Nostalrius Begins - Quality wow vanilla realm (1.12)

First, lets get this out of the way:

Source:

Info on crit and hit chances | 2006-08-25 09:54 | Aeus
The way WoW calculates crit rate is over ALL attacks. Crit rate is not based on hits only. In other words, if you have a 5% crit rate, that 5% chance includes misses.

A 1% increase in your crit chance does not equal a 1% increase in your damage output. This is because of the simple fact that a spell that does not hit cannot crit. If you have a 15% chance to crit but a 17% chance to miss, you will only crit with 12.45% of your shadow bolts, (15*0.83). Now add 1% crit: (16*0.83) = 13.28, or a .83% increase. Whereas 1% hit is basically a 1% linear increase in damage output.

Raid buffed, raid consumables (no world buffs): 602 spellpower
602 spellpower * 6/7 (shadowbolt spellpower coefficient) = +520.3 dmg
SB Rank 9 = (455+508)/2 = 481.5 avg dmg
481.5+520.3 = 1001.8 avg sb damage

Scenario 1k:
v1 base: 1000 shadowbolts (15% crit, 17% miss)
830 hits, 170 misses
830 * 0.15 = 124.5 crits
(830 - 124.5) * 1001.8 = 706,769.9 (damage from non-crits)
(124.5 * 1001.8) * 2 = 249,448.2 (Ruin empowered crits)
Total damage = 956,218.1 damage

v2 +1%crit: 1000 shadowbolts (16% crit, 17% miss)
830 hits, 170 misses
830 * 0.16 = 132.8 crits
(830 - 132.8) * 1001.8 = 698,455.0 (damage from non-crits)
(132.8 * 1001.8) *2 = 266,078.1 (Ruin empowered crits)
Total damage = 964,533.1
Diff = 8,315.0 = +0.87% damage increase

v3 +1%hit: 1000 shadowbolts (15% crit, 16% miss)
840 hits, 160 misses
840 * 0.15 = 126.0 crits
(840 - 126.0) * 1001.8 = 715,285.2 (damage from non-crits)
(126.0 * 1001.8) * 2 = 252,453.6 (Ruin empowered crits)
Total Damage = 967,738.8 damage
Diff = 11,520.7 = 1.20% damage increase

Conclution: +spell hit is better on EVERY LEVEL until you reach the hard cap of +16%, you will always have a 1% chance to miss

*EDIT* - Fixed math ratios for landed critical strikes

Sources:
http://forums.elitistjerks.com/forums/t ... and-mages/
http://www.wowhead.com/forums&topic=355 ... rit-w-ruin
http://www.wow-one.com/forum/topic/4630 ... /?p=357910
http://wowwiki.wikia.com/wiki/Spell_hit
http://wowwiki.wikia.com/wiki/Attack_table

It is definitly true for most casters, but while a frostmage scales completely linear, warlocks have an exponential scaling with all 3 damage stats and crit has a very high value because it influences the imp SB uptime way more than hit and i believe there are possible gear levels where crit is equal or slightly better than hit.

You're not accounting for the ISB debuff.

Crit is worth the same as hit when you're around 15% crit and 83% hit. If you have under 15% crit then crit will be slightly better than hit.
I'm taking into account the proc resist bug on Nost, if we assume that's fixed then hit only becomes better than crit at 19%.

This is for a shadow bolt only rotation of course, if we add corruption to the rotation then hit becomes better than crit in pretty much every scenario.

Guybrush wrote:You're not accounting for the ISB debuff.

Crit is worth the same as hit when you're around 15% crit and 83% hit. If you have under 15% crit then crit will be slightly better than hit.
I'm taking into account the proc resist bug on Nost, if we assume that's fixed then hit only becomes better than crit at 19%.

In the current 7 debuff raiding enviroment, ISB is not a consistant, reliable debuff. It should not really be factored in, other than the odd bit of bonus damage, and here's why...

Regardless whether you are a horde or alliance warlock, the following debuffs should be on every boss
1. Sunder Armor
2. Fairy Fire
3. Curse of Elements
4. Curse of Shadow

And the following should be up on most bosses (not Broodlord, Chromag, or Nef)
5. Curse of Recklessness

The following will be on those fights instead
5. Thunderclap / Thunderfury

Leaving 2 debuff slots for all the following, all of which WILL knock each other off
Spell Vulnerability (Nightfall)
Frost Vulnerability (Winter Chill)
Shadow Vulnerability (ISB)
Expose Weakness (Hunter 8/8 T2)
Taunt (3 BWL drakes)

Now ive never raided as an alliance toon, so i do not know the importance or fights in which there would be a judgment of wisdom or light, but those are on priority with the first 5, knocking all below it off.

Taking all that into consideration, ISB will realistically be up slightly more than 50% of the time. Even at 15% crit over 12 seconds, 4 SB casts will have > 50% chance to apply and have ISB available. Whereas Winter's Chill will always be applied with a mage casting FB

Again, you can't crit for damage if you don't hit.

Guybrush wrote:This is for a shadow bolt only rotation of course, if we add corruption to the rotation then hit becomes better than crit in pretty much every scenario.

yes, yes, 100% yes

I think you're missunderstanding the quote you linked.

It says that 5% critrate is accounting for misses too. This means that out of 100 casts 5 of them will be crits. Therefore, %crit is not diminished by %miss. In effect crits cannot miss.

I have already created a level 1, gotten ony buff and fired shadowbolts at a level 60 character (and a variety of other levels) to see if that was indeed the case. It was not.
After 100 shadowbolts fired at a level 60 (from a level 1), 1 of them were hits and 1 of them were crits. It could be blind luck that one of them actually crit, but if the hitcheck is applied before crits, then that one crit would be absurdly lucky. Because of this I've been wondering about some kind of cap. If crits can't miss, then obviously there would've been no hit, but more crits instead.

I still haven't reached any real conclussion, as the data I am capable of collecting myself will never be enough to show anything concrete. You'd think that shooting at a level 60 as a level 1 with high critchance would end the argument, but it didn't. It just left me with more questions.

Here's the raw data:

(In case it wasn't obvious, all 100 casts aren't shown. They don't fit in a single screenshot. Lmk if you want the excel file)


As you can see, the critrate does indeed decrease as the leveldifference grows, but not at the pace you'd expect and not as regularly as the number of hits. Statistically, 100 shadowbolt casts aren't enough to say anything and that's why I haven't changed my spreadsheet yet.

By the way, the links from the original post are too recent. As far as I can tell they're from post 2.0, in which case they're discussing TBC mechanics. Which are different.

From a strictly DPS perspective, crit will always be superior to hit for warlocks, because of Improved Shadow Bolt. This superiority decreases when you start dotting, ofcourse, but not to a point where hit becomes better.

Botmaster5 wrote:
Source:

Info on crit and hit chances | 2006-08-25 09:54 | Aeus
The way WoW calculates crit rate is over ALL attacks. Crit rate is not based on hits only. In other words, if you have a 5% crit rate, that 5% chance includes misses.

I'm not 100% certain, because this is taken out of context with no link/citation to the original post, but I'm almost certain that blue post is in reference to melee attacks (specifically auto attacks).

Even worse, the scenarios you gave actually contradict your citation. In your scenarios, your crit IS dependent upon your hit, because you're multiplying your chance to hit by your chance to crit (two dice rolls) where the blue post clearly says "crit rate is not based on hits only" (meaning it is a one dice roll where 1-15 is a crit 16-83 is a hit and 84-100 is a miss).

In any case, even though you quoted that post out of context AND misinterpreted it, the arithmetic in your scenarios doesn't make any sense either.

Scenario 1k:
v1 base: 1000 shadowbolts (15% crit, 17% miss)
830 hits, 170 misses
830 * 0.1245 = 103.3 crits
(830 - 103.3) * 1001.8 = 728,008.0 (damage from non-crits)
(103.3 * 1001.8) * 2 = 206,971.9 (Ruin empowered crits)
Total damage = 934,979.9 damage
*emphasis added

Where are you getting this "0.1245 number? I thought you had a 15% chance to crit. I'm assuming its coming from when you say:

If you have a 15% chance to crit but a 17% chance to miss, you will only crit with 12.45% of your shadow bolts,

But if you're doing this, you're counting the miss chance twice. Why would you apply that number when you've already said 170 of the casts were misses? It makes no sense. You've already calculated the miss chance, now you just multiply the raw crit chance since the miss was already factored in. If you wanted to use that number you should have just applied that to the total number of casts.

For example:

1000 casts
170 misses = (1000*.17)
830 hits = (1000*.83)
124.5 crits = (1000*.1245 OR 830*.15)

This is actually a really interesting thread. I'd assume that more or less whatever consensus is found in this thread would also apply to mages?

Not at all. As i wrote earlier, frost mages scale linear with all 3 dmg stats. Meaning if you have x% hit and y% crit, increasing your spelldmg from 300 to 400 gives you the same total dps gain as increasing your spelldmg from 600 to 700 with same hit and critchance does and its the same for crit and hit. Thats why its obvious that hit is always better than crit for frost mages because of the 2 roll system.
Its way more difficult to calculate for warlocks. spelldmg not only increases the dmg of your spells, but also reduces the number of life tabs you have to cast per shadowbolt, so spelldmg basically also reduces your casttime. hit and crit increase the everage dmg of our casts aswell, but also the uptime of our 20% dmg multiplier.
Its even more dificult for firemages. For them, the value of each stat depends on the number of mages in their raid, what stats they have, what spells they cast and even what ping they have.
So you have to determine the value of each stat for every class/specc seperatly.

Wosser wrote:Not at all. As i wrote earlier, frost mages scale linear with all 3 dmg stats. Meaning if you have x% hit and y% crit, increasing your spelldmg from 300 to 400 gives you the same total dps gain as increasing your spelldmg from 600 to 700 with same hit and critchance does and its the same for crit and hit. Thats why its obvious that hit is always better than crit for frost mages because of the 2 roll system.
Its way more difficult to calculate for warlocks. spelldmg not only increases the dmg of your spells, but also reduces the number of life tabs you have to cast per shadowbolt, so spelldmg basically also reduces your casttime. hit and crit increase the everage dmg of our casts aswell, but also the uptime of our 20% dmg multiplier.
Its even more dificult for firemages. For them, the value of each stat depends on the number of mages in their raid, what stats they have, what spells they cast and even what ping they have.
So you have to determine the value of each stat for every class/specc seperatly.

OPs math is ignoring ISB and Life Tap so its actually more useful for mages than locks, once he fixes the double miss bug.

Guybrush wrote:OPs math is ignoring ISB and Life Tap so its actually more useful for mages than locks, once he fixes the double miss bug.

im not accounting for double miss.
1000 shadowbolts
15% crit over all 1000
17%miss
= 12.45% of all LANDED shadow bolts are crits, 2.55% are rolled into your misses
830 landed SBs
830 * 0.1235 = # of landed crits

my math is sound for base variables (hit, miss, crit, damage). ISB, lifetap, spell rotation, is NOT taken into consideration, this is a pure +hit vs +crit debate

diogenes wrote:I'm not 100% certain, because this is taken out of context with no link/citation to the original post, but I'm almost certain that blue post is in reference to melee attacks (specifically auto attacks).

http://wowwiki.wikia.com/wiki/Spell_hit -> "The effect of Hit chance on Critical Hit chance"

Overall chance to crit over all spells cast is thus affected by hit rate. To calculate overall crit rate, multiplying the two chances together:

Code: Select all

crit rate over all spell casts = crit * hit

For example, a caster with no spell hit rating gear or talents, against a mob 3 levels higher (83% hit chance), and 30% crit rating from gear and talents:

Code: Select all

crit rate over all spell casts = 30% * 83% = 24.9%

In addition, direct damage spells suffer from partial resistance, but again, that has no effect on whether a spell hits or not.

im currently on the road, but will be home tomorrow to continue the debate.