Holy Paladin Information

[Arianne]

So in the first section of your post you're arguing that overheal is bad, but in the second section of your post you're talking about how 5% more healing is better. That 5% more healing is going to be overheal most likely...

[Lizana]

So in the first section of your post you're arguing that overheal is bad, but in the second section of your post you're talking about how 5% more healing is better. That 5% more healing is going to be overheal most likely...

Please read the previous posts to see i am saying you can and should control your overheal. My previous comments about over healing were directed at controllable overheal. But if as a pally your doing 60%+ overheal your healing the wrong way. You take the 5% healing when its available because it makes every heal you cast more mana efficient and increases your hpm. The last post i made was to argue against even considering for mana regen how the extra crit in the long run makes a negligible gain in the long run. Those are two different arguments, but with both of them being valid. Unlike a lot of posters i look at the argument from multiple viewpoints and can make the argument from all of them instead of name calling and being ignorant when some one brings up a different side of the issue.

[mastersam562]

Please read the previous posts to see i am saying you can and should control your overheal. My previous comments about over healing were directed at controllable overheal. But if as a pally your doing 60%+ overheal your healing the wrong way. You take the 5% healing when its available because it makes every heal you cast more mana efficient and increases your hpm. The last post i made was to argue against even considering for mana regen how the extra crit in the long run makes a negligible gain in the long run. Those are two different arguments, but with both of them being valid. Unlike a lot of posters i look at the argument from multiple viewpoints and can make the argument from all of them instead of name calling and being ignorant when some one brings up a different side of the issue.

I highly disagree with that bolded statement. I over-healing is normally over 60% and I top healing charts in any raid setting ive gone on so im doing something right. Wish WWS would work so i can post some to show you, but maybe later. Look up thilain on bloodscalp US if your curious who i am.

Edited for some reports

Wow Web Stats ( basic nax run)
Wow Web Stats (more healers stepping it up on this run it seems)

I noticed that holy paladin dont need to control our OH because its so easy to get it all back. I have over 650 haste that is key to my mana regain with seal of wisdom, that include the 2 trinket which can proc more haste. I see alot of people looking at OH as a bad thing which I embrace it. Knowing when you can pop everything going into berserk healing mode, then knowing when to ease up taking a couple of swing at the boss to regain mana is the key to a successful raiding paladin my friend.

[jere]

A lot of that is fight dependent as well. All kinds of factors can influence the mp5 output.

Take one of Premonitions Patchwerk kills:
Wow Web Stats
The World of Warcraft Armory

The paladin there is a 51/20/0 spec and has the following statistics for that fight

Healing Time:  93 sec 

Spell           casts    %crit   %OH
Holy Light      22       .63     .57
Holy Shock      2        .33     .59

Adding the 3% crit there would net 1 more HL and probably 0 HS's.
764/93*5 = 41 mp5 from that 3% crit

The range can swing wide depending on healing time, fight mechanics. I would agree that it could be low often enough, though I wouldn't think 3 talent points for an extra 21-41mp5 is a bad price at all.

Also note the OH of the paladin in question. I think I recall seeing comments made about skill versus overheal on the order of 60%. This goes to show that sometimes skill can't lower it due to game mechanics (plus I would think that a premonition pally would have some sort of skill at this point).

[Akeber]

Never Said palleis were reactive healers, but if your target is going to take 4k damage why throw out a 12k heal when it isnt needed.

That is the definition of reactive healing. You see the tank short 4K on health, so you cast a heal that will land for 4k.

The thing is, by the time you "see" that happen, process what to do, and start casting a heal, and it finally lands, the tank could easily have taken another hit. What about streaks of avoidance, or non-avoidance.

There are certain fights where tank damage is predictable. For example, if you target sarth you can watch for his cast bar to see the breath coming and time your heal on the tank accordingly. Most damage is nowhere near that predictable.

5% extra healing is curently a waste, and will probably be a waste in Uldar too. The benefit of more crit has NOTHING to do with the extra healing it provides, but is well worth the increase in mana returned form illumination.

Now about crit being the main way a pally restores mana, here is a WWS section from a recent 10 man naxx

Spell Total mana
Illumination 190,392 Mana
Seal of Wisdom 176,837 Mana
Replenishment 152,951 Mana
Divine Plea 43,326 Mana
Mana Restore 34,800 Mana
Spiritual Attunement 27,261 Mana
Judgement of Wisdom 25,306 Mana
Blessing of Sanctuary 10,557 Mana

That's great and all, but it's naxx 10, let's look at something that you can't heal while sleeping. That's what we're really talking about, progression content that requires min/maxing, and taking all the best possible choices in gems, chants, gear, spec, etc. Here is the mana returned to a holy pally (specced into ret for the 3% crit and skipping kings) from an evening of sarth 3d attempts:

Illumination 304,948 Mana
Replenishment 126,292 Mana
Mana Spring Totem 82,586 Mana
Mana Restore 62,400 Mana
Seal of Wisdom 59,824 Mana
Soul of the Dead 59,400 Mana
Divine Plea 31,515 Mana
Spiritual Attunement 19,744 Mana
Mana Tide Totem 13,690 Mana
Restore Mana 4,226 Mana
Hymn of Hope 1,858 Mana
Judgement of Wisdom 702 Mana
Replenish 504 Mana

You can spec/gear all kinds of strange ways if the content is easy, it doesn't make it right.

[Lizana]

That is the definition of reactive healing. You see the tank short 4K on health, so you cast a heal that will land for 4k.

The thing is, by the time you "see" that happen, process what to do, and start casting a heal, and it finally lands, the tank could easily have taken another hit. What about streaks of avoidance, or non-avoidance.

There are certain fights where tank damage is predictable. For example, if you target sarth you can watch for his cast bar to see the breath coming and time your heal on the tank accordingly. Most damage is nowhere near that predictable.

5% extra healing is curently a waste, and will probably be a waste in Uldar too. The benefit of more crit has NOTHING to do with the extra healing it provides, but is well worth the increase in mana returned form illumination.



That's great and all, but it's naxx 10, let's look at something that you can't heal while sleeping. That's what we're really talking about, progression content that requires min/maxing, and taking all the best possible choices in gems, chants, gear, spec, etc. Here is the mana returned to a holy pally (specced into ret for the 3% crit and skipping kings) from an evening of sarth 3d attempts:

Illumination 304,948 Mana
Replenishment 126,292 Mana
Mana Spring Totem 82,586 Mana
Mana Restore 62,400 Mana
Seal of Wisdom 59,824 Mana
Soul of the Dead 59,400 Mana
Divine Plea 31,515 Mana
Spiritual Attunement 19,744 Mana
Mana Tide Totem 13,690 Mana
Restore Mana 4,226 Mana
Hymn of Hope 1,858 Mana
Judgement of Wisdom 702 Mana
Replenish 504 Mana

You can spec/gear all kinds of strange ways if the content is easy, it doesn't make it right.

Actualy reactive healing would be waiting for the tank to take thej damage then healing it. What i said, and reading is the key here, is "but if your target is going to take ". Theres very few fights in the game that the damage output from the fight is not a predictable stream of damage to a predictable target. If you know some one is going to be taking 4k of damage and you choose to throw a 12k heal instead of a 5k heal, you wasted your mana by not throwing the appropriate heal. I am not talking about sudden burst damage on the tank, i am talking about predictable damage against anyone in the raid.

On a side note, get of your high horse and realize that not everyone is in the game for progression 25 man content. As it has been said previously in the thread, this information is intended for a a pally new to healing. I dont know any pallies that hit 80 and go strait to s3d.

. That's what we're really talking about, progression content that requires min/maxing, and taking all the best possible choices in gems, chants, gear, spec, etc.

And no, thats not what this thread is about, and as a note your "progression" content was cleared by people wearing all t6 gear thats = to level 80 blues... So please reconsider how much you have truly progressed

One last note, please post your links to the WWS so i can verfiy the information, and see the person in context

And to jere

You are right on some spam healing fights overheal is hard to control, but theres and equal number of fights where it is much easier to control. The 60% number was picked out of thin air, and was there to server a purpose. By the fact still stands, if your having a huge amount of preventable overheal and you find your having mana problems, changing your healing rotation when possible would give you better mana returns over the long run than the extra 3% crit. An extra 3% crit on average means for your 100 casts of holy light it will take you 67 seconds to regen enough mana for a "extra" holy light. To me haveing to wait on average 67 seconds to to cast an "extra" holy light isnt worth 3 points, compared to the amount of healing increase 5% healing gives over that same time period.

[Akeber]

What i said, and reading is the key here, is "but if your target is going to take ". Theres very few fights in the game that the damage output from the fight is not a predictable stream of damage to a predictable target. If you know some one is going to be taking 4k of damage and you choose to throw a 12k heal instead of a 5k heal, you wasted your mana by not throwing the appropriate heal. I am not talking about sudden burst damage on the tank, i am talking about predictable damage against anyone in the raid.

On a side note, get of your high horse and realize that not everyone is in the game for progression 25 man content. As it has been said previously in the thread, this information is intended for a a pally new to healing. I dont know any pallies that hit 80 and go strait to s3d.

In reality, the incoming "damage stream" is never that predictable. If it was, the "optimal" healing makeup would all druids, putting out enough HPS to cover the DTPS of the raid for an easy win every time. Would make for some pretty boring encounters though. No one can tell the future. You can't predict if a tank is going to avoid 50% of the next 10 swings, avoid them all, or avoid none. If you could accurately predict all incoming damage, healing would be trivial.

I realize not everyone does 25 man content, and obviously a fresh healer is not going to walk into sarth 3d. I'm assuming any encounter that is new (to them) and hard, not old and easy. That is what "progression" is. It's going into an encounter you, and likely most of your group, have never done before. In this "new to them" boss fight that our "new pally healer" (as you mention is the target of this thread) has never seen, the tank WILL be taking unpredictable damage, the raid will be making mistakes and taking avoidable raid damage. In that type of situation the mana returned from 3% crit on a paladin that is likely less than optimally geared, where he and his raid are inexperienced, will be of much more benefit than %5 increased healing.

If you want to assume a 10 man naxx run, full of experienced people that have seen all the fights 20+ times, you're really moving away from the intended target of this thread. You can do all kinds of crazy things with specs, gear, or even the number of people you bring, when the enounter is something you've done before and are experienced with.

Here is a WWS and the pally healer I was referring to:

Wow Web Stats
The World of Warcraft Armory

[jere]

One thing that needs to be fixed is the comparison being presented. We are comparing a 3 point talent to a 5 point one. The comparison either needs to be 3% crit to 3% healing or 5% crit (conviction 2/5) to 5% healing.

[Lizana]

One thing that needs to be fixed is the comparison being presented. We are comparing a 3 point talent to a 5 point one. The comparison either needs to be 3% crit to 3% healing or 5% crit (conviction 2/5) to 5% healing.

We are comparing some one picking up the Sanctified Seals, and then spending 2 points somewhere else or picking up the new prot talent at teir 1 for 5% healing. By taking 5% healing you can still pick up 5/5 conviction therefore you point is invalid.

[jere]

No my point is perfectly valid. You aren't trading 5% healing to get 3% crit, you are only trading 3% healing to get the 3% crit. How you spend the other two points is up to you. Comparing spending 3 talent points to spending 5 is invalid. If you want to compare it to 3% crit + pursuit of justice or whatever other 2 point combination with the 3% crit, then great. Just comparing the two without considering those other two points is incorrect and slanted.

[Lizana]

Because every post against kings and against the new talent is people that are speced 51/0/20. And that has been the point argued back that any healing increase is pointless because of overheal. From your post "5% crit (conviction 2/5) to 5% healing." That is an invalid point. Now it is true a pally could spend 3 for crit and get 2% increased healing. And your right i havent done the math on that, neither have you, so please instead of just trying to poke holes in my math, do some to try to base your argument.

My main argument at this point is how little mana return you actually get from the 3% crit, becuase that is THE EXACT reason who you and other pallies say they take it, not for the healing increase a crit heal gives but for the mana return aspect of crit. That is what i am and have spent the past several posts addressing. If you notice the last post with my math in it, no where in the post did i talk about the 5% healing talent, instead i just broke down your argument with numbers proving how little mana is actually regened over a boss fight from that extra 3% crit, and from your own posts that is the reason you take the talent.

[jere]

From your post "5% crit (conviction 2/5) to 5% healing." That is an invalid point.

That was just an example. You know that. Don't be needlessly obtuse. The point was to make sure you were comparing equal things, which you weren't.

My main argument at this point is how little mana return you actually get from the 3% crit, becuase that is THE EXACT reason who you and other pallies say they take it, not for the healing increase a crit heal gives but for the mana return aspect of crit. That is what i am and have spent the past several posts addressing. If you notice the last post with my math in it, no where in the post did i talk about the 5% healing talent, instead i just broke down your argument with numbers proving how little mana is actually regened over a boss fight from that extra 3% crit, and from your own posts that is the reason you take the talent.

Except all you did was post math on a particular instance of a fight. If the fight had been shorter, it woud have been more mp5. If the fight had been longer, it would have been less. Again, all your math didn't really prove anything. 3% crit is still more mana than 3% healing (which is almost none given practical real world healing situations). The amount of extra healing isn't all that much from the math you posted either. So you either put 3 points in a talent that provides a small mana regen increase or 3 points in a talent that provides a small healing increase. It doesn't change the nature of the decision. Both are small returns (at least from the math you have presented thus far), so pick which ever one you like better. It isn't some cut and dry decision like you are trying to present, it is a matter of choice.

[jere]

You need to learn how to handle critism on your data Lizana. Getting defensive over things is not the way to be in this thread. It has a lot of good information and a lot of slant. You pick and choose a small group of WWS reports and claim an average. That is not responsible.

My stance is that it comes down to which you want more as a personal choice and that the math doesn't define which is better between 3% healing and 3% crit. This stance does not require math to back up, it requires logic. I have provided that. You disagree with it. That is fine. Don't get upset when someone finds fault with your presentation of numbers. The reasons for the faults were highlighted. It wasn't a case of "I feel" or "I think". It is simple logic pure and simple. If you don't understand how the value could fluctuate based on the WWS reports you look at, then I am not sure how to convince you. Maybe it takes some experience in statistics to understand how sample sizes affect the tolerance of your results. I am assuming you have some of this basic knowledge of this though.

The arguments are not against your math per say, but in the presentation/conclusions of your data. If you don't agree with them, then that is completely fine, but don't just brush them off because you don't like them. If you want data, go back to your own example where you found mana regen in the 20's (total), that is over twice your average value. That value is pretty far off from your proposed average. How do you think sample size of the WWS reports you used for the calculation affected your average value? What are the variance and standard deviation of your results for example? You haven't even looked at these if we go off the data presented so far.

If you are gonna do the math, then please present it correctly. You are definitely very smart, but in order to back up the conclusions you have made, you are going to need to do more work than grab a few WWS reports here and there and then average them. If you have any background in statistics at all, you know this is true. And remember, you are the one providing the math. The burden of proof to back it up falls on you, not everyone else. If someone comes around and says "you forgot to do this" or "this isn't really representative", then don't take it personally. It isn't an attack on you, but an effort to refine your results to make them more correct.

In either case, you do make very good points, so just because I disagree with some of them does not diminish how good they are. You could be very well indeed be right about your average value, but you actually haven't done the work yet to get that value for sure, so you could also be very wrong. Presenting it as "the value" is not the best route to take until you have actually verified it.

EDIT: If you are still upset with me, perhaps this discussion can be taken to PMs rather than diverting the thread.

[jere]

If you are looking for math, here is an interesting parse:
http://wowwebstats.com/kv1wkeq2evmue...347&a=x1af7000

Vincevega's guild killed Patchwerk in 105 seconds
During this time:
Holy Light casts: 59
Holy Light crits: 27 (46%)
Holy Light norms: 32 (54%)

He isn't spec'd into the 3% crit, he grabs BoK

If he had spec into 3% more crit, he would have had
Holy Light casts: 59
Holy Light crits: 29 (49%)
Holy Light norms: 30 (51%)

For a total of 2 more holy lights.

2*764 = 1528 mana regened over the course of the fight, which lasted 105 seconds
1528/105 = 14.55 mps
14.55*5 = 72.76 mp5 from that 3% crit and illumination.

That isn't to say this is representative of everyone, but it is a far cry from the proposed 9.6% average. Which is interesting.

[jere]

Yes, a sample size of 95 WWS is small, but in 95 of them not to see the numbers you are referring to. At no point did i ever say i found mana regen equal to 20 mp5, the largest i have found for a fight was 5 mp5 per % crit, so with the 3% crit your looking at 15 mp5 over the fight for 3 points spent. 3.1 to 5 is not that large of a range.

I was referring to this post you made:

WWS Link - Wow Web Stats

Patchwerk Fight

25 holy lights - 55% crit = 14 crits
0 Flash of lights - 0% crit = 0 crits
5 holy shocks - 38% crit = 2 crits

Add in 3% more crit

25 holy lights - 58% crit = 15 crits
0 Flash of lights - 0% crit = 0 crits
5 holy shocks - 41% crit = 2 crits

So by increasing crit by 3% your gaining 1 holy light crit on average so your gaining 764 mana over the course of a 145 second fight or 26 mp5.

This would be about 8.7 mp5 per point from your calculations. Though looking at how you calculated the data, I can see why your numbers are lower. You aren't reading the WWS correctly. He didn't cast 25 holy lights in that fight, he cast 25 non-crit holy lights. If you expand out, the number of crits was 31 out of 56, or 55.36%. An increase of 3% (58.36%) would have put him at 32.68 holy lights. This will statistically round up to 33 more often than it will round down, which would have boosted your results to 52 mp5 in that case. So the range on that particular pally for that fight was 26-52 mp5 with a higher chance for 52 than 26.

However, this is all funky math anyway because we are adding a statistically sound 3% to a non statistically sound value of 55.36%

Did you do all your HL calculations without expanding the Healing section and just counting the crit percentage against the non-crit HL count?

On a side note, I would be interested in seeing your tallies of the 95 WWS reports. Could you post them up on Google Sites/Docs? It would be interesting to go through.

[Lizana]

And you are correct this is an extreme outlier of my data set because of the time in combat. In fact it is one of the top 5 fastest kills of patch that WWS has ever recorded. Now whos being unfair with data? I should have clarified how i picked my patch wws results. I looked at guilds with kill times between 3 -5 min to have a normal data set that wasn't biased by extremes in times.

[Meeks]

And you are correct this is an extreme outlier of my data set because of the time in combat. In fact it is one of the top 5 fastest kills of patch that WWS has ever recorded. Now whos being unfair with data? I should have clarified how i picked my patch wws results. I looked at guilds with kill times between 3 -5 min to have a normal data set that wasn't biased by extremes in times.

You do realize that time has little to do with it? If the fight was longer in fact the mp5 numbers prob would of gone up as eventually they would of added a holy shock crit which it did not in the first calculation. The longer the fight is the more spells that are cast and you crit more in a direct relationship to the increased length.

Double the time of the fight and you will double holy light crits creating identical mp5 plus possibly add a holy shock crit resulting in even higher numbers.

[Lizana]

Actualy the longer the fight goes on the less likely you are to be able to keep up a 100% holy light spam at 1.7 second cast time. Infact i dont know a single pally healer that can chain cash holy light for 300+ seconds.

[jere]

Lizana, I think you were calculating incorrectly if you saw the edit in my recent post. It looks like from the examples you were only using the non-crit HL number as the total number of HL. That is going to change some of your results depending on rounding and could double the mana gains you calculated (not necessarily the average).

Also, could you post up your spreadsheet/document up on google? It would be neat to see the results individually to see how the deviate is with respect to the mean.

[Lizana]

Yes, there was an error on some of them, hence why the data was removed from the post because i could no longer be 100% verified in their results. I am working to redo the math in the posts. Theres no reason to post multiple times that you think i did the math wrong after the posts have been removed.