Cataclysm PvE Holy Paladin Theorycrafting

37 Blood Elf Warlock
6870
02/03/2011 10:02 AMPosted by Kyaza
It seems a lot of people are confused as to why haste is more important than crit. I'm not sure if it was mentioned as I didn't read all 17 pages worth of posts, but the reason Haste is greater than Crit in terms of stats is because Holy Radiance stacks with haste and you want your haste high enough to get an extra tick off Holy Radiance.

For those who are concerned about the haste/crit comparison, checking out the guide at elitistjerks is usually a good idea.

http://elitistjerks.com/f76/t110847-holy_cataclysm_holy_compendium/

For those who are interested.


The reason is far more complicated and involved the HPM/HPS dynamic of HL and DL, combined with the Total mana over the fight length compared to the total healing, while also considering burst healing that might be needed to survive a special ability, even though it reduces your ability to heal the rest of the fight. WoW is getting pretty stale ready again, so I don't know if I will every bother to finish my spread sheet that shows all this.
Reply Quote
85 Dwarf Paladin
7025
02/03/2011 10:02 AMPosted by Kyaza
haste is more important than crit.
I agree that it is more important while it benefits Holy Radiance (adding an extra tick) but as soon as you remove that benefit, as in stacking between tick plateaus, the value drops dramatically.

The above is slightly misleading. Haste is not more important than Crit - in fact, a solid 20% or more Crit raid buffed is possibly even more important than Haste even with its benefits for Holy Radiance, due to the consistency on Conviction (which is 9% healing all around). Crit also increases cast speed occasionally...

We all know that stuff. Point is, the stats are very complex in relation and making any kind of statement one way or the other even if in a specific enough context can mislead the less attentive players - especially those looking to affirm their belief in any one stat. I'm guilty of it on occasion also I am sure. :|

Reply Quote
85 Blood Elf Paladin
8460
haste is more important than crit.


The above is slightly misleading. Haste is not more important than Crit -

Point is, the stats are very complex in relation


I disagree with the first part of what I quoted, and hopefully I will prove the second one wrong.

1% haste makes us cast spells 1% faster, so in the time it would take to cast 100 spells with 0% haste, we would be able to get off 101 spells with 1% haste.

1% crit gives our heals a 1% chance to heal 1.5% their original amount. In 100 cast we would get the throughput as if we casted 100.5 spells with 1% crit.

You can see Haste is more throughput already, now, look at how much rating it takes to get 1% of each of them.

Haste-128
Crit- 179

So 1% haste is more throughput than 1% crit, and it takes much less rating to obtain. You can use the same reasoning for mastery. You could argue that conviction plays into this aswell, but most logs are showing at least 90% uptime on conviction on players with ~20% crit buffed. Conviction changing to holy hasn't changed a thing about our stat prioritzation.

Edit: Thanks for replying to my post, Lylthe
Edited by Nycgangsta on 2/3/2011 7:35 PM PST
Reply Quote
85 Blood Elf Paladin
7350
Its twenty percent crit raid buffed to maintain conviction, or 15% unbuffed.
.
That 20% unbuffed was from a model that didnt account for PotI crits
Reply Quote
85 Dwarf Paladin
7025
Nyc we've discussed this at pretty great length. If you'd like to review the thread go ahead, but I'll give you the numbers rundown as I witnessed them to be (and seeing as I hold to the Mastery and somewhat Crit supporting side of the discussion I am wary of getting the numbers wrong here).

100 Haste provides ~0.45% additional effective HPS (between HR plateaus)
100 Mastery provides (~0.30% - Rabel's #'s) to (~0.21% - Lylthe's #'s) additional effective HPS
100 Crit provides ~0.28% additional effective HPS (before Infusion of Light)

This takes into account the differences in rating, Mastery's rate of use (a variable value depending on healing style), and other factors such as which spells the stats affect.

It does not take into account the added mana cost Haste entails, reduced GCD from Haste, and it also does NOT take into account Infusion of Light (please someone do the math for that) which could be very significant.

My guess, given a full math rundown, these stats are going to look something like this for the purposes of pure HPS:

Crit = Haste > Mastery

For maximum healing output and efficiency in raids I still firmly believe that this stat balance is accurate:

Intellect > Spirit > Haste at or above 6.04% or 14.5% > Crit to 20% > Crit & Mastery equally (by rating)

-----

Edit: Just to reaffirm my numbers again since there was perhaps some doubt that I really was producing that level of Mastery use, here is a parse from my Halfus kill this week.

http://www.worldoflogs.com/reports/c9ii1or7nt4b76ya/details/33/?s=13941&e=14292#tab-healspell

Here you can see my Mastery outputting 8.8% of my healing - pretty significant considering my Mastery was @ just over 17% at the time. At ~50% use rate that puts Mastery's value (for me) at:

100 Mastery ~= 0.348%

At least... on Halfus.
Edited by Rabel on 2/3/2011 11:36 PM PST
Reply Quote
85 Blood Elf Paladin
8460
02/03/2011 11:22 PMPosted by Rabel
Nyc we've discussed this at pretty great length. If you'd like to review the thread go ahead, but I'll give you the numbers rundown as I witnessed them to be (and seeing as I hold to the Mastery and somewhat Crit supporting side of the discussion I am wary of getting the numbers wrong here).

100 Haste provides ~0.45% additional effective HPS (between HR plateaus)
100 Mastery provides (~0.30% - Rabel's #'s) to (~0.21% - Lylthe's #'s) additional effective HPS
100 Crit provides ~0.28% additional effective HPS (before Infusion of Light)

This takes into account the differences in rating, Mastery's rate of use (a variable value depending on healing style), and other factors such as which spells the stats affect.

It does not take into account the added mana cost Haste entails, reduced GCD from Haste, and it also does NOT take into account Infusion of Light (please someone do the math for that) which could be very significant.

My guess, given a full math rundown, these stats are going to look something like this for the purposes of pure HPS:

Crit = Haste > Mastery


I'm confused as to what it is you said in your post before the last line that lead you to saying it. In fact it seems to support Haste being better. Sure, Haste cost extra mana, but there is also downsides to crit. Crit is completely RNG and can never be relied on for throughput, where as haste is always there.

Plus, Haste also effects IoL procs, and allows you to cast more HL's than DL's. Ya, you've heard this all before, but I'm really just not sure what point your holding onto to say that crit is equal to haste, when the numbers you gave me proved haste is better.
Edited by Nycgangsta on 2/4/2011 9:29 AM PST
Reply Quote
85 Dwarf Paladin
7025
02/04/2011 9:28 AMPosted by Nycgangsta
Crit is completely RNG and can never be relied on for throughput
This is completely pointless to bring up due to the nature of RNG. It could just as likely help you MORE as LESS, I think we can safely call it equivalent. Yes, if you really don't like RNG and feel more comfortable using other stats - feel free, it won't completely ruin your healing. On the other hand it might.
02/04/2011 9:28 AMPosted by Nycgangsta
Haste also effects IoL procs
Wrong. It affects the spell being affected by IoL the same as it previously did. If anything you could argue Haste has a minor effect on instants (which Holy Shock is) which might allow you to get off HS more often without interfering with your other heals, but that's not really something quantifiable I think.
My guess, given a full math rundown, these stats are going to look something like this for the purposes of pure HPS:
This is my guess based on how Infusion of Light works - it aught to have a fairly strong impact on HPS generated through Crit, and Crit isn't all that far below Haste for pure HPS output. I'm guessing that IoL actually puts Crit right up at Haste's level for HPS throughput.

02/04/2011 9:28 AMPosted by Nycgangsta
Sure, Haste cost extra mana, but there is also downsides to crit. Crit is completely RNG and can never be relied on for throughput, where as haste is always there.
Your attitude makes me infer that you're basing most of your support of Haste on a gut feeling. The way you dismiss the added mana cost of Haste's HPS increase, the way you assume Haste has 100% effectiveness moments after you proclaim Crit unreliable. In fact, Haste is only effective at its full value when you are spamming spells one after the other. The moment you stop casting both Crit and Haste lose some of their HPS value. The only stat that does NOT lose its HPS value when you pause between casts is Mastery, due to it staying on target completely independant of the caster's spell choice and cast rate.

In fact, Mastery only loses HPS value when you spam spells faster on one target, not giving them a chance to take more damage (eating up the Illuminated Healing shield) OR when there is not enough damage going around to absorb the shields fully (which does happen fairly frequently in raid encounters). Note that the decrease in value is already fairly well mathed in for Mastery in the numbers I provided.
Edited by Rabel on 2/4/2011 1:26 PM PST
Reply Quote
85 Blood Elf Paladin
8460

Haste also effects IoL procs
Wrong


More cast=More chances to crit=More IoL procs, I don't see how thats wrong. It doesn't have nearly the effect on it as crit, but an effect none the less.

Edit: Also, if your not casting, why are you worried about HPS?

Also, sorry I didn't quote this, but you said Crit is right next to haste for throughput. Haste is more then 150% more throughput then crit, using your numbers. I'd hardly call that close.
Edited by Nycgangsta on 2/4/2011 2:42 PM PST
Reply Quote
85 Blood Elf Paladin
7350
"My guess," he says. Srsly, Rabel, u were doing good a few pages ago, now ur trying to reinvent the wheel without it being round again. My guess is next time I'll be inclined to think TLDR a lot sooner.
Reply Quote
37 Blood Elf Warlock
6870
Haste and Crit both effect IoL. Crit is needed to proc IoL in such a manner that having 0 crit means 0 IoL. However, haste is analogous to casting rate, and as such IoL depends on the rate of HS, which in turn depends in part on the rate of Daybreaks, which depends on rate of HL/DL. Asside from Haste's effect on the proc rate (number) of IoL's, hate also has an effect on the EFFECT of IoL procs. Much like additional crit will make your IoL cast crit more, additional haste will make it cast faster. The cast time reduction of IoL does not affect the GCD, thus the DL/HL cast time is reduced below the current GCD. Haste is thus a linear increase in the HPS of ALL IoL casts, in the same manner that it effects regular casts.

Do I have to finish up my spread sheet just to stomp out this stupid stat debate?
Reply Quote
85 Dwarf Paladin
7025
I had a detailed math up of IoL and Daybreak all typed up but went AFK and the power went out while I was out. Very frustrating.
"My guess," he says. Srsly, Rabel, u were doing good a few pages ago, now ur trying to reinvent the wheel without it being round again.
LMAO! Are you serious? Every time anyone does any math here it keeps lending MORE and MORE toward Crit & Mastery outpacing Haste in the big picture. I'm talking about both HPS and HPM (longevity).

That's an educated guess. Educated, as in what causes me to spell it "you".

Someone still needs to find a way to compare the values of the three stats in terms of Longevity vs HPS - at what point are you sacrificing too much for the small gain in HPS? This could be mathed out (maybe), or it could just continue to be referential.

Do I have to finish up my spread sheet just to stomp out this stupid stat debate?
I'd love to see a spreadsheet that contains everything in this thread relating to Mastery, Crit, and Haste. A complete Holy Paladin sheet would be amazing, but that's a huge undertaking - what does yours encompass?
Reply Quote
85 Dwarf Paladin
7025
More haste = More casts = More Daybreak procs = More HS = More IoL
Quite right, I forgot about this.

Lets do the math. I'll use this parse for my stats (including several boss kills): http://www.worldoflogs.com/reports/c9ii1or7nt4b76ya/details/33/?enc=bosses#tab-healspell

Now to figure out IoL procs value for Crit HPS we'll need to take Crits 1 to 1 value of ~0.56% per 100 Crit Rating and multiply by the Haste value provided. In this case its a flat 250% (1.5 sec every time, applied before Haste rating). This equates to ~1.4% Haste per 100 Crit Rating. That 1.4% Haste is applied only to DL and HL casts for the duration of the fight, and also only occurs once every 6 seconds (barring Daybreak procs). Divide by 6 and reduce based on the parse data for HL and DL (24.7%) and we get:

100 Crit Rating = ~0.0576% HPS from IoL
100 Crit Rating = 0.28% + 0.0576% = ~0.3376% HPS

Significant, but not nearly as much as I thought it might be. Next lets work in Daybreak (much harder I think).

Haste's effect on Daybreak is going to be based on the number of casts of HL, DL, and FoL vs the number of GCD's spent on other spells for a fight duration. Without a mod specifically tracking that its rather difficult to nail down - this parse only indicates the number of spells HITTING targets not the number of times they are cast. Using the numbers I do have, the fight durations, I've come up with a rough ESTIMATE percentage:

~30% of my casts were HL, DL, and FoL
100 Haste Rating = ~0.78% x 30% = ~0.234%
~0.234% x 20% (Daybreak chance) = ~0.0468%
100 Haste Rating = ~0.0468% Daybreak procs

Note: When you think about it it becomes clear that Daybreak is not an HPS increase except through IoL, which after all the reductions is going to be pretty insignificant. In fact, using the extra HS from Daybreak is probably a decrease in HPS but an increase in HPM. I'm going to keep this here for use at a later date, a fairly significant HPM boost for Haste I think! For better HPS, keep casting DL. For better HPM, abuse your Holy Power, Holy Shock, Holy Radiance, and Beacon as much as possible.

The next step is to apply Daybreak and IoL together... which I don't have time for right now. Feel free to use my numbers or tell me how terrible they are. :D

Because people apparently don't like long posts in a theorycraft thread:

- TLDR!! -
IoL is a pretty solid boost in HPS for Crit
Daybreak is a boost in HPM for Haste, and with IoL a small boost to HPS (much smaller than what Crit gets)
In both cases using HS is a HPS loss over straight casting DL (I'm pretty sure)
Edited by Rabel on 2/5/2011 4:30 PM PST
Reply Quote
85 Blood Elf Paladin
8460
Thanks for taking the time to clear this up, Rabel.

Rabel looking at your log I can see why you're gearing the way you are. You seem to be casting much more DL's and very few HL's, which makes Crit and Mastery a better stat for that play style. I use a completely different play style which favors HL much more then DL, with HL coming at ~20% of my heals, Beacon ~20% of my heals, HS 10%, DL ~10%, HR ~10%, PotI ~10%, WoG/LoD ~15% then the rest is mastery/FoL, but since most of my heals come from HL and PotI (Through the beacon) it makes haste a better stat for me. I'm not sure which play style is most used, I haven't been that into logs this expansion, but I can see a difference now.
Reply Quote
85 Dwarf Paladin
7025
more DL's and very few HL's
Well, when I didn't people died, so I started healing like this. That was perhaps due to my stats at the time (almost entirely focused on Mastery).

If it ain't broke don't fix it right? *shrug* Maybe I aught to try some different healing methods.
Reply Quote
85 Blood Elf Paladin
7350
IoL was shown by EJ models to be effected by crit by a very small amount, and even then it increases hps/hpm very little, depressing yes. Just make sure you use that IoL proc if u get it after ur first HS on a Daybreak, then back to use the 2nd HS... I've gotten some lucky strings of Daybreak, HS, IoL, HL, Daybreak, HS.... that are just insanely fun and efficient. Daybreak on the other hand, can only be increased in frequency by increasing haste.

I will always have 15%+ haste self buffed for 1.8sec HL/DLs from now on. I dropped down to the HR plateau at 14.5% but those 1.9 second HL/DLs were killing me and my twitch healing, making me resort to FoL more.
14.5% Haste is very important for that extra 2nd tick of HR. Next in importance to me is crit at 15% to maintain Conviction. I'll have both of those once I get home from work tomorrow and get the DMC:Tsunami. My crit dropped a little under 15% when I regemmed a couple slots with a +20 haste slot bonus with Spirit/Int from pure Int in response to the numbers posted by Lylthe the other day.

All of our heals are increased in power by haste and most have procs or plateaus that make haste even more powerful in thier use. More haste, more HR, more HoPo, more casts under AW/DI, more HR ticks, higher hps from HL/HS, more PotI to ourselves and beacon, more targets healed, quicker use of DL spam under GotAK, more free GCDs for Judging, higher hps with HL/HS to save mana and regain mana for those times after we have to DL spam, quicker building of HoPo through ToR use for when you know that AE is coming and ur gonna have to stack (or cheesing the ToR delay to get LoDs that are a mere 2-4seconds apart), quicker building of 3 HoPo + IoL/Daybreak procs under DP to play catch up if ur DSac gets eaten, etc etc as I'm sure I'm missing something.

Sorry Rabel, but I say TLDR, because ur modeling ur own logs with an unconventional, and already mathmatically proven inferior, gearing/reforging/gemming pattern by using mastery/crit over haste. Analysis on a suboptimal method, is only providing confusion to people that come to this thread for advice. They think OH LOOK numbers, must be right, and its not right, its a theory craft on an inferior method. As evidenced by ur dependence on DLs in that log imo. And as you have missed the obvious idea that increasing crit opportunity will increase crits this late, lends little credence to your theories bro. How many ranked parses for ur raiding category have you had btw? Any?

If you tried the haste method and gave it a couple weeks to get the muscle memory down, u'd have better hps/hpm. Unless... I've forgetten if ur one of those effected by the terribad latency monster atm, then disregard that, ur gearing and healing perfectly in that case. And all math aside, the quicker our heals, the better our ability to heal stupid and save a raid is. We have the best healing coefficients of any healing class on our direct heals, and all of our offtank heals heal the bacon tank for 50%+PotI, putting us in the best position to be raid savers due to people getting RNG'd by some fight mechanic or a positioning problem/miscommunication

EJ is crap for a lot of classes and specs, but ret and holy paladins isn't in that category. Take ur math there if you really think its worth it weight and will hold up.

Lylthe on the other hands posts very good numbers, provides the complete basis of the math and doesn't state he's "guessing" repeatedly for conclusions.

Sorry if I come off harsh, I shouldn't really care so much, the greater the number of paladins preforming suboptimally the less nerfs we get. And I think I just TLDRd my own damn self, lol
Edited by Rathimis on 2/6/2011 8:44 AM PST
Reply Quote
37 Blood Elf Warlock
6870
Well to be fair, the math on EJ is mostly incomplete due to people trying to analyze a small subset of the model without understanding the implications of what they have left out. The 'math' that showed little favor for crit in regards to IoL, was also not factoring the effect of Haste o the GCD. I keep prodding these people so they realize that they have only a small picture, and a missleading one at that.

There are two factors to consider here. The 'normal' factor is "How much healing can I do during the fight, and how can I maximize this". This is the 'Fight Healing' approach that rawr initally used, and can give a good aproximation of a fight, under the assumption that the most efficient way to heal will provide the required HPS. In these situations, Haste varies from being a reduction in total healing to being a greater contributer then Crit, but this dynamic relies heavily on the number of Judgments you make relative to total time, relative to total mana. This is because of two interactions, the first being that more haste causes more mana drain, and you may be in a position where you cannot downgrade a DL into a HL using haste as a HPM gain. The other end of the spectrum comes from Judgments not being effected by Haste, resulting in too many judgments lowering the total effect of Haste as a percentage of spells cast. This dynamic can be counter intuitive, as haste falls into a bell curve of sorts for effectiveness.

However, when we consider the case where you do not meet current HPS requirements, we enter into the haste terriory. If you have an ability like Mimiron's blast, that needs X HPS for a short time, haste is the best stat to try to use a more efficient heal to cover the HPS. This aspect of a fight seems to be downplayed in normal mode encounters as proper excecution is far more effective then gearing for this, based on reports, the same is not true for Hard Mode ecnounters, where there is definatly a desired HPS level for certain abilities.

Re Rabel: My sheet currently models the interactions of every stat, and all talents that effect healing. The only things not yet implemented are a WoG - LoD option, Cooldowns like AW/DF/GoAK/LoH, and a Tower of Radiance option. As well as some buffs like Flask and potions. I might get some time tonight to work on it more so it is not so cryptic.
Reply Quote
6 Troll Shaman
0
So, should we gem for spirit bonuses or not? I have no freaking clue, but here's my first crack at it.

https://spreadsheets.google.com/ccc?key=0AhYqR2TzgIgadGNZQ09vOGF2LVo4YW80ZEdpbmRzNGc&hl=en#gid=0

Adding 10 int, makes HL hit 6.388 harder, DL 17.05118 harder, FoL 12.76251 harder. I'm going to ignore the extra mp5 since it's really really small. Also, the starting mana pool is basically normalized out, so I'm going to ignore that too.

Using the baseline from the previous gem posting, spirit is worth ~0.6 mp5/spirit. Adding 10 spirit would add about 6 mp5. The spreadsheet does a more detailed calculation.

Now things get fuzzy. The way I'm going to look at it is that I'm going to use the additional mana, and using HPM numbers convert it into equivalent healing. Now what makes this difficult is the time frame to regen. First, just because you regenned the mana, doesn't mean you can use it. You need an additional cast. Also, we don't straight spam. We occasionally throttle (at least I do) if no heals are needed.

But for the sake of simplicity, I'm going to assume that 2.0 seconds has elapsed for HL and DL and 1.2 seconds has elapsed for FoL.

Using the appropriate numbers:
+10 spirit "heals" for this much more
HL - 11.09928314
DL - 8.879223981
FoL - 4.012761832

+10 int heals for this much more
HL - 6.388650274
DL - 17.05118927
FoL - 12.762512

I'm not very happy with analysis, so any suggestions to improve it are greatly appreciated. Worst case, I'd have to write a full holy paladin simulator.
Edited by Lylthe on 2/6/2011 10:49 AM PST
Reply Quote
37 Blood Elf Warlock
6870
For my stat weight of Int:15, Spi/Haste/Crit: 5-8. This means that a 10 int bonus per hybrid gem is almost a wash, a 20 Int bonus for 1 gem is totaly worth it (obviously). However, this means that you need roughly 20 Spi/Haste/Crit for the bonus per gem to validate it.
Reply Quote

Please report any Code of Conduct violations, including:

Threats of violence. We take these seriously and will alert the proper authorities.

Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.

Harassing or discriminatory language. This will not be tolerated.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)

Reported!

[Close]