## Meta gems for warrior tanks in Cataclysm

December 12, 2010 at 10:35 am | Posted in Uncategorized | 10 CommentsPlease ignore this article. It looks like ESD will not provide a 5% buff, but instead only 1%. (21 Jan 2011)

*This article examines the relative merits of three different metagems (ASD, EfSD and ESD) for tanking in Cataclysm content. Mastery scaling for warrior tanks is discussed in order to effectively analyze the benefit of block value bonuses. I conclude that ESD is better than ASD in most cases, while fight-specific factors may influence the choice between ESD and EfSD.*

Note: substantial changes have been made to this article thanks to Button pointing out some errors I had made in my initial assumptions. (14 Dec 2010)

At present, there seem to be three tanking meta gems available in Cataclysm:

- Austere Shadowspirit Diamond (ASD): +81 stamina, +2% increased armor value from items
- Effulgent Shadowsirit Diamond (EfSD): +81 stamina, reduce spell damage taken by 2%
- Eternal Shadowspirit Diamond (ESD): +81 stamina, +5% shield block value*

(*) Currently ESD appears to be bugged. Rather than providing 5% BV, it only gives 1%. See Maintankadin and Zarko for details.

Selecting the right meta will initially depend on the expected incoming damage and the split between magical and physical damage. If your raid’s current progression fights are heavy on magic damage, EfSD may win out. If physical damage predominates, however, the choice is between ASD and ESD. Making the right choice requires some calculations.

In Wrath, AED was the clear choice because of the importance of effective health. More armor meant more effective health while the benefits of reduced overall damage were less important due to strong healer mana regeneration. In Cataclysm, these conditions no longer hold true; tanks have much more health relative to expected burst damage and healer mana now matters a great deal more.

At least for warriors, the scaling on mastery has a big effect on the value of ESD. At low mastery values, block chance will be low and, therefore, the extra 5% block value is rarely applied. But as mastery grows, more incoming swings are blocked and the 5% bonus is used more and more. In the calculations that follow, I assume that the 5% bonus is doubly useful when critical blocks occur, which is, again, a more common occurrence when mastery is high. In other words, prior to reaching unhittable, **mastery scaling is better than linear.** Recall that, due to diminishing returns, dodge and parry are *worse* than linear.

I played around with the avoidance numbers a little. Assuming a warrior tank keeps dodge, parry and mastery ratings equal and has +agility only from a leg patch, here are my conclusions (percentage represents the value of mastery relative to dodge in terms of expected damage reduction; >100% means mastery is better):

Without ESD | With ESD | |

1000 rating | 82% | 95% |

1200 rating | 85% | 100% |

1400 rating | 90% | 105% |

1600 rating | 96% | 111% |

1800 rating | 100% | 117% |

While the table above may be suggestive, it doesn’t do a good job of presenting the options available to tanks in terms of reforging. When is it best to reforge dodge (or parry, which is roughly equal to dodge in most situations) into mastery for the purpose of reducing incoming damage? Let’s look at an example that draws on Zarko’s expected raid tank stats. Our imagined warrior tank at 85 will have, prior to gemming/enchanting/reforging, but with raid buffs:

- 40477 armor with ASD (without ASD, 39768)
- 2220 dodge rating (2130 from gear)
- 1511 parry rating
- 2183 mastery rating
- 634 agility

At this point, both dodge and parry are relatively high, producing diminishing returns effects. Meanwhile mastery is high, which may make it the better choice for reforging. Only 40% of a stat can be reforged to something else, so that is a hard upper bound on redistributing ratings. Practically, though, this is not always possible because the desired stat may already be on an item, preventing reforging from one to the other. Let’s suppose 25% of the dodge rating (532) is actually available for reforging.

So, what happens to damage reduction via avoidance when we reforge? Here are my results:

no change | reforge to parry | reforge to mastery | |

dodge % | 15.84 | 13.63 | 13.63 |

parry % | 12.87 | 15.21 | 12.87 |

block % | 50.27 | 50.27 | 54.72 |

crit block % | 36.27 | 36.27 | 40.72 |

damage reduction % via avoidance (with ESD) | 57.68 | 57.81 | 58.45 |

damage reduction % via avoidance (without ESD) | 54.26 | 54.39 | 54.60 |

So, at least in terms of avoidance stats, it appears that perhaps mastery is king once tanks have moved reasonably deep into the first tier of raid content. This is independent of metagem selection, but how does ESD compare against the armor bonus of ASD? Let’s examine.

Using the level 85 armor formula, our tank with will have either 40477 armor (with ASD) or 39768 armor (without ASD). These amounts of armor will, in conjunction with defensive stance and the Inspiration buff, provide 63.88% and 63.53% reduction on physical damage taken. Let’s look at the results when we integrate the armor numbers with the avoidance numbers from above.

First, the table for using ESD:

no change | reforge to parry | reforge to mastery | |

damage reduction % via avoidance | 57.68 | 57.81 | 58.45 |

damage reduction % via armor/def stance | 63.53 | 63.53 | 63.53 |

total damage reduction % | 84.57 | 84.61 | 84.85 |

Now the table for using ASD:

no change | reforge to parry | reforge to mastery | |

damage reduction % via avoidance | 52.26 | 54.39 | 54.60 |

damage reduction % via armor/def stance | 63.88 | 63.88 | 63.88 |

total damage reduction % | 82.76 | 83.53 | 83.60 |

And, for completeness, the table for no meta:

no change | reforge to parry | reforge to mastery | |

damage reduction % via avoidance | 52.26 | 54.39 | 54.60 |

damage reduction % via armor/def stance | 63.53 | 63.53 | 63.53 |

total damage reduction % | 82.59 | 83.37 | 83.44 |

These three tables are a little tricky to interpret. Let’s put them together for comparison:

no change | reforge to parry | reforge to mastery | |

total damage reduction %, no meta | 82.59 | 83.37 |
83.44 |

total damage reduction %, ESD | 84.57 | 84.61 | 84.85 |

total damage reduction %, ASD | 82.76 | 83.53 |
83.60 |

A few things jump out from this table. First, the damage reduction benefits of of ESD under all three reforging conditions are better than those of ASD. In fact, the gain provided by ASD is less than a quarter of that produced by ESD! It is possible that ASD wins at lower levels (eg, beginning heroics and before), but as gear improves beyond the first tier of raids we can expect the gap between ESD and ASD to grow until unhitable is reached (beyond that point, more calculations are needed to evaluate the relative benefits).

Second, the best result in the table occurs when ESD and mastery reforging are used in conjunction. The two interact favorably, so this is perhaps unsurprising. This interaction is tricky and interesting because it could make a big difference in terms of which enchants are useful. Perhaps Enchant Shield – Blocking will be the preferred enchant beyond a given mastery rating.

Third, the difference between ASD/no reforge and ESD/mastery reforge is *huge*. While the absolute difference is only 2.09%, consider the increase in damage by moving from an ESD/mastery tank to an ASD/nothing tank. The ESD/m tank takes 15.15% of the incoming swing damage while the ASD/n tank takes 17.24% damage. This is a relative increase of 13.8% (=17.24/15.15), which represents a sizable amount of healer mana. With most Wrath tanks being accustomed to taking the austere meta and not having the option to reforge, I expect to see a lot of suboptimal gearing decisions for the next several months.

Please note that all of the above is modeling only passive mitigation. As always, using cooldowns, proper positioning, and self-healing abilities will all play a role in the outcomes of actual boss fights.

Anyhow, what do my wise and worldly readers think? I hope this provides an interesting resource for making gear decisions as a warrior tank. I’ve done my best to get the various calculations right, but please let me know if I’ve screwed up any of the math.

## 10 Comments »

RSS feed for comments on this post. TrackBack URI

### Leave a Reply

Blog at WordPress.com. | The Pool Theme.

Entries and comments feeds.

Definitely interesting.

One small math problem – when you stripped out Zarko’s buff numbers for devo and meta you also stripped out Toughness, which he counted as a buff for some reason. That will change the armor numbers significantly, though I can’t estimate by how much off the top of my head.

The rumor on wowhead is that this meta gem actually only increases your block value by 1% (displayed value; it could actually be 1.5%, which would make sense as that’s 5% of 30). Are you using 1%, 1.5%, or 5% in your calculations? And are you doubling the contributions of the ESD on crit block or adding it after the fact?

Comment by Button— December 13, 2010 #

Figuring out armor buffs was something that looked like a bigger project than I wanted to touch at this point, but you’re right that it introduces some error into my calculations. I suppose I should re-run the math with some arbitrary amount of armor added, just to see what happens. I may do that later this week, time allowing.

Regarding block value numbers, I am using the following values: no ESD, non-crit is 30%; no ESD, crit is 60%; ESD, non-crit is 35%; and ESD, crit is 70%. I felt the 35% number was reasonable since the Eternal meta from Wrath, which says +1% block value, produces a tooltip value of 31% rather than 30.3%. Whether the 5% should be added pre- or post-block is a question I really haven’t investigated and it could easily turn out that I am wrong.

Argh! Wow math is such a pain in the butt because percents are thrown around without any reference to per cent

of what.Comment by Lujanera— December 13, 2010 #

Oh I totally understand not wanting to use devo aura in calculations, because you don’t always have a paladin or shaman in the group. Toughness, though, a) is always on and b) is multiplicative with the 2% armor increase from ASD.

The JCs in our lives need to level so we can know which block value to use!

Comment by Button— December 14, 2010 #

Alright, I spent a little time digging through Zarko’s numbers in more depth. Looks like I overlooked a lot more than I had originally supposed. Sigh. I’m going to have to recalculate most of the numbers in this post, I think.

Comment by Lujanera— December 14, 2010 #

I played around with Zarko’s spreadsheet some more. I changed three values to change the model from ASD to ESD:

-Summary tab, cell B30: 0.7 -> 0.65 for non-crit blocks via ESD

-Summary tab, cell B32: 0.4 -> 0.3 for crit blocks via ESD

-Warrior stats tab, cell E9: 1.02 -> 1.00 for no ASD

The result? About 8% less damage taken, using his calculations. Not too bad, considering that no reforging was done in his example.

Comment by Lujanera— December 14, 2010 #

[...] Now, on to the theorycrafting! Rhidach has laid out his preferred (paladin) tanking enchants for the expansion at Righteous Defense. Because paladins can reasonably block cap his priorities and preferences are a little different from mine and many of yours, but nonetheless it’s a solid starting point for any tank. On the warrior side, Lujanera is working out which meta gem we should use – +armor, or +block value? The math is ongoing on this one, so if you’re a mathy kind of person, come weigh in. [...]

Pingback by A Little Link Love: Mo’ Content, Mo’ Problems « Uncrittable— December 13, 2010 #

I’ve updated this post substantially. The conclusions are mostly the same, though I believe I have more accurately accounted for armor and buffs in this version.

Comment by Lujanera— December 14, 2010 #

The eternal gem is intended to be 1% apparently, you can see this change now on the PTR.

Comment by Gulvak— January 21, 2011 #

Thanks for the update, Gulvak. Looks like this makes the armor meta the superior choice now.

Comment by Lujanera— January 22, 2011 #

I disagree. I’ve been doing some spreadsheeting with the 1% meta assumed and at the moment the SBV meta is performing semi-significantly better than the armor meta for fights with <10% unblockable physical damage. I'm not completely ready to release my spreadsheet but I will be shortly.

I'm on a public computer atm so apologies for the vagueness, I don't have the numbers in front of me.

That said, the armor meta does have the benefit of increased damage smoothing, so even if the Eternal does provide more overall damage reduction it isn't necessarily better.

Comment by Button— January 25, 2011 #