MoP beta discussion - Page 6

Elimbras · 05/03/12, 11:32 AM

Originally Posted by Havoc12

Basically the crit modifier of PoH is

{2+0.6*(1+mastery)/[1+0.3*(1+mastery)]}, while for our other spells it is [2+0.6*(1+mastery)]

So the crit modifier for PoH when taking aegis into account is lower and rises more slowly with mastery.

Are you sure of your formula for the crit healing ?
I thought that Aegis is 30%/60% of the actual healing (including overheal), which means that the healing of POH is

normal heal : base_heal * [1 + 0.3 * (1 + mastery) ], which is what you state
critical heal : base_heal * 2 * [ 1 + 0.6 (1 + mastery)], which gives a coefficient for crit of {2 + 1.2 * (1+ mastery)}

Havoc12 · 05/03/12, 4:04 PM

Hmmm actually you are quite right Elimbras.

Originally Posted by Havoc12

I am calculating total healing with PoH including crit and aegis by determining the crit boost with 60% aegis and then adding 30% aegis for all non crits:
$base*[1+(1+1.2*(1+mastery))*crit] + 0.3*base*(1-crit)*(1+mastery)] =$

This is the formula I posted for PoH and as you can see I am using 1.2*(1+mastery) as the modifier. All my analysis has been done using the correct formula. The 0.6 is a typo that is not present in the spreadsheet I am working on, so it does not affect any of the actual results I posted. I verified every result by looking at the values I get in game so I am certain the results I posted are correct.

The reason why PoH behaves this way is because the 30% on crit part of aegis is only applied to the healing part of PoH. So for a normal heal the whole heal more than doubles on the crit. While for PoH 70% of the heal has the same modifier as a direct heal, while 30% of it only doubles.

Perkeyone · 05/05/12, 3:01 PM

As far as i can tell crit is, and always has been, the best throughput rating stat for all disc healing spells excluding atonement, PW:S, and possibly renew (atonement on live scales best with haste, but on beta... crit, now that crits deal double damage, up from 1.5. I have not had a chance to experiment with spirit shell but I doubt it can crit, meaning it will scale best with mastery like PW:S)

I just checked on beta,
-My non-crit PoH: 14038 (5428 da)
-My crit PoH: 28764 (21506 da)
It still double dips in DA, which is the main reason crit is so strong. I have no clue why you folks seem to be railing on crit.

At lv 85 it takes 128 haste rating for a 1% increase in haste, which yields 1% more spell casts and therefore 1% more hps on all cast time spells. (it also affects renew to some extent, but I dont renew enough to consider it)
1%/128=.00781% increase per point.

At lv 85 it takes 176 crit rating for a 1% increase in crit chance, which yields 1% more crits and 1% less non-crits. Crits are also affected by the "burning" meta gem.
For a spell like flash heal (and greater heal before it was taken away, and beta-atonement)
At base mastery(8): 1%*(2*1.03*(1+.3*(1+.2))-1)/176=.01024% increase per point
With +2 mastery(10): 1%*(2*1.03*(1+.3*(1+.25))-1)/176=.01041% increase per point
For PoH
At base mastery(8): 1%*(2*1.03*(1+.6*(1+.2))-(1+.3*(1+.2)))/176=.0124% increase per point
With +2 mastery(10): 1%*(2*1.03*(1+.3*(1+.25))-(1+.3*(1+.25)))/176=.01267% increase per point

Not only is crit the best hps stat for all spells excluding PW:S, but also the best hpm stat. (and on live it procs inspiration)

EDIT (To clarify, these % increases are relative to 0% haste and 0% crit, so they both suffer a "diminishing returns" so to speak)

Szeretlek · 05/06/12, 4:26 AM

As far as i can tell crit is, and always has been, the best throughput rating stat for all disc healing spells excluding atonement

Three words: Haste Always Better.

Elimbras · 05/06/12, 10:35 AM

Originally Posted by Perkeyone

As far as i can tell crit is, and always has been, the best throughput rating stat for all disc healing spells excluding atonement, PW:S, and possibly renew

It was clearly not the case till the patch 4.3 (or 4.2) where crit heals were made from 150% to 200% healing. Up to that point, crit was not as garbage for disc as for other healers, but it was not a really interesting stat either.

And additionnally, I'm a theorycrafter, but we have to take into account that crit offers less control, less reactivity, and usually more overheal than other stats (especially haste). Since we are not healing dummies but real characters with a finite health, being reactive before they die, healing through a bad rnd no-crit sequence is more important than healing slowly big numbers that partly go into overheal (I'm volunteerly going extreme in that picture, but truth is that what I say is partly true, and that correctly identifying and quantifying these effects is really difficult. To be honest, one can also say that aegis and POH, like PW:S, allows to pre-heal (steal healing would say other healers ;-) ), which increase "efficiency", but my rough feeling is that this effect is less important than the previous one I exposed).

Now, that being said, I want to precise that I don't say that crit is garbage, or whatever. But that you should take theoretical models without overheal nor death with a grain of caution, and that numbers do not always tell you the whole picture.

Havoc12 · 05/06/12, 12:48 PM

Originally Posted by Perkeyone

As far as i can tell crit is, and always has been, the best throughput rating stat for all disc healing spells excluding atonement, PW:S, and possibly renew (atonement on live scales best with haste, but on beta... crit, now that crits deal double damage, up from 1.5. I have not had a chance to experiment with spirit shell but I doubt it can crit, meaning it will scale best with mastery like PW:S)

I just checked on beta,
-My non-crit PoH: 14038 (5428 da)
-My crit PoH: 28764 (21506 da)
It still double dips in DA, which is the main reason crit is so strong. I have no clue why you folks seem to be railing on crit.

It looks like you didn't bother to read the whole post so you don't understand what is going on at all. We are not talking about the value of crit as a throughput stat. We were discussing whether PoH really double dips with crit. The answer is no, PoH doesn't double dip. It benefits from crit less than normal heals read the rest of the thread properly if you want to know why. The difference becomes greater and greater with more mastery since the crit multiplier for PoH increases very slowly with mastery compared with direct heals. At 20% mastery its ~2.3 for PoH and 2.4 for direct heals. At higher levels of mastery things get worse. At the high mastery limit you are looking at nearly 3 for direct spells but PoH pretty much caps at like 2.7

At lv 85 it takes 128 haste rating for a 1% increase in haste, which yields 1% more spell casts and therefore 1% more hps on all cast time spells. (it also affects renew to some extent, but I dont renew enough to consider it)
1%/128=.00781% increase per point.

At lv 85 it takes 176 crit rating for a 1% increase in crit chance, which yields 1% more crits and 1% less non-crits. Crits are also affected by the "burning" meta gem.
For a spell like flash heal (and greater heal before it was taken away, and beta-atonement)
At base mastery(8): 1%*(2*1.03*(1+.3*(1+.2))-1)/176=.01024% increase per point
With +2 mastery(10): 1%*(2*1.03*(1+.3*(1+.25))-1)/176=.01041% increase per point
For PoH
At base mastery(8): 1%*(2*1.03*(1+.6*(1+.2))-(1+.3*(1+.2)))/176=.0124% increase per point
With +2 mastery(10): 1%*(2*1.03*(1+.3*(1+.25))-(1+.3*(1+.25)))/176=.01267% increase per point

Not only is crit the best hps stat for all spells excluding PW:S, but also the best hpm stat. (and on live it procs inspiration)

EDIT (To clarify, these % increases are relative to 0% haste and 0% crit, so they both suffer a "diminishing returns" so to speak)

Your calculations dont look correct to me. Aside from that. Assuming 0% haste and base mastery is not a good idea. Mastery especially. We already calculated the partial derivatives. If you want to do the calculations on the stat comparisons please use them to do the calculation properly. Remember that we now have a mastery buff so baseline mastery is no longer 20%.

128/1% for haste and 176/1% for crit looks pretty discouraging for crit. We need to calculate ofc to be sure. The reason why it looks discouraging is because of overheal and uneven healing. Lets consider a particular example. If you get one crit on 5 targets requiring two PoHs to heal to full, you will need another PoH to heal the other 4 targets anyway, so the crit on that one target was almost completely wasted except for Aegis. Most of the value of crit right now on live is just the extra aegis stacking. Haste and mastery universally produce better results so far in both firelands and DS.

As I said in my previous posts however we can't really say whether crit will be good or bad until we see what the level 90 values are, particularly for mana regen and more importantly the mechanics of the encounters. However what we can say right now is that crit is better for single target heals that it is for PoH.

Edit: I see now where you are going wrong, its not the fomrula itself as I assumed earlier:

What you are calculating is the absolute increase. You are subtracting the healing of a non crit from the healing done by a crit. This is NOT A %INCREASE. It is an absolute increase. To calculate the % increase you need to divide by the amount of healing done.

Perkeyone · 05/06/12, 6:01 PM

Originally Posted by Szeretlek

Three words: Haste Always Better.

thanks for supporting that argument. go heal with your shiny buttons and pretty colored spells.

Originally Posted by Elimbras

It was clearly not the case till the patch 4.3 (or 4.2)
...
And additionnally, I'm a theorycrafter, but we have to take into account that crit offers less control, less reactivity, and usually more overheal than other stats (especially haste).

AH, yeah sorry I forgot that was a recent change, it feels like its been longer.
As far as the RNG issue goes, with my current gear, a flask and a raid crit buff, I sit at about 30% chance to crit and then I can inner focus for an extra 25%, so I feel pretty comfortable with my odds. Ive nearly reached the point where my healing done from crits will eclipse the healing done from non-crits. For reactivity, my philosophy is... if you are in a situation where you need your greater heal to be cast more quickly, then you probably should have flash healed instead, and you are probably not taking best advantage of the core feature of disc priests... damage prevention. As far as overheal is concerned, typically I dont use direct heals on low damage phases, (I smite spam during those times), and for times when smite spam is not sufficient, there is much less chance to overheal =]. Either way, the mana cost is the same regardless of a crit, unlike with haste where the more you stack the more mana you use.

Originally Posted by Havoc12

It looks like you didn't bother to read the whole post so you don't understand what is going on at all. We are not talking about the value of crit as a throughput stat. We were discussing whether PoH really double dips with crit. The answer is no, PoH doesn't double dip. It benefits from crit less than normal heals read the rest of the thread properly if you want to know why. The difference becomes greater and greater with more mastery since the crit multiplier for PoH increases very slowly with mastery compared with direct heals. At 20% mastery its ~2.3 for PoH and 2.4 for direct heals. At higher levels of mastery things get worse. At the high mastery limit you are looking at nearly 3 for direct spells but PoH pretty much caps at like 2.7

Your calculations dont look correct to me. Aside from that. Assuming 0% haste and base mastery is not a good idea. Mastery especially. We already calculated the partial derivatives. If you want to do the calculations on the stat comparisons please use them to do the calculation properly. Remember that we now have a mastery buff so baseline mastery is no longer 20%.

128/1% for haste and 176/1% for crit looks pretty discouraging for crit. We need to calculate ofc to be sure. The reason why it looks discouraging is because of overheal and uneven healing. Lets consider a particular example. If you get one crit on 5 targets requiring two PoHs to heal to full, you will need another PoH to heal the other 4 targets anyway, so the crit on that one target was almost completely wasted except for Aegis. Most of the value of crit right now on live is just the extra aegis stacking. Haste and mastery universally produce better results so far in both firelands and DS.

As I said in my previous posts however we can't really say whether crit will be good or bad until we see what the level 90 values are, particularly for mana regen and more importantly the mechanics of the encounters. However what we can say right now is that crit is better for single target heals that it is for PoH.

Edit: Yep here is where you are going wrong:

The correct crit amount for a normal heal is [2+0.6*(1+mastery)] and the correct crit amount for PoH is [2+1.2*(1+mastery)]. You are massively inflating the aegis bonus by multiplying the aegis bonus + 1 with 1+mastery.

1st paragraph
What do you mean "really double dips in crit"? If you said "double dips in DA" then Id understand, which it does.
I see now that you are talking about the ratio of healing from a crit PoH to a non-crit being less than a single target heal's ratio... but I dont see why that is a particularly pertinent issue. Crit still yeilds the better hps and hpm than either of the other 2 ratings, at any level of mastery. And the master buff is a non-issue, itwill only provide even more benefit to crit, I demonstrated already that increasing mastery makes each point of crit more valuable when I compared 8 mastery with 10.

2nd para
In your example case, Id likely only use 1 PoH, Id try to inner focus for it, and then Id smite or PoM to top everyone off until the next round of damage. Unless there's an ability that randomly 1 shots anyone who isnt full hp, you dont need to top people off, thats what you use smart heals for. Mastery is pretty weak, its main strength is that it makes crit stronger.

3rd para
If all you meant to say is that the difference between a crit and a non crit greater heal is more significant than that of poh, I can agree, but I dont follow you on why that is an issue. Im sorry that I misunderstood your post.

4th para
We are using the same formula, you just distributed the 2
[2+0.6*(1+mastery)] == 2*[1+0.3*(1+mastery)]
[2+1.2*(1+mastery)] == 2*[1+0.6*(1+mastery)]

Havoc12 · 05/06/12, 7:00 PM

Originally Posted by Perkeyone

1st paragraph
What do you mean "really double dips in crit"? If you said "double dips in DA" then Id understand, which it does.
I see now that you are talking about the ratio of healing from a crit PoH to a non-crit being less than a single target heal's ratio... but I dont see why that is a particularly pertinent issue. Crit still yeilds the better hps and hpm than either of the other 2 ratings, at any level of mastery. And the master buff is a non-issue, itwill only provide even more benefit to crit, I demonstrated already that increasing mastery makes each point of crit more valuable when I compared 8 mastery with 10.

3rd para
If all you meant to say is that the difference between a crit and a non crit greater heal is more significant than that of poh, I can agree, but I dont follow you on why that is an issue. Im sorry that I misunderstood your post.

You should think carefully before answering. If the crit multiplier is lower, how can you possibly end up with a higher increase for PoH. It should immediatley ring alarm bells in your mind that you have made an error somewhere.

2nd para
In your example case, Id likely only use 1 PoH, Id try to inner focus for it, and then Id smite or PoM to top everyone off until the next round of damage. Unless there's an ability that randomly 1 shots anyone who isnt full hp, you dont need to top people off, thats what you use smart heals for. Mastery is pretty weak, its main strength is that it makes crit stronger.

Then you would be commiting a huge mistake costing you a massive amount of HPS and HPM. PoH on 4 targets still heals for over 80k for just 8k mana in 2.5 seconds. You would spend 2-3 times that mana to heal ppl up with smite/holy fire and it would take you ages to do so. More importantly it does not change the fact that your PoH heal crit will be wasted as the smite/holy fire that you would heal the person that received the crit would overheal. Actually it is worse because all hits and crits on your smites and holy fires would overheal, which means the value of the crit you have is zero for the the entire period where you are smiting. In fact it could be a lot worse, because you timer on the aegis left by the first PoH is ticking down and the second PoH would refresh it. On live right now, its wise to keep spamming PoH even if eveyrone is full health. The reason is you are stacking aegis and when that big hit comes it all gets absorbed giving you big chunks of healing. If you look at the top ranks overheal for PoH is extremely high because of that.

4th para
We are using the same formula, you just distributed the 2
[2+0.6*(1+mastery)] == 2*[1+0.3*(1+mastery)]
[2+1.2*(1+mastery)] == 2*[1+0.6*(1+mastery)]

I am afraid that your calculations are completely wrong. I looked more closely and I see exactly what went wrong. I thought the 1.03 from the metagem was an aegis bonus before, but I was wrong. If you remove the 1.03 you are indeed using the correct formula for a crit, but you are not calculating what you say you are. Here is what you are doing. You are subtracting the healing of a non crit from a crit and then multiplying by 1%. Then you are implicitly dividing by the base value of the spell. Thus you are making the exact same mistake that was commited before. You are normalising your results based on the base value of PoH WITHOUT taking aegis into account. This renders all the results that you have posted for PoH competely irrelevant and inaccurate.

What you should have done for PoH is

1%*(2*1.03*(1+.6*(1+.2))/(1+.3*(1+.2)) - 1)/176 = 0.009121% per stat point.

For 25% mastery it is 0.009215% per stat point.

Even that however is not really good, because you are ignoring the fact that both crit and mastery always add a fixed amount and thus have diminishing returns, while haste doesnt.

I am afraid when you do things properly you will find that mastery increases the value of crit for PoH only slightly, but it does have an effect on our other direct heals.

#Here is the correct way to calculate the relative values.

$Healing_{\rm PoH} = base*[1+crit + (0.9*crit+0.3)*(1+mastery)]$

$Healing_{\rm Direct} = base*[1+crit+0.6*crit*(1+mastery)]$

$\frac{\partial Healing_{\rm PoH}}{\partial crit} = base*[1+0.9*(1+mastery)]$

$\frac{\partial Healing_{\rm Direct}}{\partial crit} = base*[1+0.6*(1+mastery)]$

$\frac{\partial Healing_{\rm PoH}}{\partial mastery} = base*(0.9*crit+0.3)$

$\frac{\partial Healing_{\rm Direct}}{\partial mastery} = base*0.6*crit$

All the derivatives are nice and constant, which makes things easy.

The amount of healing added by x% crit or y% mastery is simply:

$\frac{\partial Healing}{\partial crit}*0.01*x$

$\frac{\partial Healing}{\partial mastery}*0.01*y$

Let us define:

CritValue = increase in healing by 1% crit
MastValue = increase in healing by 1% mastery

$CritValue_{\rm PoH} = 0.01*\frac{\partial Healing_{\rm PoH}}{\partial crit}*\frac{1}{ Healing_{\rm PoH}}$

$CritValue_{\rm PoH}= 0.01*\frac{1+0.9*(1+mastery)}{1+crit + (0.9*crit+0.3)*(1+mastery)}$

and

$MastValue_{\rm PoH} = 0.01*\frac{0.9*crit+0.3}{1+crit + (0.9*crit+0.3)*(1+mastery)}$

Similarly:

$CritValue_{\rm direct} = 0.01*\frac{1+0.6*(1+mastery)}{1+crit+0.6*(1+mastery)*crit)}$

and

$MastValue_{\rm direct} = 0.01*\frac{0.6*crit}{1+crit+0.6*(1+mastery)*crit)}$

With this the real values for additional crit and mastery can be calculated. Using the very easily attainable values of 10% crit and 38% mastery

[latex] CritValue_{\rm PoH} = 0.013685753

[latex] MastValue_{\rm PoH} = 0.002380662

[latex] CritValue_{\rm direct} = 0.015454853

[latex] MastValue_{\rm direct} 0.000507271

So 1% crit increases healing by 1.37% for PoH and 1.55% for Direct heals

The stat budget value ratio for crit relative to haste is 176/128 = 1.375.

It is immediately obvious that if we ignore overheal haste and crit have exactly the same throughput value for PoH. For direct heals crit is significantly stronger. When we factor in overheal however haste should be a much more valuable throughput stat.

Mastery on the other hand appears to have a pretty poor value for direct heals compared to PoH. 1% mastery is equal to 0.24% increase for PoH and just 0.05% increase for direct spells. For direct heals it seems that haste>crit>mastery, unless the stat budget for mastery is extremely low and overheal is very high. For PoH haste is going to be better than both crit and mastery, but its not easy to determine the relative values of mastery and crit better unless we find the stat budget for mastery and some idea of overheal rates and damage patterns. This is something that we need to return to, once level 90 dungeons and raids are available.

Addendum: I dont know if its a bug but currently haste adds multiplicatively. I.e if you have say 5% haste then your total haste on borrowed time is 1.15*1.05 and not 20% as you would normally expect. This makes haste even more valuable since it stacks multiplicatively with other haste buffs.

Perkeyone · 05/06/12, 9:11 PM

Ah I do see that I made some error. Your point about the ratios being an alarm is right.
I made a script in python from scratch to try to work out my error and I think I was successful.
Do you see any obvious error in this?

#constants

intel=7500						# intellect
sp=10000						# spell power

fh_v=0.7248*sp+(6272+7289)/2	# flash heal value
fh_bct=1.5						# flash heal base cast time

poh_v=0.340*sp+(3087+3262)/2	# poh value
poh_bct=2.5						# poh base cast time

meta=1.03						# burning meta gem bonus

# first run

cr=1000							# crit rating
hr=1000							# haste rating
mr=1000							# mastery rating

cc=(cr/179.28+intel/648.91)/100	# crit chance
cs=1+(hr/128.05701)/100			# casting speed
mb=mr/179.28*.025+.2				# mastery bonus

fh_hps_n=(cc*2*meta*fh_v*(1+.3*(1+mb))+(1-cc)*fh_v)*cs/fh_bct
poh_hps_n=(cc*2*meta*poh_v*(1+.6*(1+mb))+(1-cc)*poh_v*(1+.3*(1+mb)))*cs/poh_bct

# second run (crit)

cr=1001							# crit rating
hr=1000							# haste rating
mr=1000							# mastery rating

cc=(cr/179.28+intel/648.91)/100	# crit chance
cs=1+(hr/128.05701)/100			# casting speed
mb=mr/179.28*.025+.2				# mastery bonus

fh_hps_c=(cc*2*meta*fh_v*(1+.3*(1+mb))+(1-cc)*fh_v)*cs/fh_bct
poh_hps_c=(cc*2*meta*poh_v*(1+.6*(1+mb))+(1-cc)*poh_v*(1+.3*(1+mb)))*cs/poh_bct

# third run (haste)

cr=1000							# crit rating
hr=1001							# haste rating
mr=1000							# mastery rating

cc=(cr/179.28+intel/648.91)/100	# crit chance
cs=1+(hr/128.05701)/100			# casting speed
mb=mr/179.28*.025+.2				# mastery bonus

fh_hps_h=(cc*2*meta*fh_v*(1+.3*(1+mb))+(1-cc)*fh_v)*cs/fh_bct
poh_hps_h=(cc*2*meta*poh_v*(1+.6*(1+mb))+(1-cc)*poh_v*(1+.3*(1+mb)))*cs/poh_bct

# fouth run (mastery)

cr=1000							# crit rating
hr=1000							# haste rating
mr=1001							# mastery rating

cc=(cr/179.28+intel/648.91)/100	# crit chance
cs=1+(hr/128.05701)/100			# casting speed
mb=mr/179.28*.025+.2				# mastery bonus

fh_hps_m=(cc*2*meta*fh_v*(1+.3*(1+mb))+(1-cc)*fh_v)*cs/fh_bct
poh_hps_m=(cc*2*meta*poh_v*(1+.6*(1+mb))+(1-cc)*poh_v*(1+.3*(1+mb)))*cs/poh_bct

# results
print "flash heal hps:    "+str(fh_hps_n)
print "+1 crit rating:    "+str(fh_hps_c)+"("+str((fh_hps_c/fh_hps_n-1)*100)+"% increase)"
print "+1 haste rating:   "+str(fh_hps_h)+"("+str((fh_hps_h/fh_hps_n-1)*100)+"% increase)"
print "+1 mastery rating: "+str(fh_hps_m)+"("+str((fh_hps_m/fh_hps_n-1)*100)+"% increase)"

print "poh hps 1 target:  "+str(poh_hps_n)
print "+1 crit rating:    "+str(poh_hps_c)+"("+str((poh_hps_c/poh_hps_n-1)*100)+"% increase)"
print "+1 haste rating:   "+str(poh_hps_h)+"("+str((poh_hps_h/poh_hps_n-1)*100)+"% increase)"
print "+1 mastery rating: "+str(poh_hps_m)+"("+str((poh_hps_m/poh_hps_n-1)*100)+"% increase)"

Havoc12 · 05/07/12, 7:57 AM

Don't see any obvious errors in this. What is the stat budget for mastery?

Havoc12 · 05/07/12, 10:55 AM

GC said we are getting no CD on PWS, greater heal and they are reworking spirit shell next patch.

Perkeyone · 05/07/12, 3:02 PM

I appreciate you taking the time to explain all this to me, Havoc. It is frustrating for me and probably for you too to spend so much time calculate all of this only to come to the exact conclusion that was just stated.

Ya know, I jokingly tell guildmates and friends that in order to do more dps/healing/etc all you have to do is hit buttons faster. Im a bit disappointed that its not far from the truth.

with 1000 points of crit, haste, and mastery

flash heal hps:    13343.7641385
poh hps 1 target:   5098.10249904

+1 crit rating:    13344.8257819 (0.00795610186339% increase)
+1 crit rating:     5098.46836681(0.00717654782574% increase)

+1 haste rating:   13344.7306786 (0.00724338445399% increase)
+1 haste rating:    5098.47177421(0.00724338445399% increase)

+1 mastery rating: 13343.913026  (0.00111578365636% increase)
+1 mastery rating:  5098.28450226(0.00357001889351% increase)

Havoc12 · 05/08/12, 6:19 PM

Discipline
- Grace now affects Greater Heal and Heal, no longer affects Spirit Shell.
- Rapture now also removes the cooldown on Power Word: Shield.
- Spirit Shell has been reworked - For the next 15 sec, your Heal, Flash Heal, Greater Heal, and Prayer of Healing no longer heal but instead create absorption shields that last 15 sec. Instant. 60 sec cooldown.
- Strength of Soul (New) - When you heal a target with your Heal, Greater Heal or Flash Heal spell, the duration of the Weakened Soul debuff on the target is reduced by 2 sec.

Very interesting changes. Spirit shell in particular is just want discipline needed a new way to build large shields. Now the real question does the new spirit shell stack. If it does in a similar way to aegis then this is a very exciting cooldown.

Vintoran · 05/09/12, 4:32 AM

As of this moment SS works like this on beta:
Instant (no gcd), no mana cost, 60 sec cd, 15 sec buff;
If you cast a PoH, FH, Heal or GH during those 15 seconds you create a bubble with the name of the spell that triggered it.
Different spells stacks, the same spells don't stack. That means if you cast the sequence SS, FH, FH you create one absorb shield called "Flash Heal" for the amount of only one FH. If you cast SS, FH, GH you create one absorb shield called "Flash Heal" and one called "Greater Heal" and so forth.
Spells cast under SS renew, apply and benefit from Grace.
The check if SS is active appears to be made at the end of the cast, so if you have 1 second left on the SS buff and cast a GH that heals after the SS buff runs out, you don't get an absorb shield, only a normal heal.

One possible problem is that this could lead to overpowered Disc priests in certain situations. Imagine a fight like Freya HM where you had a hard hitting raid aoe (something like 80% hp of nontanks) on top of "normal" raid damage. You could have enough (2 for 30 second cd aoe for example) Disc priests to have a SS up before every hit and severely reduce chance of oneshots, something that only Resto shaman can do as well (and only if grouped). In 15 seconds you could easily cast 6 PoHs, which is something like 10% extra hp for everyone in the raid (the 6th for a group that has taken damage in the meantime), exactly when you need it.
The other situation where this could be overpowered is predictable tank bursts like Madness. You could prestack a FH, GH, Heal combo and even PoH with aegis if you really want to for something like 140k plus PWS for 40k absorb.

That would leave Disc priests with 1 tank cd, 1 raid cd and 1 that can be used for either, which is more than anyone else has. At least in many Cata hardmodes it used to be: the guy with the least cds sits out on progress.

I also noticed that Divine Insight is probably bugged for Disc: if you cast GH, you have a chance (probably not small, had a couple of them - most likely 40%) of getting the no cd, self jumping PoM. Conversely I casted 10 Penance casts and didn't get a single DI proc from that (odds of that happening with 40% chance = 0.6^10 ~= 0.6%).

edit: It seems I overlooked that the heals during SS can't crit and PoH doesn't proc any DA, which reduces the usefulness severly. Singletarget it's most likely an hps loss (unless you have >~65% overheal) and for aoe it's more or less the same hps. The use as a third cooldown is still possible, even if it won't work in some fights (with continuous or unpredictable damage).

Szeretlek · 05/09/12, 5:04 AM

Its not so cool after I tested it.
When you trigger SS, aegis portion of all spells are gone, that means your PoH without SS will heal for 13k and make aegis for 6k, it could crit and make 26k heal and 24k aegis. With SS PoH will make only 13k bubble that cant stack and cant crit. It is not viable to turn SS on when raid take damage, because you lose Aegis and Crit totally.
Lets assume you cant overheal, your aegis do 50% of heal to bubble and you have 30% crit rating
Direct Heal WITHOUT SS does X number of heal per second with chance 70% and 2*(X+0.5*X) with 30% chance. Totally it will be 1.6*X Heal per second.
Direct Heal WITH SS will do only X Heal per second.

PoH WITHOUT SS does Y+Y*0.5 hps with 70% chance and 2*(Y+2*Y*0.5) with 30% chance. Totally it will be 2.25*Y hps.
PoH WITH SS do only Y hps.

So SS drops your hps for direct heals by 38%. and for PoH it lowers your hps by 55%.

=(

Another solution - prestack bubbles.
PoH Without SS stack 0.5*Y aegis per second with 70% chance and 2*2*Y*0.5 with 30% chance. Totally you will stack aegis with 0.95*Y hps.
PoH with SS stack Y hps

Pure PoH stack aegis at even rate compared to PoH+SS. But Pure PoH CAN stack and do some heal and SS cant stack

Conclusion: Spirit Shell is garbage.