A number of shamans (including Philondra and Grayshock in this thread) noticed that less mana is granted from Improved Water Shield procs than from Water Shield orbs triggered the "normal" way (i.e. by mob damage and environmental factors.) We speculated as to what the mechanic might be, but had no hard data to validate our assumptions. I decided to gather some data to see if we could figure out what the issue was. Please check to see if my calculations and conclusions are accurate.

Abbreviations:
2T7 = the two piece tier 7 set bonus, which increases the strength of Water Shield by 10%
IS = Improved Shields
IWS = Improved Water Shield
GoWM = Glyph of Water Mastery

Methodology:
1.) I started with a standard 0/14/54 + 3 spec with 0/3 IS. I removed all T7 gear and water-shield related glyphs, and went to Crystalsong Forest.
2.) I went to Crystalsong Forest and activated WS.
3.) I attacked a mob and recorded the mana return from each WS orb.
4.) I refreshed WS until I had 50 orbs had triggered, then killed the mob.
5.) Once out of combat, I cast LHW until I recorded 50 crits (refreshing WS as necessary) and recorded the mana return from 50 IWS procs.
6.) I then equipped 2T7 and repeated steps 2-5.
7.) I returned to the Lexicon of Power in Dalaran, added GoWM, removed 2T7, went back to Crystalsong Forest, and repeated steps 2-6.
8.) One I had tested the returns for all permutations of 2T7 and GoWM with 0 points in IS, I added 1 point in improved shields, removed GoWM, and repeated testing as above. I did the same for 2/3 and 3/3 improved shields.

Notes:
1.) As anticipated, GoWM had no effect on mana granted from Water Shield Orbs. As such, data points for tests without GoWM are omitted.
2.) Data listed below for mana returns are averages rounded to the nearest integer. In no case was the variation between procs greater than +/- 1 mana.
3.) In "Test Condition" below, the percentage in parentheses represents the expected gain in mana granted by water shield orbs according to the combination of IS points and the presence or absence of 2T7.

From the data above, we see the following trends:
* The difference in mana granted between a baseline (0/3 IS, no 2T7) WS proc and a baseline IWS proc is 50 mana, or 50% of the passive mp5 granted by a baseline WS.
* Both IS and 2T7 increase the amount of mana granted by both WS and IWS.
* Although the absolute difference between the two increases, the percentage difference remains the same, and is in line with what we predicted should happen given the wording of the IS talent and the 2T7 bonus.

Conclusions:
* For a max rank Water Shield, standard WS procs grant B*(O+P) mana, where B is the total percent bonus to each orb granted by talents and bonuses that improve the per-orb mana return of WS (namely IS and 2T7), O is the tooltip mana per orb for a naked, 0/0/0 spec shaman, and P is 50% of the base passive mp5 of an unglyphed water shield.
* IWS procs, in comparison, only gain mana equal to (B*O); they do not gain any bonus, scaling or otherwise, from the passive mp5 of water shield (P).
* In other words, WS procs gain a baseline modifier to mana return (that happens to scale with IS and 2T7), but IWS procs do not. It is unclear why this discrepancy exists.

The above conclusions are counter intuitive for two reasons:
1.) A logical reader of the tooltip would come to the conclusion that IWS and WS procs should return the same amount of mana. This is clearly not the case as of patch 3.0.3.
2.) The bonus modifier P, which is calculated from the passive mp5 of water shield, is unaffected by abilities that increase only the passive mp5 (GoWS), but *is* affected by abilities that increase only the WS orb procs (Improved Shields) and abilities that affect both (2T7).

I'm not sure if the 50 mana difference really depends on WS's passive mp5 amount. If it does, the difference at level 70 should be 25 mana.

Also, I wonder why the base amount for an IWS proc (untalented and without any bonuses) is 427 mana, and not 400 as the tooltip suggests.
EDIT: Found the answer in the linked thread (level-based scaling).

Your work looks very solid and I can't see any flaws.

Can someone confirm that the values for damage-triggered charges are the same without the IWS talents?

Interesting! Thank you so much for conducting this test. Your table and method were very easy to understand too.

Before wrath came out, I was unsure about the benefits of IWS, but I really like it now and have itemized for crit because of it. Even with the discrepancy, IWS is win.

That being said, this is probably a bug, yes? Has anyone seen blue posts regarding the discrepancy? If it gets fixed in an upcoming patch, that means IWS will be even more OP. Exciting!

I believe this is a bug with Rank 9 of Water Shield. I'm still using Rank 8 and have the exact same procs from combat and heals. (246-247 in my case.)

I'm using Water Mastery, Imp Shields, and IWS and therefore have all the factors except 2T7 in play. Perhaps you could check a Rank 8 as max level and see if you have the same results.

Before I posted I went and checked all ranks of WS available to me at 75, and found that all ranks except Rank 2 are proccing the same from combat and IWS. Rank 2 actually procs more from IWS than combat. (93 IWS, 84 damage).

what about with the water shield totem?

I have done similar testing and noticed the same difference between on the on hit and on crit values.

If I may add to your table ...

Where is that from? I can calculate the IWS return fine, but the on hit value is constantly higher and I can't figure out the source of it either. From these numbers, it looks much like a fraction of the improved values. Could something like purification affect the value AFTER it has received improvements from talents and set?

A minor point, but S5 is supposed to start Dec 17, and presumably the totem would be available after that point if you control wintergrasp.

I tested all ranks of Water Shield (via the spellbook) and Rank 1, 3, 4, 5, 6, 7, and 8 all react the same from IWS and damage. Rank 2 returns more mana from IWS than damage, and Rank 9 returns less mana from IWS than damage.

It has nothing to do with whatever rank your max is, it is a bug with Rank 9.

Has anyone on the PTR determined the % for IWS and LHW in 3.08. I will hop on and run some tests if noone else has, but if the work has already been done I dont want to waste time.


Obviously this wont effect the LHW/ES spam HPS but it will change the HPM, and some numbers will help us reevaluate the LHW spam validity.


Also, has anyone tested the R9 bug in the PTR? Once again I will be happy to test if noone else has.

As of 3 days ago on the PTR:

You have a 33/66/100% chance to instantly consume a Water Shield Orb when you gain a critical effect from your Healing Wave or Riptide spells, and a 19.8/39.6/60% chance when you gain a critical effect from your Lesser Healing Wave spell.

Handyhoof listed the proc percentages, but I have been unable to confirm anything related to WS9, as I have been unable to access the PTR due to the account-flagging issue various people are reporting.

I can also report that although I was unable to get Totem of the Thunderhead after 16 Underbog runs, one of my fellow guild members did -- and he reported that Improved Shield and 2T7 *were* affecting the mana granted from the totem, giving us a maximum of 569 mana from IWS or 631(!) mana from normal triggering.

In other words, with 5/5 Tidal Focus, 3/3 Imp Shields, 2T7 and Totem of the Thunderhead, one normal WS proc will actually give you enough mana to cast one LHW with 5 mana to spare.

Blue post says they are checking this bug out.

World of Warcraft - English (NA) Forums -> [Bug] Improved Water Shield procs

As you mentioned in the bug report thread, I'm pretty sure our data conclusively shows that the problem is NOT with the improved shields talent, as many people assume and I had initially postulated myself, but rather with some sort of invisble modifier being applied to WS9. Hopefully Blizzard can make the WS and IWS mana returns match up so that this issue can be laid to rest.

Seems this isn't being addressed in the upcoming patch 3.0.8 if the official PTR notes are anything to go by (link)

Something I observed which may help with testing water shield.

If you stand on the top step of one of the stores next to the Eventide bank water shield (and lightning shield) will endlessly proc.

Not entirely sure what causes this effect and I certainly agree it is strange...

The reason you proc it is because the no-fly zone is only laid out around the 'outside' parts of the city (with the logic that inside is a no-fly zone anyway). Because WS literally procs from everything, you will get the mana return everytime you enter the no-fly zone and gain a hidden debuff/trigger.

The RNG factor of a 60% procc over the length of the fight is insignificant.

In a parse of my latest Kel Thuzad fight I had 47 crit LHW I don't see this as being enough to make the RNG factor insignificant at all.

In particular, say in a short fight that takes loads of mana (Patchwerk) one good be very low on mana and have bad RNG luck and go oom where without that factor one would not have.

Now, I'll admit it really isn't that big of a deal, but the crux of my argument is that the reason behind the change was IWS being too efficient so there's NO specific reason to change the proc chance instead of lower the amount of mana returned. On the other hand I feel that there is a case for changing from reduced proc chance to just lowering the amount of mana returned.

I apologize in advance for the main question being a derailment.

Those who use Improved Water Shield on Patchwerk: Don't you think there's a higher probability of bad RNG that you just refreshed WaterShield as Patchwerk landed a hateful, leaving you out of the healing loop for an entire GCD. Leaving your target vulnerable for a 2nd hit before you heal them?

However, on almost all other fights, how do users of IWS feel? I am a massive Intellect stacker with no points in IWS and rely on replenishment. Currently my mana pool hits 30k when i use elixirs and almost 29k when i use flask of resto. I've had no problems spamming HW on Patchwerk in 25man situations, using keyboard movement to interrupt if the heal isn't needed. All other fights, including Sarthion 3drake are easy mode in terms of mana management.

We have two hateful tanks. On Patchwerk the way I've done water shield refreshes is to make sure that both tidal waves is up and that both tanks are fully topped off, so that I should have time to land a heal in addition to my water shield refresh even in a worst case scenario of him getting hit and the other hateful tank receiving a hit with no avoidance for either. I also want riptide to be off cooldown during this time as I can follow the heal with that if necessary assuming nature's swiftness is down.

I have found that improved water shield allows me, on lesser healing wave friendly fights due to distance between players on my assignment, to put some more throughput into my gear. It is definitely possible to reach an abundance of mana with int stacking, but I lose more throughput then if I shift three points into improved water shield and use fewer int gems and a spellpower based flask and food I have found.

With regard to your first question, in 25 man PW I have never found a situation in which I couldn't spare 1 GCD to refresh WS, even undermanning with a relatively low number of healers (Our PW kill last night was 18 man with 4 healers - 1x disc, 1x druid, 2x shaman.). Even with low levels of haste, HW will cast in ~1.7s with Tidal Waves, so I tend to bounce a CH off my tank target whenever I want the GCD to refresh WS.

Fully glyphed and talented (GoWM, GoWS, 2T7, and 3/3 IS), WS gives 140 mp5 and 535 mana per IWS proc (i.e. 2140 mana for 4 orbs). In other words, 4 IWS procs is the equivalent of ~76 seconds of passive WS regen, so IWS is a net gain as long as you can find one GCD to refresh WS within 1.25 minutes of it falling off. Even this is somewhat of a worst case scenario, as it assumes that you were unable to fresh WS when it had between 1 and 3 charges remaining. If, like me, you don't think that GoWM is worth a glyph slot, then 4 IWS procs is equivalent to ~97s of passive WS regen.

I would say that if you have the time and latency to successfully cast-cancel on PW, then you should also have the time to refresh Water Shield approximately 1-2 times during the fight. With 30k mana, however, you should be getting around ~375 mp5 from replenishment alone, so depending on how long it takes your guild to kill PW you might not care about any other sources of regen. Frankly, with that much of a mana buffer I wouldn't even bother with cast-cancel - but that might be my 400ms latency talking.

I use crit heals that top off the tank to refresh my water shield. If the tank gets hit in the intervening GCD, the AA will land on the other side along with (normally) the ES.

For patch, I reactively LHW heal OT #2, which is normally one of our DK tanks. A TW LHW is more than fast enough (with 40 ms ping) to beat out the next hateful and our DKs can take 2 hatefuls with only a LHW in between. It's fairly unusual, but seems to work fine.