 |
02/01/08, 8:59 AM
|
#31
|
|
Bald Bull
|
Illumination scales based on how much mana you are burning (the more you use the more it returns). Holy Light costs more mana than Flash of Light for about the same levels of healing. Hence, Holy Light does give you more Mp5 from Illumination than FoL. Therefore, crit is generally better for Holy Light builds than FoL spambots.
And stop it with the overhealing argument. I've seen FoL spambots overheal more than HL builds simply because FoL is so efficient you don't bother to cast-cancel if your target is full. And of course, overhealing doesn't matter as long as you don't run out of mana. Its a really moot argument unless you're trying to top meters, in which case you have more problems than crit overhealing (and if your only concern is topping meters I encourage you to reroll a resto shaman or CoH priest).
The one thing you really don't seem to be understanding about HL builds is that you don't spam one rank of HL. HL downranking requires you to think and adjust your rank for every time you heal depending on how much damage your target has. You can't model that mathematically because damage is a very random thing. You can try to average out how much of each rank you use or try to use WWS to find it (but, at least to the best of my knowledge, it still can't divide by rank due to the nature of the combat log), but it won't be accurate. You can go on and on with the "more HPS and HPM from this rank of FoL compared to this rank of HL", but in practice it doesn't matter. You're entirely correct that I'm saying "fuck the theorycraft". Real game experience is much more important for healers than a theoretical maximum you obtain on paper, at least in my opinion. And my experience shows me that both builds are perfectly viable.
You're perfectly welcome to try and create an "ideal" gearset, but I would suggest you change the title to "Holy itemization for best FoL performance". If you're going to ignore one playstyle might as well make sure everyone clicking on the link knows it.
|
|
|
|
|
02/01/08, 9:31 AM
|
#32
|
|
Bald Bull
Blood Elf Paladin
Darksorrow (EU)
|
|
Illumination scales based on how much mana you are burning (the more you use the more it returns). Holy Light costs more mana than Flash of Light for about the same levels of healing. Hence, Holy Light does give you more Mp5 from Illumination than FoL. Therefore, crit is generally better for Holy Light builds than FoL spambots.
|
This is flat out wrong. I explained this a milllion times already, mana doesn't matter if you didn't run out, and if you ran out illumination returned the exact same amount of mana for FoL as it did for HL (aside from the extra crit on HL of course, but as I already said it doesn't really have much effect on how it actually scales with additional crit). You're saying crit is better for HL becuase it burns mana faster, while in reality you're comparing a strategy that uses all your mana to a strategy that does not. Illumination does not work better when you spend more mana, it works better when you have more mana, and HL vs FoL have no effect on this (outside of the base extra crit which again doesn't have much effect on the benefit of extra crit). I mean who cares how much mana illumination returned if you didn't use it???
Besides my theorycraft doesn't enforce FoL spamming and no HL... All it says is you cast FoL when possible and when more HPS is needed you cast HL. Heck you could even adjust it to completely ignore the existance of FoL and could still use my theorycraft to maximize your gear, you'd just have to assume you turn a max rank HL into a low rank HL instead of a FoL for the sake of HPS->MP5 conversion. And you'd still get very useful numbers, which may or may not favor crit depending on the parameters you put in.
As for crit overhealing, this is compeltely NOT bullshit. I don't care how much you actually overheal, but the fact you assume your heal will not crit means that if it crits it will overheal. This results in more overhealing than you would otherwise have. This doesn't happen on all heals, but happens on a non-neglicible portion of them. Therefore if you assume a crit heal does 1.5x healing you're straight-up overevaluating crit.
|
|
|
|
|
02/01/08, 10:16 AM
|
#33
|
|
Bald Bull
|
Originally Posted by galzohar
This is flat out wrong. I explained this a milllion times already, mana doesn't matter if you didn't run out, and if you ran out illumination returned the exact same amount of mana for FoL as it did for HL (aside from the extra crit on HL of course, but as I already said it doesn't really have much effect on how it actually scales with additional crit). You're saying crit is better for HL becuase it burns mana faster, while in reality you're comparing a strategy that uses all your mana to a strategy that does not. Illumination does not work better when you spend more mana, it works better when you have more mana, and HL vs FoL have no effect on this (outside of the base extra crit which again doesn't have much effect on the benefit of extra crit). I mean who cares how much mana illumination returned if you didn't use it???
|
In all respects Illumination is not regen. It simply reduces the effective cost of a heal. Since its a percent of the base mana cost of a spell, and Holy Light has higher base mana costs, you do gain more raw mana from Illumination for using HL over FoL, even at the exact same crit chances. HL then gets an additional 11% crit over FoL (9% if you use S3 gloves, but whatever), which is a sizable amount. You can say that with a 40% HL crit the effective cost of HL V (for example) is not 275 mana but 209 because of the 2/5 of the spells having a reduced cost. By the same token FoL will still cost 149 mana with the 29% crit chance it would have in the same gear. So that crit chance reduces the effective cost of your Holy Light by 24%, whereas its only reducing the cost of FoL by ~17%. Yes, this is mostly thanks to the extra 11% crit you get, but ignoring the fact that because of that 11% extra crit HL gets more out of crit is completely foolish.
And if mana doesn't matter who do you make such a big deal about HPM?
Originally Posted by galzohar
As for crit overhealing, this is compeltely NOT bullshit. I don't care how much you actually overheal, but the fact you assume your heal will not crit means that if it crits it will overheal. This results in more overhealing than you would otherwise have. This doesn't happen on all heals, but happens on a non-neglicible portion of them. Therefore if you assume a crit heal does 1.5x healing you're straight-up overevaluating crit.
|
You're already assuming all your pretty little FoL's go into effective healing, which is bullshit to begin with since you don't cast-cancel (and holy shit batman, FoL can crit too). Honestly, you can't pick and choose certain conditions to support your idea and expect people to take you seriously.
|
|
|
|
|
02/01/08, 10:22 AM
|
#34
|
|
Bald Bull
|
Originally Posted by galzohar
This is flat out wrong. I explained this a milllion times already, mana doesn't matter if you didn't run out, and if you ran out illumination returned the exact same amount of mana for FoL as it did for HL (aside from the extra crit on HL of course, but as I already said it doesn't really have much effect on how it actually scales with additional crit). You're saying crit is better for HL becuase it burns mana faster, while in reality you're comparing a strategy that uses all your mana to a strategy that does not. Illumination does not work better when you spend more mana, it works better when you have more mana, and HL vs FoL have no effect on this (outside of the base extra crit which again doesn't have much effect on the benefit of extra crit). I mean who cares how much mana illumination returned if you didn't use it?
|
Gearing for HL vs. FoL does have an effect on how much mana you have. If you're gearing with the intent to spam FoL, your mana regen from gear should be pretty much zero. With HL, the value of +healing drops due to that the additional hp/sec is less relevant, and the value of mana regen rises due to that it's actually needed.
Also, crit vs. mp5 scales directly with mana/sec spent. Regardless of how many semantic hoops you want to jump through, 1% crit gives you (mana spent per 5 seconds * .006) mp5. If you spend more mana, crit is worth more relative to mp5. Whether or not you actually need more mana is completly irrelevant. If you don't need more mana, you shouldn't be adding either of them.
|
|
|
|
|
02/01/08, 11:04 AM
|
#35
|
|
Glass Joe
|
I knew that flame post would get me in trouble...
I guess you can see Galzohar that were never going to agree with you. This forum is populated by the best in the game, and we got to where we are by knowing our shit. Do you notice that the focus on paladin items in BT/HJ have a ton of crit,(Hammer of Atonement, Savior's Grasp, Glimmering Steel Mantle, Blessed Adamantite Bracers, Girdle of Hope, ect.) that have no mp5 on them at all? Yellow Sockets are all over items(not that socket bonuses should ever be taken into account unless the gain is greater than the loss).
Until you get to BT/HJ, and get enough crit rating to achieve the HL paladins goal of 40%+ HL crit, you are right, FoL spam bot is more efficient and worthwhile. So If you like change your title to Pre-HJ/BT Holy raid itemization for the best performance. I'm not saying that I was spamming HL R5 in kara gear, thats just stupid. I was a FoL bot untill I got the gear that let me change into a more powerful healer, with the kind of power and longevity that is needed for BT/HJ that really cant be done effectively with FoL spam.
|
Mashing two buttons is the ONLY way to heal.
|
|
|
02/01/08, 11:20 AM
|
#36
|
|
Bald Bull
Blood Elf Paladin
Darksorrow (EU)
|
|
Gearing for HL vs. FoL does have an effect on how much mana you have. If you're gearing with the intent to spam FoL, your mana regen from gear should be pretty much zero.
|
If you read the original post, my assumption isn't you just cast FoL and nothing else. The opposite - my assumption is that FoL spamming will, at certain points, not be enough and you'll need to HL sometimes to make up for the lack of HPS. However you won't need to *all* the time or you'll go oom (or use low ranks of HL which don't really help anyway, but you can easily adjust the spreadsheet and the theorycraft to account for that playstyle as well, although those that support ignoring FoL as a spell prefer to ignore that fact).
"Mana returned from illumination" if you didn't go oom is a completely meaningless number. Additional mana is pointless if you're not using it, and therefore if you care at all about mana you should look at how much mana you could spend in a fight, not how much mana you regened. You really have to decide if mana matters to you or not, as the "spam HL with high crit to restore more mana" is complete BS. The more mana you spend the more mana illumination restores, but that doesn't matter unless all mana is used, in which case wether you used HL or FoL is irrelevent beyond the 11% crit difference.
As for crit scaling of HL VS FoL: Let's ignore crit overhealing for a sec here and assume both get the same healing multiplier and ignore it.
Mana after crits = mana before crits * (1/1-0.6c). This is explained in the main post.
Mana multiplier is 1/(1-0.6c).
dMultiplier/dCrit%=0.6/(1-0.6c)^2
crit chance multiplier dMultiplier/dCrit%
25% 1.176470588 0.830449827
26% 1.184834123 0.84229914
27% 1.193317422 0.854403882
28% 1.201923077 0.86677145
29% 1.210653753 0.879409506
30% 1.219512195 0.892325996
31% 1.228501229 0.905529161
32% 1.237623762 0.919027546
33% 1.246882793 0.93283002
34% 1.256281407 0.946945784
35% 1.265822785 0.961384394
36% 1.275510204 0.976155768
37% 1.285347044 0.991270214
38% 1.295336788 1.006738436
39% 1.305483029 1.022571563
40% 1.315789474 1.038781163
41% 1.326259947 1.055379268
42% 1.336898396 1.072378392
43% 1.347708895 1.089791559
44% 1.358695652 1.107632325
45% 1.369863014 1.125914806
46% 1.38121547 1.144653704
47% 1.39275766 1.16386434
48% 1.404494382 1.183562681
49% 1.416430595 1.203765378
50% 1.428571429 1.224489796
As you can see if you add 11% crit the amount of extra mana you get from crit doesn't change much at all. For example going from 39% FoL crit to 50% HL crit makes additional crit increae your mana multiplier 20% more effectively. Going from 25% to 36% makes additional crit increase your multiplier 9% more effectively.
Also those kinds of differences can be easily seen in the spreadsheet and then used to determine how good crit actually is given how you evaluate your effective healing from crits. The spreadsheet/theory in no way forces you to spam flashes, it works just as correctly to optimize HL spamming if the correct numbers are plugged in it.
It all sums up to how much you evaluate your crit overhealing - which isn't nuisance as you do overheal more with higher crit if you don't rely on crits. There's no way to avoid the increased overhealing from crits without relying on the crits. It doesn't really matter how much overhealing you would've had if you didn't crit, as long as with crits you definitely are overhealing more than that. A lot more since you choose the size of the heal based on the damage the target is taking assuming it does NOT crit. This is very different from +healing as you CAN rely on +healing and base your healing on that, and will not have the same overhealing effect that crit has.
Note that the basic theory says nothing about what stats you should stack. It just says how to calculate which stats to stack, which was the original purpose of this thread, which seems to be largely ignored. The "crit stacking and HL spamming isn't worth it" is the results I was getting with the variety of stats I was putting in it. If you put in your stats in that theory and get different results, I'd love to see it. If you find a problem with the basic theory, I'd love to see it even more. But saying "your theory doesn't work because crit stacking and low rank HL spamming is the way we clear the game" just shows you didn't even read the original post from start to finish.
If you're going to disagree, at least show how your way of healing is better, rather than say "I do it this way, it works, I'm cool, you're not".
Last edited by galzohar : 02/01/08 at 11:33 AM.
|
|
|
|
|
02/01/08, 11:32 AM
|
#37
|
|
Bald Bull
Blood Elf Paladin
Darksorrow (EU)
|
Main post edited to show you can do the same math with downranked HL as your baseline efficient heal if you really want to, and it'll work just fine (in terms of showing you which stats would help you the most).
|
|
|
|
|
02/01/08, 12:22 PM
|
#38
|
|
Bald Bull
|
You're still completly missing the point. Yes, over the course of spending a fixed amount of mana Illum will return the exact same amount of mana regardless of what spell you cast. However, over the course of spending a fixed amount of mana, mp5 will not return the exact same amount of mana. Suppose you have 10,000 mana, 100 mp5, and 10% crit:
With a spell that costs 100 mana/second, you burn your initial 10k mana in 100 seconds. After 100 seconds, you will have 2000 mana from your mp5 and 600 mana from your crit.
With a spell that costs 500 mana/second, you burn your initial 10k mana in 20 seconds. After 20 seconds, you will have 400 mana from your mp5 and 600 mana from your crit. By multiplying your mana/second spent by 5, the value of crit relative to mp5 has increased fivefold.
To look at it in terms of time to OOM:
10,000 mana, 0 mp5, 0% crit, 100 mana/second spent: 100 seconds
10,000 mana, 100 mp5, 0% crit, 100 mana/second spent: 125 seconds (+25%)
10,000 mana, 0 mp5, 33% crit, 100 mana/second spent: 125 seconds (+25%)
10,000 mana, 0 mp5, 0% crit, 500 mana/second spent: 20 seconds
10,000 mana, 100 mp5, 0% crit, 500 mana/second spent: 20.8 seconds (+4%)
10,000 mana, 0 mp5, 33% crit, 500 mana/second spent: 25 seconds (+25%)
By increasing the mana spent per second, the value of crit relative to mp5 for increasing time until OOM has vastly increased.
To take it to the logical extreme:
1000 mana, 5 mp5, 0% crit, 1 mana/second spent: Infinite
1000 mana, 0 mp5, 100% crit, 1 mana/second spent: 2500 seconds
1000 mana, 5 mp5, 0% crit, 10 mana/second spent: 111 seconds
1000 mana, 0 mp5, 100% crit, 10 mana/second spent: 250 seconds
I'm not necessarily disagreeing with your spreadsheet's conclusions -- it pretty much says to do what EJ's paladins are doing, so at the minimum it isn't recommending something that doesn't work. HL vs. FoL is completely unrelated to whether or not crit has a larger benefit for HL than FoL. The value of crit in a vacuum is meaningless -- it's the value of crit relative to the other stats you could have that matters.
|
|
|
|
|
02/01/08, 12:27 PM
|
#39
|
|
Bald Bull
|
Originally Posted by galzohar
If you're going to disagree, at least show how your way of healing is better, rather than say "I do it this way, it works, I'm cool, you're not".
|
You're still missing the point of what everyone is saying. On paper your FoL build that has 0 regen but 5000 +healing will beat out a HL build (incidentally this is the line that basically made me write you off) in theoretical healing. No one is going to argue that.
What we're saying is that in game you can be just as effective using a downrank system, but you have to itemize differently. That is why this topic is rubbing a lot of us the wrong way. You say all we have to provide is experience, but by the same token you only have numbers to provide to us. I know your paper results say that we are terrible healers and stupid for clinging to an outdated system, but your spreadsheet doesn't account for the flexibility (and the ability to keep me awake) you gain from a downrank system over 1-button spam.
And I will say it one last time, Illumination does not return mana. It simply reduces the cost of a heal. You are still losing mana, not like "pure" regen in Mp5 (why I personally think comparing the two is difficult at best).
Just to put it in perspective, with 29% paper doll Holy crit the effective cost of our heals are as follows. Base costs are in ().
FoL VI: 149 (180)
HL IV: 144 (190)
HL V: 209 (275)
HL VI: 277 (365)
HL VII: 353 (465)
HL VIII: 441 (580)
HL IX: 502 (660)
HL X: 540 (710)
HL XI: 638 (840)
Of course, this means almost nothing since you rarely chain cast the same rank of HL. You get the point though that you're gaining more efficiency from crit with an HL system.
|
|
|
|
|
02/01/08, 12:33 PM
|
#40
|
|
Bald Bull
Blood Elf Paladin
Darksorrow (EU)
|
Your mana doesn't determine the fight duration. Nor does any other stat you have. To you duration of the fight is fixed, and the only thing you can change is how much healing you can do in that duration. That's why "time till oom" is completely meaningless. If a fight is to last 5 minutes, it'll last 5 minutes no matter what you do, and you want to maximize the healing you can do in 5 minutes. During those 5 minutes non-crit sources will regenerate a pre-determined amount of mana, which is independant of what spells you use, and then multiplied by the crit modifier.
Just to illustrate why "time till oom" is meaningless using a wierd example:
I use 200 mana/sec, and regen 199 mana/sec. This will result in T "time till oom".
Adding 1 mana/sec in regen will increase my "time till oom" to infinity.
According to your "time till oom" system, in that situation 5 mp5 is infnitely better than anything else. This just isn't true, considering no matter how much regen you have you can pretty much always spend it faster if you actually choose to and have targets to heal.
Crit is an average mana cost reduction, yes, which is exactly the same as a mana multiplier. Having everything cost 1/2 is exactly the same as having 2x mana (total, not base, of course).
And you keep saying I smash 1 button with no flexibility, which shows you didn't bother even reading the whole post. My whole assumption is that you have to HL due to lack of FoL HPS, just not all the time, due to variation of damage.
|
|
|
|
|
02/01/08, 12:52 PM
|
#41
|
|
Yet again, dead again.
|
A few points on crit that are often ignored:
I treat crit not as a regen stat, but as an efficiency stat. Instead of calculating the mana returned and trying to compare it to the mana returned from mana/5, I treat it as a spell cost reduction. With each percent of spell crit equaling to a .6% reduction in spell cost. The main reason being, the more expensive spells you cast, the more mana returned (saved) by crit.
Next, using the method that galzohar mentions, spam FoL and HL as necessity, the overheal from crit often helps make that necessity less frequent. The need to HL comes from two situations, incredibly high burst damage, which crit wont often make up for, and unlucky strings of non-avoidance, which crit will often make up for. Quite often when you switch to HL it is just because the tank has stopped dodging a few hits, and while FoL might be enough during "typical" avoidance rates, when it gets a string of hits, it might not be able to cut it. If you have some crits at that time, you might not need to switch. This is not an uncommon occurrence when you have ~30% crit rate.
Also, when you are primarily using a FoL healing method, having to switch to HL is typically reactive, and relatively infrequent, which means often times you will have cooldowns like divine favor and divine illumination available. And as I said earlier, I fully believe that with more crit, the frequency that you have to switch to HL is reduced, so there is a greater uptime where you can get by with cooldowns.
Mana regen is important, and I don't consider crit mana regen. Crit supplements mana regen as it reduces the cost of your spells, and allows you to go longer without having to switch to high mana spells, and it means a larger percent of your high mana spells are cast during cooldowns. All 3 stats have a place, and they are difficult to model. Healing from crits is unreliable, but by no means not helpful.
|
|
|
|
|
02/01/08, 12:55 PM
|
#42
|
|
Bald Bull
|
Your "varriation of damage requiring HL" is just an "Oh Shit Macro Time!". It is not flexibility. Flexibility is being able to see a person at -2.7k, throw them one heal and move on. Flexibility is to go straight from someone at -1.5k to someone at -4k without missing a beat. FoL does not have that, whether you're at 5 healing or 5000, and you can't pretend to have it without going into (*gasp*) using different ranks of heals.
And maximizing healing is stupid. Your job is just to keep the assignment alive unless you are assigned to do otherwise. People who try to maximize healing are the people who dip into raid healing when they're supposed to be on one assignment and can't respond fast enough to their assignment's damage. That leads to wipes. If you want to see big numbers on meters go reroll a DPS class. Meters and "maximized healing" mean jack shit for a healer.
|
|
|
|
|
02/01/08, 1:15 PM
|
#43
|
|
Bald Bull
|
Originally Posted by galzohar
Your mana doesn't determine the fight duration. Nor does any other stat you have. To you duration of the fight is fixed, and the only thing you can change is how much healing you can do in that duration. That's why "time till oom" is completely meaningless. If a fight is to last 5 minutes, it'll last 5 minutes no matter what you do, and you want to maximize the healing you can do in 5 minutes. During those 5 minutes non-crit sources will regenerate a pre-determined amount of mana, which is independant of what spells you use, and then multiplied by the crit modifier.
|
You mean when you hold the amount of mana spent constant, things which scale with how much mana you spend also stay constant?
Holy shit. What a revolutionary idea.
If you consider the length of the fight and amount of mp5 you have on gear constant while also stipulating that you use all of your mana (as more mana regen is pointless if you don't), then the amount of mana you're spending is constant. However, a HL-focused build doesn't spend the same amount of mana per second as a FoL-focused build, and the amount of mp5 on your gear isn't fixed.
For a 200 second fight, if you have 10,000 base mana:
75 mana per second spent: 125 mp5 or 55.55% crit required to make it through the fight
100 mana per second spent: 250 mp5 or 83% crit requried to make it through the fight
Increasing the mana/second required to keep the target alive by 33.33% has doubled the mp5 required, while only increasing the crit required by 50%. The value of crit relative to mp5 has increased with the increased mana requirement.
Originally Posted by galzohar
And you keep saying I smash 1 button with no flexibility, which shows you didn't bother even reading the whole post. My whole assumption is that you have to HL due to lack of FoL HPS, just not all the time, due to variation of damage.
|
If this is supposed to be a response to me and not just shitty quoting, what the fuck are you talking about?
|
|
|
|
|
02/01/08, 1:21 PM
|
#44
|
|
Von Kaiser
Blood Elf Paladin
Stormreaver (EU)
|
Originally Posted by flyingtoastr
Your "varriation of damage requiring HL" is just an "Oh Shit Macro Time!". It is not flexibility. Flexibility is being able to see a person at -2.7k, throw them one heal and move on. Flexibility is to go straight from someone at -1.5k to someone at -4k without missing a beat. FoL does not have that, whether you're at 5 healing or 5000, and you can't pretend to have it without going into (*gasp*) using different ranks of heals.
|
Hasn't he already said that he switches to HL when necessary? I would guess that if he was assigned to raid healing, as an FoL spammer he would do exactly the same as you given the same hp deficits, so I'm not sure what your point is.
Originally Posted by flyingtoastr
And maximizing healing is stupid. Your job is just to keep the assignment alive unless you are assigned to do otherwise. People who try to maximize healing are the people who dip into raid healing when they're supposed to be on one assignment and can't respond fast enough to their assignment's damage. That leads to wipes. If you want to see big numbers on meters go reroll a DPS class. Meters and "maximized healing" mean jack shit for a healer.
|
Again I think he was referring to maximising his hps on whatever his assigned target is to reduce the chance of things going wrong.
|
|
|
|
|
02/01/08, 1:41 PM
|
#45
|
|
Bald Bull
|
Originally Posted by Quozzy
Hasn't he already said that he switches to HL when necessary? I would guess that if he was assigned to raid healing, as an FoL spammer he would do exactly the same as you given the same hp deficits, so I'm not sure what your point is.
|
He is talking about those times when you tank does take a huge spike and you have to blow your macro (DF + HL XI) to get the highest HPS possible. That is the right thing to do for either build and I will never dispute it.
I don't believe he means that he switches to HL IX when a person is 4.5k down. An FoL spammer will typically either HL XI (and overheal) or if there is no danger simpley FoL VII him 3-4 times. A HL downranker on the other hand will go straight to a different rank HL, heal him once and move on. If that same person then takes 1.8k damage the FoL spammer will throw a single FoL, the HL downranker will switch and throw a Rank V HL.
Thats what I mean by flexibility. The ability to change exactly how much HPS you're pushing based on the required situation is the single greatest advantage a HL build has over an FoL.
|
|
|
|
|
|