Thanks, given your 4,663,806 raw healing and 29% uptime (high from what I have seen of other logs) the total accredited to the shield seems about right but does assume 100% shield usage.

The reason you are getting 20% out of the shield is that you have 80% overhealing, the more actual healing done the less the shield will contribute to your performance. if the same proc had occured to your holy priest Kinsie it would have generated only 9% of thier actual healing. The other issue is how many of the shield procs actually got used which sadly we can't track. Not having done hard mode Mim how many people will take further damage within the 8 second window? my experience of divine Aegis is that a supprisingly high % of shields are not consumed.

I wouldn't say that the mace favors overhealing. The problem with your comparison is that you are comparing shield value as a percentage of effective healing instead of just total shield value. If the shields will mostly get used no matter what (like on most of Mimiron Hard Mode), the percentage of your effective healing that comes from the shields increases as your overhealing increases. This is because the value of the shields is based on effective healing + overhealing. So increasing the percentage of overhealing doesn't affect the value of the Val'anyr shields, it just decreases effective healing from normal spells so increases the percentage of effective healing that comes from the shields.

That's an... interesting way to look at it I suppose. It still favors overhealing, and yes, it was used in an ideal situation in Mimiron to the tune of preventative effective healing with the HL glyph splash and it appears to inflate the effectiveness of the weapon. I mostly posted to show how effective it CAN be (as a percentage of effective healing). It is a bit ridiculous that one must overheal that much to get a lot out of it.

My primary shield almost always gets used on any fight (tank healing paladin POV), and that for sure favors overhealing. I suppose that's just another reason why a raid healer is less effective with the mace, there may not be "guaranteed" incoming damage if they are primarily healing the raid. Most Ulduar hard fights have some sort of regular raid damage though, so the fights mostly ensure that a larger shield on several raid targets will be consumed. Entirely fight dependent though.

Even from my POV I find that an orange that scales with high throughput + overhealing is a stupid thing for Blizzard to implement. For a healing item that is supposed to be the best in the game for a long period of time, they can do better. They designed it entirely from a DPS itemization viewpoint.

Val'anyr scales with throughput, not with overhealing. Here's a chart showing how thinking that Val'anyr scales with overhealing is misleading. This is in a very simplified case of 2 heals for 100 HP, the first of which procs Val'anyr (so the second generates a 15 HP shield).

Overhealing (%)	Effective Healing from spells	Effective healing from Val'anyr Shield	Percentage of Total Effective healing from Val'anyr Shield
0%	200	15	7%
25%	150	15	9%
50%	100	15	13%
75%	50	15	23%
90%	20	15	43%
99%	2	15	88%

As you can see, more overhealing doesn't result in more effective healing (damage prevention) from shields from Val'anyr, it just inflates the percentage. The only thing that Val'anyr Shield effectiveness scales with is the total healing (including overhealing) output during the proc. This is all assuming that all shields actually get used up, of course. During a proc, it is in your best interest to overheal if the shields will be used up, but saying that this means that it favors overhealing seems a little overblown to me.

Mimiron Hard 25 Paladin with Val'Anyr, WoL: World of Logs - Real Time Raid Analysis

Zoning in on P2, which ensures most shields are used due to the pervasive fire nova: World of Logs - Real Time Raid Analysis
* Protection of Ancient Kings: 92259

Rough P2 split using WWS: Wow Web Stats

First Val'anyr proc WWS split: Wow Web Stats - 125330 total healing output

Second Val'anyr proc WWS split: Wow Web Stats - 205524 total healing output

The theoretical absorb value should then be: (.15)(125330) + (.15)(205524) = 49627, half of the reported WoL absorb value

Digging a little deeper, let's see who used the shields:

WoL: World of Logs - Real Time Raid Analysis
* Protection of Ancient Kings on myself (I beacon myself in P2 as a point of context): 12474
* Protection of Ancient Kings on Duoctane: 936

Using the Val'anyr WWS splits above:
* Total healing on myself respectively: 23,086 + 34,884 = 57970, or an expected potential 8695 absorbs
* Total healing on Duoctane respectively: 0 + 4084 = 4084 or an expected 613 potential absorbs

There's a lot more digging to be done here (e.g, rather than cross-referencing against WWS, split to the Val'anyr procs in WoL directly to see what its reported values are). Assuming WWS is accurate in its total healing, it appears that either WoL inflates the absorb value or the 15% in the tooltip is incorrect (very likely that WoL is inflating due to the testing nature of the absorb parser). It would be one thing if WoL was reporting a lesser absorb value than that which would be possible, but the fact that its values exceed the theoretical maximum either points to a problem.

I'm using mostly an empirical look; if anyone wants to do a mathematical approach, by all means. My next step is likely to set up some type of test to validate the 15% number listed in the tooltip just to get that unknown off the table.

Val'anyr just got a nice buff on the PTR.

Quoted by mmo-champion, picture of the "new" one:
http://static.mmo-champion.com/mmoc/.../valanyr32.jpg

It's a buff all right but I still suspect that many healers would prefer the hard mode 701 spellpower weapons to this because of the erratic nature of the proc. No mention of whether or not they have fixed the known bugs with hots etc.

Clearly they don't want weapon swapping for the proc effect.

Pala of my guild did some tests:
- Sacred Shield spamming procs it
- beacon spamming doesn't proc it
- rezz spamming doesn't proc it (we were bored)
- overhealing doesn't proc it (old news)
- blessing spam doesn't proc it
- JoL does not proc it

So it boils down to: SS and any effective direct healing.

I'm also wondering of WoL is parsing the proc healing correctly, because it seems we (or at least I) have underestimated the mace proc a bit in that case. In an entire night of Ulduar, it's about 15% effective healing, topping to 25% on most high AoE dmg bossfights.

edit: he just did loatheb, you can proc it with SS and just spam HL during the aura and people will get shields it seems.
edit2: as the posters behind me say, it procs from overhealing. My mistake.

The hammer definitely procs off of direct heals which are 100% overheal; I was pretty sure this was the case, but I just logged on my Paladin to double check and spammed some heals on myself in Orgrimmar. I believe it is HoTs which are 100% overheal that do not cause it to proc.

It definitely procs from overhealing with direct healing. Just to re-affirm.

Going to add one more piece of random facts: once you receive the buff, you can swap weapons and will keep the buff. (Of course you lose a GCD by swapping the weapon).

Can people please stop linking World of Logs parses until they have fixed the well documented flaws in how they estimate the effects. All the usual flaws and artefacts are present in the linked parses. You generally need to drop the contribution from the hammer by 30% or more because of flaws in how they record things.

That said the most obvious flaw using the thorim parse is that to get a shield size of 1441.7 on average the heal proccing it needs to be 9600. this is far far more than any holy spell save greater heal can generate. If you look at the procs on yourself it is even worse, you procced it 3 times for an average of 2360 each, this requires an activating heal of 15733 which is patently idiotic. I could go on in this vein for a while but I don't want to bore everyone to tears, if you want to know more about these flaws send me a private tell.