Jump to content


Photo

[Ret 5.4 MoP] Retribution Concordance (RetCon) - Siege the Day


  • Please log in to reply
971 replies to this topic

#1 Exemplar

Exemplar

    The One-Eyed Man

  • • Guide Author
  • 1,802 posts

Posted 24 September 2012 - 03:58 PM

This post has been promoted to an article

Edited by Exemplar, 30 October 2013 - 12:39 PM.

Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

#2 Ronark

Ronark

    Don Flamenco

  • Members
  • 345 posts

Posted 24 September 2012 - 04:24 PM

I believe that SoB affects both TV and DS, as was evident in last night's H Madness run.

#3 Wrathblood

Wrathblood

    Don Flamenco

  • • Guide Author
  • 351 posts

Posted 24 September 2012 - 04:26 PM

Great stuff. I noticed one thing, though. Arent gems +100% itemization for secondary stats, rather than 50%? If so, would suggest gemmingforhit is optimal, then STR after cap, then reforging for Exp. Also, socket bonuses are fairly powerful. With a three socket piece, and a single yellow socket, using an orange str/haste gem would likely be worth it, instead of staying with hit or str gems and losing bonus.

#4 Fordragon

Fordragon

    Von Kaiser

  • Members
  • 24 posts

Posted 24 September 2012 - 05:12 PM

Gemming str-haste on reds and full haste yellows,yields more dps than going full str gems since they have double secondary stats.

#5 Exemplar

Exemplar

    The One-Eyed Man

  • • Guide Author
  • 1,802 posts

Posted 24 September 2012 - 05:35 PM

I started to rewrite the gem section, then stopped. I'm still inclined to say gem for Str. Here's why.

Let's say you have gear where it is possible to reforge stats to hit and expertise cap. Your options are:
Gem hit/exp to (near) cap. Reforge lesser stats to better stats.
Gem Str. Reforge lesser stats to hit/expertise.

In both you would achieve Hit/Expertise cap. In the first you're just gaining more Secondary Stats, but in the second you gain Str.

Example with numbers:
Hit/Exp cap in gems, 500 extra Haste, 300 Crit, 200 Mastery (total 1000 secondary stats) on gear.
500 Str in gems, Hit/Exp cap via reforge.

Using the values Balhale previously calculated for stat weights (listed in OP):
500 Str vs (500 * 0.54) + (300 * 0.43) + (200 * 0.43) Secondaries
500 vs (270) + (129) + (86)
500 vs 485

Str won. Cap Hit and Expertise via reforging, if possible. Unless you hunt for Exp/Hit items in Dungeons, it's actually hard to cap without gems. Absolutely use the gems if you need to reach caps. However, by the Heroic Epics stage there's a lot of Exp/Hit and it's easy to cap (easy to overcap, actually).

That said, Fordragon is accurate on Yellow sockets. OP will be updated appropriately.
Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

#6 Zephur

Zephur

    Von Kaiser

  • Members
  • 25 posts

Posted 24 September 2012 - 06:07 PM

I think you misunderstood the argument, Exemplar. Because of double secondary stats on gems, and 2 haste > 1 strength, it's always better to prioritize haste over strength, which means:

  • Fierce Vermilion Onyx (80STR/160HAS) in red slots
  • Quick Sun's Radiance (320HAS) in yellow slots
  • Lightning Wild Jade (160HAS/160HIT) in blue slots

If you find yourself short of the hit or exp caps, even after reforging, use the appropriate hit>exp>haste>str gems instead:

  • Precise Primordial Ruby (320EXP) or Accurate Imperial Amethyst (160EXP/160HIT) or Etched Imperial Amethyst (80STR/160HIT) in red slots
  • Rigid River's Heart (320HIT) or Accurate Imperial Amethyst (160EXP/160HIT) or Etched Imperial Amethyst (80STR/160HIT) in blue slots
  • Wicked Vermilion Onyx (160EXP/160HAS) or Lightning Wild Jade (160HIT/160HAS) in yellow slots


#7 Balhale

Balhale

    Piston Honda

  • Members
  • 156 posts

Posted 24 September 2012 - 07:48 PM

I'm going to fix the reforges on my pre-raid gear set today and run a stat weight sim at the pre raid gear level to see if it's still worth gemming hybrids. I'll probably do a normal one either tomorrow or the day after.

Also, it's worth matching anything that has a yellow or red socket, and probably most blue sockets as well, especially at lower gear levels where hit/exp overcapping is not as much of a concern.

#8 gibborim

gibborim

    Glass Joe

  • Members
  • 2 posts

Posted 24 September 2012 - 08:20 PM

I think you misunderstood the argument, Exemplar. Because of double secondary stats on gems, and 2 haste > 1 strength, it's always better to prioritize haste over strength, which means:


The stat weights you are basing this assertion on are the T14 Heroic Gear stats. Stat weights for pre-raid and early raid are more favorable to Strength and lead to 1 Str > 2 Haste. Seems better to have a separate weighting/section for release gearing and remove it after a few weeks.

#9 Exemplar

Exemplar

    The One-Eyed Man

  • • Guide Author
  • 1,802 posts

Posted 24 September 2012 - 08:56 PM

I think you misunderstood the argument, Exemplar. Because of double secondary stats on gems, and 2 haste > 1 strength.


The argument by Wrathblood was clearly Hit/Exp over Str. Which I explained is unlikely to be a net gain. Haste over Str, as Fordragon suggests, is definitely wise. The OP was updated to cover both portions.

Balhale - in a thrown together Dungeon gear set I get Haste at 0.46 Str, Crit at 0.29 Str, and Mastery 0.28.

So rule of thumb would be to gem Str until you get epics. You could continue to do so and have very little lost opportunity.

By the end of the tier you can follow:
Red - Str/Haste gems (Orange - Fierce).
Yellow - Haste gems (Yellow - Quick).
Blue - Str/Hit gems (Purple - Etched).

You always match colours (or typically hybrid colours) for socket bonuses. They're worth it.

Only use further Hit (or any Expertise) gems if you cannot reforge or enchant your way to caps.

OP updated per gibborim's suggestion. Pre-raid and raid gear gemming differentiated.

Somewhat related: I'm inclined towards Accuracy cloak, unless you literally cannot reforge away enough Hit. Why? Worst case you wind up with extra Mastery/Crit (effectively same as enchanting Superior Crit), best case you net some extra Haste and an overall gain (since it's better than Crit). Since the boot enchant is guaranteed Haste, I wouldn't try the same gimmick there.
Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

#10 Tobrexa

Tobrexa

    Von Kaiser

  • Members
  • 89 posts

Posted 24 September 2012 - 11:18 PM

Blue - Str/Hit gems (Purple - Etched).


Shouldn't it be: Blue - Haste/Hit gems (Green - "Lightning")?


Edit: to reiterate my comment on SoR proccing on DS: it hits only once per single DS hit: you could also say that the initial DS hit procs a normal SoR. SoR scales linearly with the number of mobs hit by DS

Edit2: tested HotR, and it seems that it applies the active seal on the primary target, that is to say: SoT/Censure and SoR AoE. Weird behaviour of SoJ: also hit every target in range, not just primary target... couldn't verify if it applied snare debuff

Edit3: HotR applies SoJ and SoJ debuff to every target in range

Go tell the Spartans, passerby:
That here, by Spartan law, we lie.


#11 Balhale

Balhale

    Piston Honda

  • Members
  • 156 posts

Posted 24 September 2012 - 11:20 PM

You literally cannot reforge away enough hit in T14H gear if you also want to match the blue sockets and the hit bonus on the boots. As I'm going through getting my pre-raid (463) set reoptimized it's exactly the opposite. I can barely scrape together enough hit and exp.

Why str/hit in blues instead of haste/hit? If str/haste is better than str then haste/hit should be better than str/hit.

Anyway, here are the results for the pre-raid gear set stat sim both using SW:

This (higher dps) set has as much haste as I could get without dropping horribly under hit/exp cap (14160 str, 2549 hit, 2535 exp, 4725 haste, 2564 crit, 917 mastery):
http://5.chart.apis....&chts=dddddd,18

Normalized to Str:
Str - 1.00
Hit - 0.79
Exp - 0.56
Haste - 0.48
Mastery - 0.30
Crit - 0.30

My results agree pretty well with yours then.

With a less haste focused set (14337 str, 2550 hit, 2542 exp, 2901 haste, 2725 crit, 2237 mastery), I got:
http://4.chart.apis....&chts=dddddd,18

Normalized to Str:
Str - 1.00
Hit - 0.77
Exp - 0.54
Haste - 0.59
Mastery - 0.30
Crit - 0.31

Both sets are gemmed for strength.

Getting a Lessons of the Darkmaster trinket is a huge priority - the thing has about a quarter of the needed expertise on it.

#12 Balhale

Balhale

    Piston Honda

  • Members
  • 156 posts

Posted 24 September 2012 - 11:41 PM

I also ran some Windsong vs. Dancing Steel runs.

Windsong (1500 crit,mastery,haste):
PreRaid - 64166 dps, 19.5% uptime on each buff
T14H - 113210 dps, 21.8% uptime on each buff

Dancing Steel (1650 str):
PreRaid - 64619 dps, 45.5% uptime on the buff
T14H - 114763 dps, 51.4% uptime on the buff

At preraid gear level, Dancing is a 0.7% increase over Windsong. At T14H gear level, it is a 1.4% increase over Windsong.

Your enthusiastic praise of Windsong over Dancing should probably be toned down, but Windsong isn't a bad alternative for people who can't afford Dancing.

#13 Exemplar

Exemplar

    The One-Eyed Man

  • • Guide Author
  • 1,802 posts

Posted 25 September 2012 - 12:52 PM

Shouldn't it be: Blue - Haste/Hit gems (Green - "Lightning")?

Probably. OP updated for Epic Raid gear gemming.

Not sure I understand your DS+SoR comment. DS hits 4 targets, do you get 4 SoR, or 16 (1 per target, per hit)? 4 SoR is original assumption (one proc that cleaves), while 16 is the definition of exponential (4^2). Original assumption requires 6 targets. Exponent would reduce necessary total targets.

I also ran some Windsong vs. Dancing Steel runs.


OP updated with your info and more caveats, but I do not understand some of your numbers. According to Blizzard the mechanics are:
Dancing Steel: 1PPM on melee attacks that land.
Windsong: 1PPM on melee damage, or non-periodic spell damage/healing, with a 1-second cooldown.

So Windsong has more proc chances (we'll have to determine if things like apply/refresh Censure count). Windsong can simul-proc all 3 buffs. Both enchants have identical 12 sec durations. I don't see anything to incline to more/longer procs for Dancing Steel.

I'm not sure how Dancing Steel could have up-time of more than twice Windsong, or are you saying Haste has a 19.5-21.8% up-time, Mastery a separate 19.5-21.8% up-time, and Crit a separate 19.5-21.8% (with reasonable overlap between the three)? If Censure applications or Seal procs can proc Windsong (remember the proc-fest that was Avalanche?), it could have significantly higher uptime than presently predicted by your Sim.

I'm unaware of extensive Ret testing in Beta. As far as I can tell no one even tested to figure out PPM, we eventually were handed them, much less test edge cases specific to individual classes. Pretty easy test - autoattack for prolonged period with no haste and see if we're getting 1 (only autoattack), 2 (seal or censure app, not both), 3 (auto, seal, and censure app), or more PPM.
Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

#14 Balhale

Balhale

    Piston Honda

  • Members
  • 156 posts

Posted 25 September 2012 - 03:44 PM

Dancing Steel: 1PPM on melee attacks that land.
Windsong: 1PPM on melee damage, or non-periodic spell damage/healing, with a 1-second cooldown.

So Windsong has more proc chances (we'll have to determine if things like apply/refresh Censure count). Windsong can simul-proc all 3 buffs. Both enchants have identical 12 sec durations. I don't see anything to incline to more/longer procs for Dancing Steel.

I'm not sure how Dancing Steel could have up-time of more than twice Windsong, or are you saying Haste has a 19.5-21.8% up-time, Mastery a separate 19.5-21.8% up-time, and Crit a separate 19.5-21.8% (with reasonable overlap between the three)? If Censure applications or Seal procs can proc Windsong (remember the proc-fest that was Avalanche?), it could have significantly higher uptime than presently predicted by your Sim.

I'm unaware of extensive Ret testing in Beta. As far as I can tell no one even tested to figure out PPM, we eventually were handed them, much less test edge cases specific to individual classes. Pretty easy test - autoattack for prolonged period with no haste and see if we're getting 1 (only autoattack), 2 (seal or censure app, not both), 3 (auto, seal, and censure app), or more PPM.


I'm saying that each separate buff has 19.5-21.8% uptime. So total Windsong has 60-65% uptime vs. Dancing's 45-50%. Anyways, I'll do the check when I get a weapon with Windsong on live.

However, those buff numbers are roughly what I remember from beta when I messed around with Windsong.

#15 Tobrexa

Tobrexa

    Von Kaiser

  • Members
  • 89 posts

Posted 25 September 2012 - 06:23 PM

Not sure I understand your DS+SoR comment. DS hits 4 targets, do you get 4 SoR, or 16 (1 per target, per hit)? 4 SoR is original assumption (one proc that cleaves), while 16 is the definition of exponential (4^2). Original assumption requires 6 targets. Exponent would reduce necessary total targets.


It's linear, not exponential. example: 4 DS hits, 4 SoR hits. I thought DS would not proc any seal.

Go tell the Spartans, passerby:
That here, by Spartan law, we lie.


#16 Exemplar

Exemplar

    The One-Eyed Man

  • • Guide Author
  • 1,802 posts

Posted 25 September 2012 - 07:07 PM

It's linear, not exponential. example: 4 DS hits, 4 SoR hits. I thought DS would not proc any seal.


Then SoR takes 6+ targets. Not counting loss incurred by GCD to swap seals. DS was resolved to only proc one seal on use an Xpac or two ago. I misunderstood your comment in the other thread and took it to mean this fix had been broken in MoP. OP tweaked back to original SoR commentary.
Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

#17 Elidra

Elidra

    Glass Joe

  • Members
  • 19 posts

Posted 26 September 2012 - 10:00 AM

If haste gems are superior to strength in raid gear and beyond, shouldn't the BS sockets be filled with pure haste gems instead of strength?

#18 Fordragon

Fordragon

    Von Kaiser

  • Members
  • 24 posts

Posted 26 September 2012 - 11:50 AM

Btw something that was bothering me Balhale,is SimC accounting the 10% crit from Inq?

#19 d07.RiV

d07.RiV

    Glass Joe

  • Members
  • 6 posts

Posted 26 September 2012 - 12:42 PM

So it would appear that as long as you can reforge out of hit/exp into haste, it could be benefitial to gem full exp/hit/haste over str? It would be a bit risky because 2 haste is only slightly better than 1 str, and you can't always reforge perfectly so there's always risk of losing more stats than gaining. Also, losing a lot of str could cause haste value to drop below 50%.

#20 Zephur

Zephur

    Von Kaiser

  • Members
  • 25 posts

Posted 26 September 2012 - 03:27 PM

Btw something that was bothering me Balhale,is SimC accounting the 10% crit from Inq?


Yes, it is. The raidbuffed stats snapshot does not show 10% from Inquisition because it takes place before combat begins. The same thing happened with Eye of Unmaking; the strength doesn't show up in the snapshot, because the buff is only up during combat.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users