From what I can tell, the IBF buff was directed more at PvP than PvE. IBF is generally only used during the "oh crap" moments (like Festergut 3rd inhale), and going from 47% reduction to 57% reduction, while significant, isn't really meaningful. Yeah, you'll be taking a fair bit less damage, but chances are all the healers are bombing you anyway, so the added reduction isn't THAT helpful.

As for between CDs: I've never understood the claims of "spiky damage." By not having a block mechanism, all of our DR comes from armor (and talents). We have, more or less, the same avoidance as the other tank classes. That means any given hit will hit us for around the same damage, as compared to say, a warrior, who can take wildly different amounts of damage due to crit blocking. The only other cause of perceived spike damage I can think of is avoidance strings, which as I mentioned, should be pretty similar across classes.

If you are talking about the miss cap as found by Whitetooth that is not a hard cap (e.i. limits miss chance to maximum of 16%) but instead relates to the diminishing returns formula affecting how quickly subsequent gains are reduced.

I have not seen any data indicating a hard miss cap and simple testing on low level mobs, such as that done in Whitetooth's own tests, shows that you can get miss rates higher than 16%.

Ismail: An excellent dig up on an old thread. I did not however see in the entire post anything that suggested it isn't a hard cap. Several tests demonstrate how "the miss cap is 16". White's tests consecutively demonstrate "cap<20%", "cap<17%", and then three tests which all basically show "cap=16%".

I'm slightly puzzled how you came to the conclusion that it's somehow a non-hard cap. In which way is the cap not hard? Usually soft-cap denotes an inflection point in incremental functionality where a variable is partially negated, lowered in effectivness, or pointless. We say Exp soft-caps when it stops giving dodge protection, a point where it's value effectively halves. How do you propose the miss cap is "soft"? DR does not cause inflection points, it's a gradual decrease, however DR and hard-cap are two separate functions.

I'd like to see some more evidence of the miss cap not being hard-set at 16%.

Separate post because it's a separate question: Now that the changes to what procs what mean Necrosis isn't the TPS increase it once was, are we better off dumping that last floating point of Blood Tanking specs into Subversion?

I will explain in two ways first you can look at the screen shots and see that see that the achieved miss rate is above 16% and that the miss gained from defense rating changes when he adjusts the cap which he does until the achieved miss rate matches the expected rate. The expected miss rate was 20.37% in the first test and 19.49% in the last test and is always above the cap being tested. Second has a specific value relating to his main post in the thread.

Some background math on DR

We know how to calculate DR

is the diminished stat before converting to IEEE754.
is the stat before diminishing returns (can be found with simple math).
is the cap of the stat, and changes with class (value being test).
is is a value that changes with class (found in other testing).

We also know if we are right at 100% avoidance

Here
Thus , , and are known terms so can be solved for.

They way he confirmed the in his tests was to guess at the value of , this then changed which he tested for by gearing for 100% avoidance using that number. If his was to high then his would be to high and he would be hit. He then lowered the giving a lower and regeared to 100% and retested until he found a value that matched his test results.

He had 10.6% miss from the level difference and 8.89% from base plus gear. Since he expects 19.49% miss total and since you can easily see on low level mobs that the miss rate goes higher than 16% you would have to argue that miss based on level difference is excluded from the hypothetical hard cap. But his 8.89% was not high enough to test a 16% hard miss cap that excludes miss chance from level difference.

So if there is a 16% hard miss cap on base miss chance + miss from talents and abilities + miss from gear it was not found in this set of testing or any other set I have seen or been able to find.

I would but also for Festergut threat reduction during your turn to pwn.

Subv and Necrosis were very similar in threat value prior to the RS change (1.1% vs 1.0% TPS in favour of Necrosis prior to patch, based on Karhorie with i245 gear test), so whilst I haven't tested again, I'd suspect Subv is now equal TPS and the minus-threat advantage swings it in Subv's favour, for me.

I've always been skeptical about the 3s ICD of Bone Shield; it just doesn't seem to hold true. So I went off to the Valley of Lost Heroes to test it. The idea was to get as many mobs attacking me at once as possible, ensuring hits by turning my back to them. Sadly, the combat log doesn't seem to log the loss of individual bone shield charges, but it does log the cast and total removal, of course:

As you can see, the timing between the AURA_APPLIED and AURA_REMOVED are 7.01 and 6.60 seconds, respectively. The second test involved twice as many mobs. I was glyphed for this test, so at cast there were four charges. This would imply:

The exact values for XX depend on the mob swing times (I had around 10 mobs on the first pull and 20 on the second), but overall, it is clear the internal cooldown is not 3s as is generally accepted. I think it is more likely to be 2s as that allows for swing time (when engaging 20 mobs of the same type, their swing timers generally were synchronized, so it wasn't a perfectly steady flow of hits).

The key takeaway is that the uptime of Bone Shield, sadly, can be 33% lower than was previously believed; rather than 9s-12s, it is more like 6-8s, depending on boss swing time etc. Of course, avoidance is still quite nice and one should measure their own uptime via WoL or other tools, but the generally accepted 3s definitely is not correct.

Sudden Doom vs. Necrosis in a blood build is not a simple mathematical issue. Good Blood tanks often choose Death Strike over 2xHeart Strike during periods of high incoming damage to themselves or the raid.

The "floater" point in my blood build is best in Sudden Doom from a TPS perspective when HS is prioritized exclusively. I haven't determined which is better when DS is prioritized. If Necrosis is better in that scenario, it can then also be determined where the threshold is between the two.

In either case, the difference is very small so I haven't messed with it yet.


I'm still looking for more feedback about the Unholy and Frost sections of the original post, so please chime in folks.

Just to further elaborate on this, Whitetooth's testing was in relation to the diminishing returns; I'm fairly certain that it's been demonstrated that talents such as Anticipation (5% dodge) do not contribute to the curve or the cap in any way, and are simply directly added on after diminishing effects are applied to the rating curve.

Thank you for testing that. I had thought BS used a 2.5s ICD, but I had not seen someone test it and show me logs.
Your data sugggests in fact the ICD is less than 2.0s, though.

You generously looked at the 7s duration, when the 6.6s is more accurate of course, since we're testing ICD, the fastest-possible loss of BS is the better measure.

Thus:

Which is 1.65s ICD, perhaps for Blizzard neatness it's actually 1.5s... which is really lame.

Suno, re OP Unholy, this might be a point to consider:
it's a good spec for kiting oozes (bone shield reduces your damage taken for most of the fight, plus its 6% magic reduction is a nice bonus); possibly will be valuable for hard-mode kiting.

You had back turned and very low incoming avoidance, too, right, which means the only other factor would be level difference between mobs and you (they'd be 10%+ chance to miss?). Do you think the number of mobs attacking you compensated for their higher miss chance?

I tend to think so; I had 40+ mobs attacking me, and they seem to have a 2s swing timer. I only saw around a 14.9% miss rate (127 attacks, 19 misses; zero dodges/parries as my back was turned). This is just over the 6s when I had bone shield up. 127 landed attacks over 6s is roughly 20 per second; there were some delays of 0.2-0.3s but in general it was less than 0.05s.

One thing that struck me as funny in regards to the BIS list, is that with the gear that is suggested, you only get around 4.25% hit rating, and you get 27 expertise(very nice btw). I know as a tank(at least in my case), I'd want at least 1.5% more hit than that. 6% is where I'm most comfortable at because we always make sure to have a Dranaei in the group with tanks, so therefore I'd end up with 7% hit(granted I could still miss, but very unlikely).

I'm just wondering if that was intended or not, because if you would swap out the Tier 10 helm for the Broken Ram Skull head piece for Lady Deathwhisper, you'd end up with the extra 1.5% to be back up to 6%, eat some hit food and figure on 6.5% or so hit.

I'm not saying that the list is wrong, but I always thought trying to maintain a decent amount of hit(while of course being able to survive and hold aggro) was our jobs.

You can use the rep ring (55 hit), put 20 hit to your gloves and use Vivid gem (10 hit/15 stam for yellow slots). With Tricks being around, missing a few attacks isn't much of an issue.

Honestly, the only time you should really be worried about your hit is when you're having threat issues against your dps. Outside of that, with two taunts at our disposal to help with misses during tank swaps, hit doesn't have as much value as additional armor, health, or avoidance stats. So far none of the fights in ICC seem to be hugely threat sensitive. We'll see how the hard modes work out, but right now I wouldn't trade out those pieces unless you found yourself unable to stay ahead of your dps in threat.

Which extra survival talents are you talking about then? You could probably get away with ditching some threat talents, or at least a few points from them, but what would you replace them with? As blood, there's nothing else worthwhile I can reach at least.

Besides, I wouldn't count on a continuous stream of MD. Our rogues trick each other after the initial TotT on me on the pull, and although possible you'd probably have to pester the hunters for MD's alot (esp. with hardmodes in mind it might be unwhise to have to use all MD's on the boss rather than e.g. adds).

Yes, unlike the other tanks that have 50 points of a Prot tree and then spend another 20 on DPS talents, we have about 20 points of survival talents and spend the rest on DPS talents.

Every once in a while someone will "discover" the 23/43/5 (physical) or 23/5/43 (magical) type builds but outside of a situation like Sarth where threat literally does not matter, this is just a terrible idea. The last 5-point talents in each respective tree represent an incredible amount of TPS and there's no way to make up for the lack of Tundra Stalker or Rage of Rivendare with Tricks/MD.

Do you have any threat for survival tradeoffs in mind? With the current structure of our talent trees there really doesn't seem to be any real gains to be had unless you completely abandon all pretense at TPS.

I haven't actually looked specifically. Off the top of my head I imagine you could shift points and have Rune Tap/Imp Rune Tap/Mark of Blood as part of a standard build or just keep Acclimation all the time.

A lot of this really just sprung up as part of the in-guild debate over the recent paladin nerf/IBF, Frost Presence buff. (Who doesn't love the occasional and completely subjective better tank debate?). Our prot paladin seems to be on the "threat no issue" side and i'm actually curious if this is the case. If it is, surely a short cooldown self heal would occasionally serve a purpose whereas a threat talent (Sudden Doom? Necrosis?) would not if the whole "threat no issue" thing is actually solid.

For Blood tanks it's certainly acceptable to drop Necrosis/SD/Subversion for Imp RT but that may depend on your healers. Our healing team consists of at least 2 paladins that bomb the tanks anyway where it matters (3 stack Fester and so on) together with 2 druids keeping up hots so I find the self heal is usually completely obsolete. However, I may be blessed with excellent healers so YMMV. If necessary you can still line up DS with incoming hits or pop Vamp Blood first for even more (self) healing, for the moment I'll stick to threat over RT but that might change when we start doing hardmodes

Speaking of Necrosis having lost favour, I'm wondering if we're wasting the three points on Morbidity: DnD CD is irrelevant outside trash (unless you're DnD on single-target? Does anyone do that?) so strictly, you could ditch not only the point from necrosis but also the three points from that, and spec into SD and Sub. Personally I've gone for Spell Deflection, which may not be the hottest choice on the block but I've seen it go off quite a bit.

Somebody mentioned slime tanking. I think there's good mileage to be had from a modified UH build that'll swap Rage of Rivendare for more blood talents, aiming to get Spell Deflection and with a weapon enchanted with Spellshattering. Imp. UH Presence is also a no-brainer for this setup, however we'll have to see if it's warranted first.

That's assuming we will actually need to tank them on heroic mode (or at least, for some periods of time). In normal mode we can get by without gtting hit a single time, even running through the slime as long as you save AMS for it).

I'm already using 3/3 SV and 1/3 SD, I'm not sure swapping is worth it. To start with, I really like the short DnD cd for picking up waves of adds like on Valithria. I'd gain 10% per HS (1point atm) for a free DC - which in turn hits for less dmg due to 0/3 Morbidity. Yes it's extra threat but at a cost of some flexibility.

Swapping 1 point from Necrosis to SD sounds worthwhile but as Suno mentioned they're still quite close on threat - esp. in heavier healing fights where you use death runes to DS more.