue whos name doesn't come to the top of my head who had mastery crit.

Its really weird how they did it.

That would increase the amount of possible stat combinations even further, and beyond what we already saw in T11. Why they would do such a thing is beyond me, I guess we will have to wait for more drops to occur or some bluepost to highlight the randomstat bracers.

Regarding combat modelling: While trying to get some proper combat EP values out of shadowcraft, something really started baffling me. Whatever way I change certain items/gearslots around, having presumably BiS gear in all slots with the exception of ignoring set-items completely, mastery is sitting at a whopping 1.7 EP, while haste is down at 1.3. I have yet to find the point where mastery suddenly takes over by such a huge margin, since my current gear puts mastery at 1.15 and haste at 1.27.

Maybe someone can share some insight regarding this issue, and while modelling combat accurately isn't a priority considering the recent changes to assassination, it is hard to compare the two specs using shadowcraft right now when something obviously is interfering with modelling it correctly at some point - unless these values are in fact correct, which for the life of me I cannot imagine to be true.

Edit: So far it seems to be a problem with the massive amounts of haste one can get in the current tier by completely ignoring the set items and socketing for it. Mastery is bound to take over at some point (changing gear bit by bit it happened at about 1.36 EP each, with ~14k haste and 6.6k mastery). Setting the gem priority to agility delays it but having complete (presumed) BiS Gear is bound to give the edge to mastery. That is without the new trinkets, since I am not sure yet whether the procs are modelled correctly.

Edit2: After changing the gems around completely, with Blue -> Glinting, Red -> Delicate and Yellow -> Deft, haste sits at the top again by a considerable margin. Looks like we're going to have to balance things out via gems to get the absolute theoretical maximum out of our gear.

Pathal · 03/08/13, 9:53 AM

We can now calculate the GCD cap in combat calculations, and therefore wasted energy from haste. I'm sure it's going to throw a lot of people off when their haste diminishes some, but it makes sense once you think about it, I promise.

If you think it's dropping significantly faster than it should be however, some stat totals (and the resulting EP weights) would greatly help.

EDIT: I just discovered and fixed a combat bug where MG wasn't being subjected to energy capping. I don't think it will have any massive impact on EP weights or gearing, but should it discourage mastery from overtaking haste quite so easily. I don't know when it will make it to the frontend, and I'll continue to monitor the situation.

Seliathan · 03/08/13, 1:39 PM

Thanks for the update Pathal, I edited my post above with some numbers and changes I made to counteract the haste/mastery inflation.

Another thing I just noticed: Seeing the numbers simcraft produced I tried to get the same result with the shadowcraft engine. I expected some deviation in the 5% margin, since as I understand shadowcraft does not run the same amount of iterations. Using the same gear as above but switching out the legs/gloves/head/spaulders for the T15 provided a +22k DPS jump when using the 4piece bonus.

Looking at the values simcraft provides with an average of 63 APM, 30% idletime and only 3% energy overcapping, it appears that instead of the 0.2sec gcd reduction SB provides while the buff is up, it is calculated for the whole duration of the fight, which in turn explains the 22k DPS the 4 piece bonus is currently modelled at.

Nalinaa · 03/08/13, 2:29 PM

Are we sure that Vicious Talisman of Shado-pan Assault is being modeled correctly? It has 600 more EP value than heroic Rune of Re-Origination in assassination spec and 375 more EP than heroic Talisman of bloodlust for combat.

Is the ICD set to 105?

Pathal · 03/08/13, 2:52 PM

https://github.com/dazer/ShadowCraft...c_data.py#L480

'vicious_talisman_of_the_shado-pan_assault': {
        'icd': 105,
        'proc_chance': .15,
        'trigger': 'all_attacks'
    },

It's statistically more likely that it's an issue with the RPPM trinket, not the normal one.

Kryptomaniac · 03/10/13, 5:02 AM

So I was just messing around in shadowcraft, looking at higher gear levels. And as I understand it with RPPM Trinkets, haste becomes a higher valued secondary stat, but it is a high enough secondary stat that mastery becomes almost obsolete?

Im getting that the EP values for Assassination switch to 1.538 Haste, 1.439 Crit, and 1.101 Mastery. This also brings to mind that if indeed Rune could possible be BIS, current setup (not including thunderforged items except Ra-den) would give me more haste then mastery, by about a whopping 2496 rating difference. Wouldn't that cause issues with RORO.

Just looking for some clarification.

Edit: just for sake of thoroughness.
Puts me at:
Critical Strike Rating: 6600
Haste: 9547
Expertise: 1970
Hit: 2861
Mastery: 7051

Pathal · 03/10/13, 9:37 AM

Are you doing that with RoRO?

Rune biases even distribution of stats. It's going to encourage EP weights to shift dramatically.