Prayer of Mending

[Pokemaster]

I don't use, because I am level 64

[Argan]

if you dont use pom in raids, stop playing wow, i put it on CD on a tank and it makes 30% of my heals all the time, its fucking great

[harky]

Simbert, no I don't have my tests purely because it was so simple to show that it wasn't smart. If you want more sources just Google looking for PoM and 'smart heal' and you'll find plenty of people complaining about how PoM not being a smart heal makes it difficult to use. It's pretty common knowledge and has been since PoM was first introduced. And yes, despite your pseudo-statistical math if you're only counting 20 casts that is in fact a very low sample size.

You mentioned it only being a .0000000286% chance? You yourself admitted it could have been even a 20% margin of error on your part, but you calculated that margin of error completely wrong. That does not make it a .055% chance. It makes it almost completely random. With that margin of error the chances could be as high as 1%. Still, that's not the point. The point is you didn't even test properly making the sample size much lower than 20 casts. That is you did not isolate a control. You did not have one person who intentionally had the lowest percentage/absolute health.

I haven't tested it in over a year, but the spell hasn't changed in any way since then. I haven't tested more because it's so easy to show if you isolate an intended target and track bounces, which is not what you did. If you wanted to test it in a dungeon you would need a target who was definitely at the lowest health and then track how often it bounced to them over other targets. The only known quirk to PoM at this point is that it tends to bounce back to its caster at a larger than average rate.

But again; even if it were smart (which it isn't), it would not actually improve the spell because if it accurately picks targets based on health it still can not predict incoming damage so being smart or not does not impact the spells usage in any way.

[Simbert]

I haven't tested it in over a year, but the spell hasn't changed in any way since then. I haven't tested more because it's so easy to show if you isolate an intended target and track bounces, which is not what you did. If you wanted to test it in a dungeon you would need a target who was definitely at the lowest health and then track how often it bounced to them over other targets. The only known quirk to PoM at this point is that it tends to bounce back to its caster at a larger than average rate.

Thats exactly what i did. And no, my math is not off. What i calculated was the probability that the theory "PoM jumps randomly" could still be true when using my data. And that is in fact the number in my above post.

Obviously, my first set of data was not very good, but, if you would look at it, the second is way better, while showing the same results.

If it is that simple to show that is jumps randomly, and not to the person with the lowest health, then do it. In my opinion, i showed pretty convincingly that it does, in fact, jump to the person with the lowest health. I tested it in a dungeons because, as i said, i had no people available who had nothing to do.

But, if you would read my edit, which gives more useful data than the estimates above, i tested only in situation where there was no AE, making incoming damage absolutely predictable. I than lowered my own health below 100% using SW, and kept the rest of the group at 100%. In every single one of those 30 tests, PoM jumped to me. In pretty much all of them, there were at least 1 other person, and 2 pets, in range for it to jump to. If, for some strange reason, there would have been AE which i would not have noticed, it would have no effect if the jumping was indeed absolutely random. If however, it was smart, that would have lead to it jumping to some other person, which it did not do. So, either you say that my PoM always jumps to me, no matter the circumstances (which is obviously wrong, everyone has encountered the PoM NOT jumping to him at some point, since it happens more often than not if you do not intentionally lower your HP, like i did.

So, if you do not think that i am lying to you about my data (I can not anything to you if you think otherwise anyway), the position that PoM jumps randomly is simply one that cannot be held.

For you to prove that PoM not always jumps to the person with the lowest hp in range, you would need only one single event of it not doing so, however, it would be better if this would be in a replicatable situation.

If you, however, choose not to try it for yourself, i must simply assume that you are totally wrong, since my first-handed experience, tells me otherwise. And everyone interested in the matter can see it for themselves anyway. Your "evidence" however seems to be mostly hearsay. I mean, did you even do that google search yourself before advising me to do it? It does, in fact, on the first two pages, with the exception of this very thread, only yield uninteresting results, or results that say it is a smart heal.

[harky]

I replied before you edit, but your edit has an additional flaw in that you used yourself. It's been noted numerous times already that PoM prefers the caster. If you repeated the test with you at full health you'd likely see a very similar, if not identical result. The person at low health can not be the caster for this reason.

[Eletharin]

PoM is always at least good, and usually its awesome.

Worst case: throw it on the tank, it bounces once to a pet or some crap ... its still good. Not great, but hpm and hps are still respectable although not a contender when matched up with other tools in the priests arsenal.

Bounce twice, its starting to get awesome. Good hpm and hps are competitive even just at this level.

Bounce any more ... OMFG its out of control wicked awesome. If it bounces at least 3 times (and its guaranteed to on a lot of fights) its by far the best spells in the priests arsenal.


My complaint with the spell is its somewhat mindless. Its too reminiscent of paladin beacon or shaman chain heal healing where pressing the button is just as important as having any thought about what you are doing. CoH could fall here too, but if you're just mashing CoH you're not getting the most out of it. Knowledge of raid position and specifically targeting people in the center of damage spikes improves its viability a lot. PoM on the other hand ... you can macro target it to a tank, hit it every 6 seconds, and you're optimizing the ability. I would honestly like to see it removed from the game, but it would have to be accompanied by a restructuring of other mindless healing throughput so priests didn't fall behind with respect to other healing classes (without PoM we would stand even less of a chance).

[Simbert]

I replied before you edit, but your edit has an additional flaw in that you used yourself. It's been noted numerous times already that PoM prefers the caster. If you repeated the test with you at full health you'd likely see a very similar, if not identical result. The person at low health can not be the caster for this reason.

Well, if you have no intent in trying it out for yourself, than keep your strange belief that PoM jumps randomly, but more often to the caster. It does not effect me, and i believe that i did enough to try and convince you of the truth. Try it. You will see i am right.

[harky]

Nope. Just got home from work.

Took 3 casts to get a conflicting test from yours. Testing was done using a group of three: My priest a rogue and a DK. Tests were done by having the Rogue pull a mob and get him to deficit health, then having the DK taunt the mob and putting a PoM on him.

1st: Rogue
2nd: Rogue
3rd: Me
4th: Me
5th: Rogue
6th: Rogue
7th: Rogue
8th: Me
9th: Me
10th: Rogue

I was at full health for each cast. First 5 tests I was at melee range. Next 5 I was about 15 yards away.

Repeated the test 5 more times with different parameters. This time I hit the mob with SW while it was on the rogue while making sure the rogue had taken more damage than me (10-15k health vs ~20k), then had the DK taunt off and PoM'd again.

1st: Me
2nd: Me
3rd: Me
4th: Me
5th: Me

Don't take what I'm saying as 'it's totally random', because that isn't at all. However it does not display the behavior of a true smart heal like Hymn. Hymn prioritizes targets based on two criteria: Absolute and deficit health. It will always target the lowest absolute health target unless another target has higher deficit health. So if you have someone with 25k health and someone with 50k health at both at 20k current health it will heal the one with 50k max health, but if the lower max health target had 19k health they would be healed instead. PoM does not do that. If anything PoM has a list that is either 'full health', 'caster', or 'deficit health'. It does in fact bounce to people who have taken damage, but it does not prioritize beyond that. However, it also flags you as an available target if the PoM was cast by you. It also seems to prioritize you if you have taken damage and are the caster. Maybe my PoM is just bugged...

[Talhooo]

I love it so much I still use 2T9 :P

check how much it does on my sindra kill.

http://worldoflogs.com/reports/rt-dc...=11367&e=11848

[Deleted]

One of the best Heal Skills ingame, insane amount of heal for 1 GCD and for a small amount of mana. But it's not accurate in choosing the player who obtains the greatest amount of damage

[kenn9530]

pom is awsome i use it almost every time i can off cd along with CoH and spamming renew

[Rifee]

Awesome with flaws. The problem is, of course, when it spends it's entire cooldown only being used once because some poor sap was lucky enough to get it.

alla bilities have flaws and wekanesse. it's by design:P

[harky]

Grammar aside... that's really not true. Most abilities don't have flaws. Yes, they have pros and cons, but not flaws.

PoM has 4 big flaws:
1. The cooldown encourages it to be used more often than it can support. Using it on cooldown will almost always waste 20% of the spell.
2. It is a proactive cast with a reactive heal. Prayer of Mending is the only heal in game which can not prevent someone from dying directly.
3. To be usable the fight must include consistent AOE damage. This restricts the use of one of only two spells that makes Priests unique (the other being PoH).
4. One Priest's PoM can 'eat' another one. This makes PoM useless in some fights if there is another Priest around.

What are the flaws of something like Renew? You might claim it heals for less than Rejuv, but that's under-tuned, not flawed. You might claim it doesn't heal fast enough in certain situations, but that isn't a flaw, the spell isn't designed for that situation. It would be like calling a knife flawed because it didn't shoot bullets. Or calling a gun flawed because it wasn't also a microwave.

Really, the first flaw up there is the big one though. If the talents to reduce its cooldown did not have other applications those talents would literally be worthless. There's no reason for PoM to have a cooldown lower than 10 seconds. It would be like the Renew glyph reducing Renew's duration, but not increasing its healing. Not that the Renew glyph isn't bad already, but that's about the equivalent reduction of power that using PoM 'on cooldown' implies. Not stacking with other PoMs is pretty bad too. It hearks back to early Vanilla when HoTs didn't stack from multiple sources.

[Zeuq]

harky, while I agree with your points, I'm not really sure "flaw" is the appropriate word to describe all of those. In fact, I think only the fourth point is really a "flaw" in it's design, the others are really just about how situational the spell is.

With point 1, I actually think the cooldown is one of the small ways that differentiates intelligent use compared with mindless spamming. I've found that in the vast majority of cases with persistent AOE, refreshing PoM on CD isn't necessary, and I can delay it by a GCD or two; for someone who does refresh it immediately on CD, they're wasting potential throughput. I can only recall one encounter where I found I could often recast on CD without wasting charges, and that's Twin Val'kyrs. So, I agree they could safely leave the CD longer and that the talent to reduce it, if it weren't part of another necessary talent, would be largely useless; however, it would just provide less difference between using it skillfully and just watching the timers and using it immediately when it's up. It also does reduce the penalty when it bounces in a stupid way, so we can get it bouncing how we'd like it to instead.

For instance, with point 2, I don't think a flaw as simply part of what prevents the ability from being used as an instant heal to save someone's life in the way SoL or Empowered Renew can. If it could serve that purpose as well, especially in situations where it just happens to be on the right target at the right time and saves them from a death blow without us even having to react, that would be a little OP, I think.

For point 3, I'll argue that it loses some utility in fights that don't have consistent AOE, but even in non-consistent AOE, I can still usually get 2-3 bounces out of each cast. Sure, that's not optimal, but in those situations, a lot of out other abilities like PoH and CoH aren't terribly effective either. So I wouldn't call that so much a flaw either as just a situation where it isn't as useful as it could be.

Finally, with point 4, I completely agree that it's a flaw, but I'm non completely sure how they should go about fixing it. It should probably prefer not to bounce to a target that already has PoM, but what about situations where it gets cast on the same target, or a target with PoM is the only one in range? Do they stack such that both heal on the next damage taken, or do they heal damage taken sequentially in the order they arrived? When they bounce, do they bounce to the same target or calculate targets separately? Personally, I don't really care how they fix it, but I would like to see it fixed in Cataclysm.

[harky]

I consider the first point a flaw on the basis of Blizzard's own statements regarding other spells. That is that the cooldown of abilities should not be shorter than their effective usage. Wild Growth is one of the big examples. Wild Growth, like PoM, lasts longer than its cooldown. Blizz cited that as a flaw in design in regards to Cata changes. The problem really is the whole 'skillful use' thing doesn't exist. There is no way to take advantage of the 8 second cooldown except as a slightly larger Flash Heal on tanks. The only other time the shorter cooldown is remotely useful is if you notice someone else had their PoM eat your PoM, so you can toss out another one. So at best the shorter cooldown is compensation for another big flaw and at worst it's a flawed poorly tuned mechanic that encourages bad play.

For the second point the issue isn't all the complicated and is linked to the fourth issue and no, I have no idea how to solve the problem. The issue really is that there is absolutely no other heal which can not save someone from dying. All other heals have that potential and while I am a huge proponent of stabilization as a better judgment of the worth of healers, I don't agree with the implementation of abilities that do not have any potential to directly prevent death.

My problem on the third issue is that while you can't expect every spell to shine on every fight, it takes a very specific type of fight for PoM to ever be of full worth. You mentioned PoH and CoH, but that's a different issue. With CoH you can cast it irregularly when it's needed in those fights. Something like Rotface. Rotface is not a fight where you'll spam CoH, or PoH. However, if people do make a mistake they suddenly become very reliable. You have control over where the healing goes and when. With PoM you have something that is crippled by many fights (ie; Any fight without an aura), but is also not in your control.

The forth point... yeah, lulz.

[Azyoulike]

I guess I'd agree with Zeuq in that I really only think that the final point is a true flaw.

1) The short CD would be a flaw, because it means you can 'clip' your own PoM, but the alternative would be like a DoT with a CD equal to the duration - We used to have one of those - DP - and they removed the CD because it turned out being able to choose to clip was less annoying than not being able to do so.

2) That the heal is retrospective is bad from a 'the target can die' perspective, but good from a 'rarely a totally wasted cast' perspective. Casting Holy Shock for example on a Mark target in Saurfang would be a very bad thing most of the time because it means he won't get healed (from you) for another 2 seconds. Casting PoM on a mark is generally an excellent thing by comparison. If I could trade Holy Shock for PoM on my pally I'd do it in a heartbeat, and this 'flaw' is the main reason.

3) PoM is a strong single target heal, it hits harder than FH in the same cast time. Even on fights where it is not jumping it is still better than filler since it provides the most efficient way of maintaining inspiritation on a tank.

4) Colliding PoMS just suck. I actually think this is the reason it's not a true smart heal, because then collisions would become far too prevalant. The 'returns to caster' behaviour smells like some heuristic stuff added to try to reduce this.

[harky]

The first I mostly consider due to Blizzard's comments. Otherwise it's pointless, but not obstructive. You used DP as an example, but DP also had a large core flaw. It could only be cast on one target, but it still had a cooldown, etc. Removing the cooldown allowed DP to be used on your primary target easier as opposed to complicating things and limiting output. DP is actually a good example of the same flaw and one way to fix it. If PoM had a 10 second cooldown, or no cooldown at all then I wouldn't call it a flaw because it would either allow for optimal usage, or completely free use. The second I would say is a fairly bad flaw. If you want to compare compare it to something worthwhile. Holy Shock never gets cast because it's just horribly bad, that really doesn't excuse a spell that requires a very specific type of fight to be used as intended and instead relegated to a slightly stronger single target heal. But yeah, the fourth is the really big part. The first three I'd call technical flaws, but the fourth is a core flaw. So even if I conceded the first three the fourth is hard to excuse.