Thursday, March 6, 2008

One More Time!

Sorry for the spam, but as a reminder, I have moved! This is the last post I will make here! :D

Tuesday, March 4, 2008

I've Moved!

Over to WordPress!

All of the posts and comments should be over there.

The new Of Light And Fury!

More on Spell Crit vs. Intellect vs. Mp5

A couple months back, I wrote a post detailing the relativity of the various paladin healing stats. Today I'd like to introduce a separate school of thought, albeit briefly.

At least...more brief than said previous post :)

This idea is that spell crit is better equated to a larger mana pool (more similar to intellect) than to mana regeneration (a la mp5).

Why? Well, it does make sense. Mp5 works all the time, while spell crit only returns mana if you have mana to cast in the first place. So, is spell crit more like intellect? Does it increase your mana pool, rather than regenerate mana?

When you cast a healing spell, and that spell crits, you get 60% of the mana cost of that spell back. For highest-rank Flash of Light, this is 108 mana, and for Holy Light, 504.

My previous line of thought was of course that this was more like mana regeneration, simply because this is what it looks like and is reported as in a combat log. However, it doesn't regen anything if you don't cast, or don't have the mana to cast. It only behaves like mp5 because a paladin is chain-casting by nature.

Let's look at an example to demonstrate. Say you only have 1000 mana, but somehow managed to gather up 75% holy spell crit chance, and 0 mp5. (Let's ignore mana regen from spirit for this example, as it's not really a factor worthy of consideration to any paladin.) According to my previous post, the spell crit -> mp5 conversion is C*X, where C is your spell crit and X is a number based on the spell you're casting. That means that with 75% holy spell crit chance, you regenerate 270 mp5 chain casting Flash of Light, and 945 mp5 doing the same with Holy Light.

Let's try casting a couple Flashes of Light. Many of them will crit (3 for every 4), resulting in 270 mana per 5 seconds returned to you. However, Flash of Light still costs 600 mp5 to chain cast. 600 minus 270 is 330 mp5 that you're still losing, and you'll be OOM in approximately 15 seconds (1000 mana / 330 mp5 = 3 (and change), multiplied by 5 seconds = 15), give or take depending on how the RNG (random number generator) favors you that day.

Considering Holy Light, I think it's pretty obvious what's going to happen here. You've got a 1 in 4 chance it won't crit, and if that happens you're unable to cast another. Even if it does crit, you're left with 664 mana, still not enough.

And then, because you have 0 mp5, you're out of luck. Again, it's unrealistic, but it serves its purpose.

So, is it better to look at spell crit as a % increase to your mana pool?

You have a % chance (your spell crit value) to cause each spell to have 60% of its cost refunded. The line of thought here, is that after chain casting a bunch of spells, each 60% from each crit is virtually tacked on to your mana total.

Using the above example again, with 1000 mana and 75% crit chance, it's probably better to look at it that way, because after so many spells you just won't be able to cast any more, at all, ever (without potions or being able to drink, etc etc) because you have zero real mp5.

What's the new formula then?

Before, when Illumination returned 100% of the spell's cost, you could view this as a % increase to your mana pool exactly equal to your spell crit value. If you had 15% spell crit, you could say you had 15% more mana, because you would lose 0 mana on a heal crit, effectively meaning you had that much more mana to begin with.

Nowadays, that number needs to be calculated, thankfully with a simple formula. And that formula is:

C * 0.6

Where C is your spell crit value, in percentage form. The resulting number is another percent, equalling what % extra mana your spell crit is providing.

Back to the example with 1000 mana and 75% spell crit and applying the formula, you end up with 45% as the resulting value. 45% extra mana from spell crit, meaning you virtually have 1450 mana, on average. (Remember it's still only a % chance for each spell to crit, so while you might get so many crits one time, you might get fewer (or more) the next time you try.)

In Zerei's case, she has 10800 mana and I think 24% holy spell crit chance in her pvp gear. Applying the formula, that's 13.8% (0.24 * 0.6) extra mana, or about 1555, totalling 12355.

In her healing gear, and these are guesses because I can't remember exactly the specific values, I think she has (about) 11810 mana and 18% holy spell crit chance. 10.8% extra mana, or 1275, totalling 13085.

Again, these end mana values are approximate because of the RNG. If the RNG gods are not favoring you one day, you'll experience slightly worse returns from Illumination.

I'm going to leave the old post there as it is, as I still believe it is valid and provides a basis with which to compare spell crit and mp5 when considering gear upgrades. However, I'm now of the opinion that spell crit as an extension of your mana pool is a more accurate depiction of the statistic. :D

Monday, March 3, 2008

Attack Power! (or stuff I think about when bored)

Sad? Probably.

Anyway, Attack Power is a numeric value attributed to a combination of things, building onto weapon dps. Basically, the more you have, the more damage per hit you do. For every 14 attack power you have, your damage per second is increased by 1.

280 attack power? 20 dps.

As such, it is one of the stats in high demand for physical dps characters. Largely because it is one of the most easily scaled attributes for many classes.

Why is that? There are a number of reasons.

First and foremost, any buff to your character's stats is a buff to attack power. Attack power is a stat that is at least partially derived from stats. For warriors, paladins, shaman, and druids, 1 strength is 2 attack power (it's actually more complicated for druids, as it can be dependent on form). For rogues, 1 strength is 1 attack power, and 1 agility is 1 attack power. For hunters, 1 agility is 1 ranged attack power (and 1 strength is 1 melee attack power, but lets pretend that doesn't exist).

Increase your warrior's strength by 10% through Blessing of Kings, and you've effectively increased the toon's attack power by 10% as well.

500 str = 1000 ap
550 str (500 + 10%) = 1100 ap

For this reason alone, strength is a powerful stat for those classes that benefit so heavily from it. If they possess a lot of gear that buffs attack power but not strength, Blessing of Kings gives them less than it normally would, because it only buffs attack power by buffing your base stats.

So for warriors, paladins, shaman, and druids, when you're looking at gear upgrades, you need to factor strength more heavily than it looks. This is especially true when socketing gems. Always go for strength over attack power if you can.

For instance, if a potential gear upgrade looks like this:

Old
36 Strength
34 Stamina
24 crit rating

New
30 Stamina
25 crit rating
74 attack power

74 attack power sounds like a lot, but remember that 36 strength by itself is 72. So the new item is an upgrade right?

Well if attack power is the only consideration, remember that with Kings, your strength is increased by 10%. That 36 strength on the old item becomes 39.6, and thus rounded out at somewhere between 79 and 80 attack power, while the new item's 72 will remain unchanged.

Of course this all relies on Blessing of Kings, but every raid should have it, if not for this than to also increase all the other stats for more hit points, mana, and mana regeneration!

However, and I'm starting to ramble a bit, but on the subject of blessings, Might is in general a more powerful blessing than Kings, considering attack power alone. You would need something in the range of 1000 strength to come remotely close to getting the same amount of attack power from Kings as you would Might. However, the increased hit points from the increased stamina (and even increased crit chance from increased agility) may let Kings win out, but that decision is, in my experience, up to the paladins.

This is a long way to say to warriors, paladins, and shaman in particular: strength > attack power in gem sockets!

Friday, February 29, 2008

So, this Holy Shock thing...

Here's what I said before on it. Since then, I've been using it so much more!

One nifty little trick...well, it's not really a trick, and debatably nifty, but anyway...is to use it right after a Flash of Light when doing spot healing. Sort of like how a Shadow Priest might queue up a Shadow Word: Death right after a Mind Blast, or a Fire Mage might use Fire Blast right after Fireball, it helps you do some burst healing, while giving you more freedom to move since the Shock is an instant spell.

Speaking of moving, it is of course our one instant-cast healing spell (aside from Lay on Hands...). With the buffs it's getting in 2.4, it will be even better at this. Any time you need to heal while moving you'll have a nice and powerful spell at your disposal. Remember the range limitation though, you have to be 10 yards closer to use HS than FoL and HL.

Speaking of burst, you can use it similarly to damage opponents. Start casting Hammer of Wrath, and once that finishes hit HS. Bam, burst 2k damage (more if you can manage crits out of them).

It's still not a very mana efficient spell, but then most instant-cast spells aren't. I think it's also further weighted down by its utility as a damage spell and healing spell. The long cooldown makes timing on it difficult as well, it may not always be there when you might need it.

But for an ability with such limitations, it's not as bad as I once thought it to be :)

Title Change, sort of

Since the ol' blog has deviated quite a bit from being a paladin-only talk place thing (I talk good), I've changed the title but not the address. Might also go through and do some colorizing and cleaning up here and there, but we'll see. Also might end up changing the address, but not yet.

Also, this is the blog formerly known as Blood Paladin, for anyone using a Feed Reading utility of some sort.