I'm not experienced enough to play around with numbers, but do not forget that mastery will start affecting Earth Shield too in 4.1, so we will not be seeing a straight 20% drop in the healing done by Earth Shield. Most likely the reduced base heal is to cover mastery affecting ES at lower healths. (Source)

Again -- the conflict shouldn't be between Earth Shield and Riptide, or Water Shield and Earth Shield. It should be between Water Shield and Earthliving Weapon. All those calculations you're providing us with completely miss the point.

Looking at your log linked above, your Earthliving Weapon (Viveterre in French) overhealed for 55.9% over the course of the night.

61.2% on Halfus.
52.7% on Magmaw.
67.9% on Omnitron.
61.7% on Maloriak.
3.7% on Chimaeron. (Mostly useless healing anyway.)
57.0% on Atramedes.
54.4% on Nefarian.

If we take out Chimaeron as an anomaly, your overhealing on the remaining fights for ELW was 59.15%. Given your healing style and healing assignments, it appears an extra 20% to Earthliving Weapon is not benefiting you very much.

I did some testing on the strange results of Healing Stream Totem. The following samples are taken in elemental spec, swapping gear around and using Earthliving Weapon, [Flask of the Draconic Mind], Power Torrent and [Volcanic Potion] to temporarily boost my spell power. All observed values are the lower of the two, essentially rounding down.

Observed	Tooltip	Spell power
43	39	129
270	265	2870
305	300	3295
336	332	3675
358	354	3938
379	375	4193
418	413	4659
440	435	4927
468	463	5264
614	610	7034
617	550	7058*+
642	638	7371
678	612	7800*
704	638	8115*
731	665	8452*
801	729	9297*
849	777	9875*
916	844	10683*
964	892	11261*

All *-marked spell power values are with Earthliving Waepon. The tooltip matches the wowhead fomula:

28 + 0.0827 * $SP

but the observed values are off. Substracting 744 for 2/2 talented Earthliving Weapon and substracting the expected tooltip, the average rest is 4.51. Keeping in mind that all observed values are rouned down, I tried an offset of 5, which fits for all but one sample (+), which is off by one and might be an error on my side.

So the formula for the coefficient table should look like:

Healing Stream Totem expected:
(33 + 0.0827 * $healing_power) * (1 + 0.25 * $purification) * (1 + 0.25 * $soothing_rains) * (1 + 1.02 * $spark_of_life) * (1 + 1.05 * $target_spark_of_life)

Except that does produce wrong results. I tested with 7017 spell power (no Earthliving) and the predicted value would be 1401.79, the observed is 1233. That is exactly 1.1/1.25 of the expected value, which means: There was no Purification hotfix for Healing Stream Totem.

The tooltip value is even more off. I thought the tooltip should list 1.1 for Purification (old), 1.15 for Spark of Life (bugged) and 1.5 for Soothing Rains. The tooltip was so far off, that changing any one of these would not get to the right value. I had to respec and check every talent to get to the right formula:

Healing Stream Totem real:
(33 + 0.0827 * $healing_power) * (1 + 0.1 * $purification) * (1 + 0.25 * $soothing_rains) * (1 + 1.02 * $spark_of_life) * (1 + 1.05 * $target_spark_of_life)

Healing Stream Totem tooltip:
28 + (0.0827 * $spell_power) * (1 + 0.1 * $purification) * (1 + 0.25 * $soothing_rains)

Differences:

• tooltip scales with spell power, spell with healing power
• tooltip forgets to apply modifiers to the base heal, which is incorrect anyway
• tooltip doesn't care about spark of life
• both tooltip and spell work with 10% Purification

Edit: If you know what to look for, then it seems others have made that observation before me. Apologies if that was already known.

Are you saying if someone in your raid had a 5k health deficit, you or someone else in your raid would actively try to heal them (and by active I mean using non-smart or aoe zone heals)? Spell discipline is a significant part of your skill as a healer this expansion, and saying that the earthliving glyph provides little or no efficiency gains because you can't stand when someone isn't topped off means you need to readjust your priorities as a healer.

Obviously there are plenty of examples where earthliving overheals or contributes to raid overheal, but that doesn't make it nearly as bad as you make it seem, especially when it's usually the #3 highest source of throughput for resto shamans in 25 mans.

I made a ranked list of trinkets. It includes blue trinkets from heroic dungeons, as well as trinkets with damage procs that really should go to dps before a healer. Trinkets that boost mana tide for other healers should be valued higher for healers raiding 25-player dungeons.

https://spreadsheets.google.com/ccc?...thkey=CI-0-uoP

I use slightly different stat weights than the ones in the guide at the beginning of this thread. I don't at all mean to imply that those are wrong and mine are right. Rather, this reinforces Aanzeijar's point. Stat weights for a healer should only be taken for what they are, a place to get you started. They'll be different for you depending on your personal gear, talents, and healing style.

With that said, the precise values and ranking are only a rough guide. Trinkets close together in value are just that -- close together in value. But this list lets me make broad judgment statements like "Tyrande's Doll is amazing," and "Heart of Ignacious is terrible."

edit:

Fixed the suggestions made so far.

Static spirit effects including reforging have been adjusted to account for increased personal benefit from mana tide. This still doesn't account for the benefit to 2 other healers in a 10-player raid, or 6 others in a 25-player raid. These benefits are impossible to quantify without some sort of judgment about the value of spirit to the healers in your raid.

Value for intellect increased to better account for BoK, mail specialization, and regen from increased mana pool, replenishment, and spirit regen.

Value for mp5 decreased to reflect a more accurate spirit conversion with current gear.

edit2:

Further adjustments to mp5 and intellect HEP based on Casey's arithmetic. Only very minor changes to ranking resulted from this change. I still am not providing a numeric value for mana tide regen for other healers. I can't see how that would be quantifiable, and I'm not sure it would even be useful. If your other healers need more mana to do their jobs, that's something to consider subjectively against personal benefits.

edit3:

I separated spirit into static and stacking stats that can affect mana tide, and procs that can't. Hopefully this change didn't introduce new errors.

edit4:

Added Tendrils of Burrowing Dark.

Very nice! A couple notes:

1. For trinkets which affect mana tide, the weight of spirit should be increased by 36% (9% uptime * 400% increase in personal spirit) (if column D is 'Y', then multiply the term from the spirit column by 1.36).
2. It seems like either int is valued too low or spell power and crit are valued too high; converting Intellect with Gift/Kings into an equivalent spell power and crit rating and adding them would give 1.22 total stat weight from that point of intellect instead of the 1.2 weight which intellect actually has in that list [see CaseyTheRetard's post below for a better estimate of the value of int], and that ignores the effect of replenishment, mana pool, and the effect of intellect on spirit regeneration
3. [Fall of Mortality] is no longer the same 1926 proc as its regular version. It's now 2178 spirit, which averages out to 363 spirit instead of 321.
4. [Jar of Ancient Remedies] and its heroic version are significantly under-ranked (since the proc isn't reflected), but I'm not sure what the best way to represent them is (an additional mp5 from the on-use based on the lost spirit regeneration from 515 spirit from some reasonable intellect pool?)

Heroic Jar should be more like 425 spirit (3/4 of 580 due to uptime) + 302.5 Mp5 (7260 Mana every 2min on average) which totals to 927 "stat points" and brings the trinket to the third place on your list while still completely ignoring the Mana Tide buff it provides to your raid (580spirit *4=2320 spirit for 16.8sec every 180sec which means an additional 216.5spirit for every healer in your raid).

You have a value on your sheet for personal spirit. That presence would indicate that (a) you saw how it was quantifiable, and (b) you thought that quantifying it was useful. If every other healer in your raid did the same, you would have a reasonable estimate for the value of MTT spirit to your raid. It's not a simple problem, but it's far from being intractable. Certainly we need not throw our hands in the air by claiming the issue is "subjective."

At the very least, it's useful in a tool of this nature to have a knob to turn to see how varying the raid's value for spirit affects the relative ordering of the trinkets. While one may not agree with my approximation of "each other raid healer values spirit the same as I do", which is certainly fair since I'm not sure I agree with it myself, I think you would likely agree with my tool's placement of DMC:T and Mandala above e.g. Tyrande's Favorite Doll for 25-man raids.

The power of MTT in raids is in its ability to provide mana to other healers, certainly that factor is more significant and therefore more deserving of examination than simply quantifying the self-return.

In the guide it recommends the Int Flask and Int Food for consumables. Granted, the guide also recommends some other things that don't necessary work with my play style (and specifically 10 man HC raiding), but I'd like to get some clarification from other shamans on which consumables they are using and why.

Until this point I've been using the spirit food/flask to help with mana problems. After looking at World of Logs parses, I've changed my glyphs up some (dropping ELW for WS because my EL was doing major overhealing - Big thanks to Amilie for opening my eyes up to check on that), so that should help some with my mana problems. However, I'm not sure if I should be looking towards the INT based consumables.

Is there a reason the int food/flask is recommended over the spirit? I've seen some references to int helping regen by providing a bigger pool, but haven't been able to find a straight forward math explanation on the subject. I did some searching in the old thread (and this one) that referenced people using the spirit consumables for MTT, but no real arguments on spirit vs int consumables.

Is there a really glaring reason I've yet to find on why I should be using the Int flask/food over spirit? Granted, maybe I'm looking into something that makes little difference, but I like min/maxing where possible, especially if one method is clearly superior to another.

Well, for one, in terms of pure stat points, Int consumables give 5% more stats than Spirit ones due to Mail Specialization -- at least Draconic Mind does, I haven't checked Severed Sagefish Head. So unless you're raiding in a world where Spirit is worth more than Int, you should always get Int buffs.

But beyond that: you always gear, gem, and enchant for Int over Spirit. Why wouldn't your consumables follow the same philosophy? Why are you using Power Torrent on your weapon instead of Heartsong, if you're vacillating between Int and Spirit consumables? It's not like they follow special rules. They're just a bag o' stats like gems or enchants.

I've got similar results. Why does Unleash Life buff Riptide only by 16%? Why does Focused Insight buff Riptide only by 24%? And why is the combined buff 40%? Earth Shield buffs correctly by 15%.