I was reading up on this Paladin thread on the WoW forums on yellow gems. Aus (Vis Maior) made a statement saying the following:

Luminous Pyrestones are bad. Stop using them.



That's one of the excerpts I gathered from the entire thread.

Check it out. Most of it's a good read. If you ignore the flaming and name calling at least.

http://forums.worldofwarcraft.com/threa ... 4170&sid=1

It's dangerous to simply cite someone who doesn't provide numbers to support their explanation, regardless of what guild they come from.

He is correct in asserting that intellect produces the lowest returns of all the stats. A 10 INT gem will provide 12.1 INT talented and raid buffed. That is the equivalent of: 0.15% crit, 4 +heal and about 2.5 mp5 over a 6 minute fight (much less for a longer fight). The crit will tack on some additional theoretical HPS and mp5, but 0.15% will barely contribute to those numbers. But he fails to even acknowledge the +healing component of the pyrestone, which is its strong point. So to argue against its intellect component is fallacious.

But I'm not an expert in this area, so I can't really provide insight beyond this. I know it is commonly believed that 22 heal, 11 heal / mp5, and 10 crit or 10 haste gems are ideal in later levels of progression. Which moreso than the others depends on other stats.

But I've also had people argue that the +healing component of 5 INT / 11 heal gems make them surpass that of crit and haste when you've only just begun gearing. With naturally low +heal stacked on your gear, the HPS potential of crit and haste are both greatly diminished. So I've had people argue to me that int/heal gems are ideal for meeting the insightful earthstorm meta requirement.

Anyway, I will leave that to others to provide the numbers to either prove me right or wrong. But really my point is that we shouldn't be relying on vague assertions, regardless of who they come from. Numbers or at least empirical data are key.

I'd second Lume's comments, and further assert that "it all depends".

Epic gems are far more accessible now than when they were solely the province of high-end raiders, which means that the old assumptions about what gear they're going into are not entirely correct any more. There may be a "generally ideal" set of gems for paladins hitting specific BT/SW fights in X many bits of T6, for instance, but that ideal doesn't necessarily apply to people who are putting epic gems into T5 gear, or badge gear, or whatever.

Tier 5, for instance, has 0 red sockets, and 5 yellows - but stacking spell haste or spell crit at the expense of solid +heal, at the Tier 5 content level, can be fairly problematic in its own right, not to mention the fact that the healadin metagem of choice, the Insightful Earthstorm, requires 2R/2Y/2B for activation. (Yes, of course one can ignore the socket bonuses, but few of them are totally worthless.)

As a general rule, yes, Int is less useful than other stats, and the 11 heal/4 int gem is relatively weaker than the 22 heal, or even the 11 int/2 mp5 gem (although I'd argue that the 2mp5 on that gem is pathetic and weakens it compared with the equivalent range of rare gems). However, socket bonuses and metagem activation concerns are also relevant, and I can see entirely feasible situations where 11 heal/4 int > 10 spell crit, 10 spellhaste or something un-yellow that loses you a socket bonus or metagem activation.

This was a post I wrote discussing gems for healadins; it's not aimed at people who are using epic gems (since they ought to know enough not to need a guide ), but the general principles are still relevant, as is this portion:



In other words: yes, the Luminous Pyrestone is weaker than some of the other gem alternatives, but that doesn't always make it a bad choice.

Now, if only there was a JC-cuttable equivalent of the Iridescent Fire Opal (11 heal, 4 spell crit, drops from Heroic Ramps), yellow gem choice wouldn't even be an issue any more.

It was somewhere buried in... page 5.

I think.

I see it. Page 4. He's absolutely right, if you compare individual stats to what other gems can bring. But he doesn't compare HPS throughput of +heal vs. haste vs. crit at various levels of gear. The luminous pyrestone is the only yellow gem with +heal, so citing the impotence of intellect and then comparing its +heal component only to that of the teardrop spinel is what makes his developing argument fallacious.

He should be comparing each gem. So let's use this gear as an example:

http://www.chardev.org/?template=69940

As you can see, I've left two yellow slots open so I can figure out what would be best to use in meeting the meta requirement (and the meta is more than worth it, because it's a huge amount of mp5). We're going to use holy light as an example, because it is the spell most paladins use when HPS is key.

Raid buffed, you have 2156 heal, and a 25.05% crit chance with holy light. Blessing of light adds 580 to the holy light. The basic amount of haste this gear provides is 1.59%, bringing holy light's cast time to 1.96 seconds with light's grace active.

Two 5 INT / 11 +heal Gems

Socketing 10 INT and 22 +heal at a basic level. But talented and raid buffed, this is actually 12 INT, which produces an additional 4 heal and 0.15% crit, bringing up your healing to 2182 and your crit to 25.2%. So your holy light will receive 1559 bonus from your +heal at a 1.96 second cast time. Your average non-crit max rank holy light will heal for 4460. Your average crit max rank holy light will heal for 6690.

(6690 * 0.252 + 4460 * 0.748) / 1.96 = 2562 HPS.

Two 10 crit Gems

Socketing 20 crit for an additional 0.91% crit. Your crit then becomes 25.96%. 1540 bonus from your +heal. 4441 average non-crit. 6661.5 average crit. 1.96 second cast time.

(6661.5 * 0.2596 + 4441 * 0.7404) / 1.96 = 2560 HPS.

Two 10 haste Gems

Socketing 20 haste for an additional 1.27% haste. Your holy light now has a 1.93 second cast time. 1540 bonus from your +heal. 4441 average non-crit. 6661.5 average crit. 25.05% crit.

(6661.5 * 0.2505 + 4441 * 0.7495) / 1.93 = 2589 HPS.


Conclusion and Other Tangibles

Even with nearly all badge gear, haste seems to trump the pyrestones and crit in pure HPS potential. But at this level, the pyrestones outdo the crit. However, there are other factors to consider, such as the extra mp5 that crit brings and the theoretical loss of mp5 you get by using haste. Not being a paladin myself, I'll leave it to others to worry about that stuff.

I might have missed something with my math. The last time I worried about paladins was when I raided with one as my main back in Q1 2005. So I don't exactly concern myself with it. But by approaching it in this manner, I think it helps people actually begin to understand the possibilities involving their class and stat choices. And this is what Aus fails to do by opening with a vague assertion and developing the argument only half of the way.

That thread (and the 2 400-post ones that followed it) amused me greatly. They actually inspired my last blog post on the Johnny archetype.

I posted on page 13 a different way of comparing the two gems:

Luminous Pyrestone vs Spell Crit gem
= 11 heal +5 int vs 10 crit rating
= 12 heal + 0.06 crit% vs 0.45 crit%
= 12 heal vs 0.39 crit%
= 12 heal vs 8.6 crit rating

That's essentially the real choice: Would you rather net 12 heal or 8.6 crit rating?

Unless you're severly undervaluing crit rating, 8.6 crit rating is a lot better than 12 healing. Haste vs Crit is a lot more debatable, as is sacrificing socket bonuses for a red +heal gem. But it's pretty clear that Luminous gems are not as good as the alternatives.

One thing all the calculations seem to miss off is your raid group make up..

As a Pally healer starting the T6 content I ALWAYS have both a Shammy and a Shadow Priest in my group.. this TOTALY devalues the Mp5 stats from the "Elitist Jerks" values, and makes Haste a hell of a lot more attractive.

Another thing to think about -- is there an "ideal" level where you should stop stacking + healing and instead start stacking other stats ?
(I currently have about +1900 healing and in gear sets I can have up to 173haste (10%) -- which I am REALLY enjoying)

Altogether I think gem choice is VERY gear dependent

Man, jealous Asros is jealous. Wish that was the case for most paladins, but in my experience, its not.




Well, you should be focusing on crit as well, keeping over 30% buffed crit is a good. I have 2200ish healing going into sunwell and was close to 33%+ buffed crit(depending on trinkets and swappable gear). Now moving on to brutallus, Haste is king. Look at SK Gaming paladins....they are stacking 1 stat and 1 stat only.



Yeah, i guess, but for holy paladins, there isnt much of a choice on gear. A couple non tier pieces, but most of the non tier pieces dont have sockets.

But again, you're looking at only one tier. For people looking at different tiers - working out how to gem their T4 or T5 or badge gear - it's a very different picture.

Good point. I have tunnel vision from trying to regem for brutallus.


However, healing and + spell crit are your 2 main stats. I was gemming alot of INT in t4/t5 instances since it adds to crit/int/mp5 and scales with kings.

I decided to delete my post so I wouldn't look totally lost. You caught me before I could, though, I guess! ^_^

The way I worked it was as such:

T4/T5, I ignored the set bonuses and predominantly gemmed +Heal once I'd managed to activate my desired meta.

In T6 content I've gone for +Spell Haste in all the yellow sockets (of which there are a hell of a lot in T6 gear)

For fights like Brutallus you *need* the Haste and by the time you are in T6 gear your +Heal/+Crit/Mana Regen will be sufficient without needing to gem for it, yet your haste will most likely be suffering.

I keep a variety of "haste" sets enabling me to have anywhere from 2000+Heal/350 haste to 2200ish+heal/0 Haste all while maintaining the same MP5/Crit pretty much.

There are so many things that come into play, one cannot truly get a straight answer via math.

However, a suitable argument can be made based grossly on level of progression.
T4/ beginning of T5: +mp5 is appropriate
end of T5/ beginning of T6: +crit is appropriate
Sunwell: +haste is appropriate

This is pretty well accepted among the paladin community, there is nothing new here. The reason for the change with progression is also pretty well accepted: fol spam vs. holy light spam. Comparing fol vs holy light ratios of what each individual paladin has, you can easily convert mp5 to crit.

There are exceptions to this however, for my level of progression (cleared Hyjal, 5/9 BT), I have been stacking crit. My spell crit is extremely high right now, and mp5 relatively low. However, a guildie of mine, still stacks mp5 and +heal. We heal completely different in the same level of progression and there is not one clear choice as to what is better. They are just different.

More exceptions:

Group make-up, mp5, crit, mana-efficiency:
- Shadow priest and shaman giving you mana regen? You can stack haste as you dont have to depend on crit or mp5.
- Spamming holy light (down ranked as appropriate)? crit will normally have a higher value of return for each itemized point vs. mp5
- Spamming fol? mp5 will have a higher value of return for each itemized point vs crit.

Gimmick fights:
- bloodboil for instance: I can spam max rank holy light a significant portion of the fight, because of massive damage I am taking. Never mis

Spell crit:
The theory behind alot of the numbers, is to not take into account the increase in hps due to crits, as overhealing does not count. Crit is a way of sustaining yourself throughout the fight due to mana return. While crit will add some +healing the number crunching usually discounts it.

Intellect:
The theory behind alot of the numbers, is that max mana doesn't matter. Except for gimmick fights, you are at max mana only once during the fight, and that is right before you cast your first heal. Int gives you virtually no gain on mana except for the first cast (with the miniscule portion of crit that the intellect gives). Because of this, my suggestion is to base theory crafting numbers for int, based on an equivalency of +heal or hps, and not mana-efficiency.

Yellow gems:
Yellow gems become more of a personal preference. Ask yourself the following questions, and you may need to revisit these questions as progression happens:

1. Do I consistantly get a shadow priest/shaman in my group?
(If yes, +heal and +haste, depending on progression)
2. What is my ratio of FoL to Holy Light?
(If mostly FoL spam, +heal/int gem)
(If mostly HL spam, +crit)

A copy and paste of some numbers I crunched a while back transforming crit into mp5:



As you can see there, I value crit much more than mp5, as a +10 crit gem is roughly equivalent to 4.7 mp5, and you cannot get a +5mp5 gem to replace it.

:sigh: Sorry for the long drawn out response. But that is my take on the gems.

I am one of those gear confused Paladins Siha has mentioned. My guild currently farms T4 content with little prospect of making it into serious T5 before the xpac. All those Kara runs means I am heavy on badges and will probably be getting all the helpful badge gear and epic gems I can. With the new badge gear rated far above my own game play I am stuck in Healadin limbo.

With easy sauce Kara I don't know what to gem for since any improvement in gear wont change a thing in my curent game play. (Weeps a little) I'm thinking I should gem out in hopes of some shining knight sweeping me off my feet with a request to heal a TK run.

Now I feel I am rambling, and really haven't added anything to the topic, but I thought I would give what Siha said some credence. Yes, the confused badge over loaded paladins does exist, and they are stuck in a sad gemming purgatory.

I really wish that there was a craftable version of this. As it is, I try to run heroic ramparts as often as possible in order to collect more of them.