Working Theories of Theorycrafting as of 2.1 - Page 7

MeCh · 08/11/07, 3:52 AM

You still need to update the windfury nerf.

◊ manly · 08/13/07, 5:25 AM

Originally Posted by Hate Monkey

I know that this is widely believed, but it's been bugging me ever since people stated it to be true. Been going through LOTS of WWS reports for various fights, mages only, and for every Fire mage with 30 penetration, I'm seeing ~3% partial resist rate. But for mages with higher than 30 penetration, mainly 35-50, I see 1% or less partial resist rate. Going to make a compiled list here soon of about 50-100 mages for each fight that doesn't "violate" the rule.

That would be a good thing, since it always hurt to look at the [Mantle of the Tempest] and keep grinning at the -18 resistance (old term for +18 penetration).

In all seriousness, there were a few theories back in the ZG-era that binary spells single miss/resist roll could be 'abused' in such a way that excessive hit ratings could overcome the innate resists. However, I remember having never seen any data proving it. As such, perhaps the innate resist can be overcome on binary spells, possibly even with spell penetration.

Gruntle · 08/13/07, 9:05 AM

You may want to update the miss rates, based on the testing in the weapon skill testing thread it seems very likely that the miss rate is considerably higher without weapon skill (check that thread for details, particularly post 318). The effect of weapon skill on misses is still not sorted out completely, but is likely nonlinear.

Hate Monkey · 08/13/07, 3:11 PM

Originally Posted by manly

That would be a good thing, since it always hurt to look at the [Mantle of the Tempest] and keep grinning at the -18 resistance (old term for +18 penetration).

In all seriousness, there were a few theories back in the ZG-era that binary spells single miss/resist roll could be 'abused' in such a way that excessive hit ratings could overcome the innate resists. However, I remember having never seen any data proving it. As such, perhaps the innate resist can be overcome on binary spells, possibly even with spell penetration.

THIS IS FOR MAGES ONLY!

Been trying and trying to find a decent way to set this up so it's in a readable format, but just due to time of reports and how long has passed since then, it's to much work to try and keep it updated.

But as with what I am able to see with WWS logs and fights that don't have weird resists to them, this is what a general rule I'm seeing:
All assuming CoE/S and a PvE spec, not PvP.
0 Spell Penetration = ~8% of total damage lost to partial resists
10 Spell Penetration = ~6% of total damage lost to partial resists
30 Spell Penetration = ~4% of total damage lost to partial resists
40 Spell Penetration = ~2% of total damage lost to partial resists

Didn't see a single mage with only 20 penetration.

Of my searches I'm unable to see anyone with higher than 40 spell penetration.

One thing to note is that I ran across people with 32-38 spell penetration who did not see a different level of partial resists than that of 30 spell penetration people.

Pheroz · 08/13/07, 3:53 PM

Hate,

Can we move the partial resist/penetration theories to a different thread? We really shouldn't muck up this thread with stuff that, as of this point in time, isn't the generally accepted theory.

Hate Monkey · 08/13/07, 4:42 PM

Originally Posted by Pheroz

Hate,

Can we move the partial resist/penetration theories to a different thread? We really shouldn't muck up this thread with stuff that, as of this point in time, isn't the generally accepted theory.

Well with as little of time as I've spent on that, I just proved the "You can't get lower than 5% on partial resists" bunk, so that is important to this thread.

tetracycloide · 08/13/07, 5:02 PM

Originally Posted by Hate Monkey

Well with as little of time as I've spent on that, I just proved the "You can't get lower than 5% on partial resists" bunk, so that is important to this thread.

Going to need to back up this statment, preferably in another thread, with a full breakout of the data sets you looked at and the math you used to build the confidance intervals necessary to say this to a degree of certainty 90% or greater. Looking up a few parses and eyeballing the trends and cross referencing that with armory stats that may not have been accurate at that point hardly constitues proof, just cause for further investegation would be a more accurate description.

Hate Monkey · 08/13/07, 5:04 PM

Originally Posted by tetracycloide

Going to need to back up this statment, preferably in another thread, with a full breakout of the data sets you looked at and the math you used to build the confidance intervals necessary to say this to a degree of certainty 90% or greater. Looking up a few parses and eyeballing the trends and cross referencing that with armory stats that may not have been accurate at that point hardly constitues proof, just cause for further investegation would be a more accurate description.

I know that, thats why I limited my original post. You can also go back to the previous page too and look at my post there about it.

But due to some logs not staying up on WWS for a long time, and gear upgrades/changes along with spec changes, it's a lot harder to do than you would think.

Pheroz · 08/13/07, 5:46 PM

Originally Posted by Hate Monkey

Well with as little of time as I've spent on that, I just proved the "You can't get lower than 5% on partial resists" bunk, so that is important to this thread.

It's really not. That's just statistics and probability in action. And it also doesn't mean there's a correlation between spell penetration and partial resists.

Comparing 4 fire mages from my guilds last night Gruul kill, we see a horrible jumble of information that is not conclusive and can be misinterpretted lots of ways becuase of bias.

Mage      |  # of fireballs (hits + crits)   | Spell Penetration   |  Mitigated %
Pheroz                   53                                   10                     2.7%
Karha                    54                                   33                     4.0%
Kelthus                  55                                    30                     1.2%
Hypnotiq                 64                                    53                     4.7%

Obviously not even close to a statistically significant sample. But thats the kind of thing your looking it. If you want to make blanket generalizations from eyeballing series of data like that, people are going to want to see your back up and your methods. Without that, your just challenging conventional wisdom without much ground to stand on.

It certainly merits discusiion, but this thread is a place where conclusions are far more valuable then the specific discussions, and I think we're pretty far from a conclusion right now.

Hate Monkey · 08/13/07, 5:59 PM

Originally Posted by Pheroz

Obviously not even close to a statistically significant sample. But thats the kind of thing your looking it. If you want to make blanket generalizations from eyeballing series of data like that, people are going to want to see your back up and your methods. Without that, your just challenging conventional wisdom without much ground to stand on.

It certainly merits discusiion, but this thread is a place where conclusions are far more valuable then the specific discussions, and I think we're pretty far from a conclusion right now.

Well see thats the way you're doing it, but you don't know how I'm doing it. I posted a general information header that so far I've only looked at mages that are in PvE gear and PvE spec and having CoE/S. Now you just went and pulled the numbers from a WWS parse, but didn't bother with looking at specs, debuffs on the mob, or gear the mage is wearing.

I've looked through enough logs so far that I feel comfortable in saying that the 5% rule is bunk, and it'll still be time before I can compile a decent list of information and present it for public scrutiny.

Just stop replying saying what I'm doing is wrong when you have no idea what I'm actually doing. I'll make a nice new thread with discussion and all that when I can get all the information lined up nicely.

Zurgat · 08/14/07, 8:37 AM

On the subject of elemental mobs / bosses and +- attack power. (Such as Hydross)

Melee attacks made by elementals
* Can't be mitigated by armor. (Plate or leather makes no difference)
* Can't be blocked. (Only use a shield to make threat, or for stats)
* Can be parried.
* Can be dodged.
* Can be partially resisted with resist gear.
* Attack power modification debuffs "do" affect their damage.

Nite_Moogle · 08/14/07, 10:19 AM

Originally Posted by Zurgat

On the subject of elemental mobs / bosses and +- attack power. (Such as Hydross)

Good addition.

Foofu · 08/14/07, 1:05 PM

Originally Posted by Hate Monkey

Well see thats the way you're doing it, but you don't know how I'm doing it. I posted a general information header that so far I've only looked at mages that are in PvE gear and PvE spec and having CoE/S. Now you just went and pulled the numbers from a WWS parse, but didn't bother with looking at specs, debuffs on the mob, or gear the mage is wearing.

I've looked through enough logs so far that I feel comfortable in saying that the 5% rule is bunk, and it'll still be time before I can compile a decent list of information and present it for public scrutiny.

Just stop replying saying what I'm doing is wrong when you have no idea what I'm actually doing. I'll make a nice new thread with discussion and all that when I can get all the information lined up nicely.

The point is no-one here knows exactly what you did, so it's hard for us to back up your findings and sticky them in the "tested and accepted mechanics" thread. You need to disclose all your data and the math you used to draw your conclusions so peer review can happen so we can all get behind you and accept this as proven, and that needs to happen in a separate thread so Moogle can link to your detailed findings, and so anyone trying to debunk them doesn't muck up this thread.

Zurgat · 08/16/07, 10:29 AM

I'm curious about this, on the subject melee :
- 1h/Shield, 2h: 8.64% [137 hit rating]

Does this include druids in cat/bearform and players wearing "no" weapons ?
Dual fists counts as dual wielding(unarmed skill), but when unequipping both weapons, are you considered to be using only a 1h attack, or is this dependant on class ?

Nite_Moogle · 08/16/07, 11:52 AM

Feral attacks are considered to be one-handed. Not using a weapon is somewhat irrelevant :P

sp00n · 08/17/07, 12:45 PM

Originally Posted by Zurgat

I'm curious about this, on the subject melee :
- 1h/Shield, 2h: 8.64% [137 hit rating]

Does this include druids in cat/bearform and players wearing "no" weapons ?
Dual fists counts as dual wielding(unarmed skill), but when unequipping both weapons, are you considered to be using only a 1h attack, or is this dependant on class ?

Moreover it's outdated.
Current findings show that the miss % is around 9% (and clearly above 8.64%) against a level 73 mob and respectively 28% for dual wielding.
We haven't figured it out yet exactly, but it strongly points to 9%.

Any additional point in weapon skill above 350 changes this value, with a so far still to to test change when you hit 355 skill (unknown before, 0.1% hit per point of weapon skill after).

Zurgat · 08/21/07, 8:57 AM

Found this bit about hunters & their pets which might be useful.
I don't see any info about warlock pets, but i suspect it follows a similar method.

Any use for ilvl calculations? http://www.wowwiki.com/Ilvl

Hunter pet scaling & info
Source : HunterGuide » Theorycraft

Pets get about 30% of the hunter's stamina added to their stamina.
Pets get about 35% of the hunter's armor added to their armor.
Pets get about 22% of the hunter's ranged attack power added to their melee attack power.
Pets get about 12.5% of the hunter's ranged attack power added to their spell damage.
Pets get about 40% of the hunter's resistances added to their own resistances.

Exact stats vary slightly based on pet family.
All scaling is based on total stats (after items, buffs, etc) of the hunter.

Training Points
Your pet's Training Points are based on two things, Loyalty Level and Pet Level. To get more Training Points, you have to increase one of these variables.

TrainingPoints = PetLevel * (LoyaltyLevel - 1)

Info bits
Base focus regen rate = 24 focus every 4 seconds.

Ferocious Inspiration stacks from multiple hunters in the same group.

Expose Weakness does not stack from multiple hunters attacking the same target. The most powerful version takes precedence.

alkis · 08/22/07, 12:03 AM

One thing to add is that the equipped weapon's skill affects the hit, miss, parry, dodge chance of shield related abilities on warriors. Not sure if it this is obvious to everyone but this is quite bad when you get disarmed and you have low unarmed skill and you pretty much can't shield bash or slam.

Gruntle · 08/22/07, 4:56 AM

Originally Posted by alkis

One thing to add is that the equipped weapon's skill affects the hit, miss, parry, dodge chance of shield related abilities on warriors. Not sure if it this is obvious to everyone but this is quite bad when you get disarmed and you have low unarmed skill and you pretty much can't shield bash or slam.

Problem is that we actually don't really know how that works yet. Tests are being reported and discussed in the [Rogue/Warrior] Weapon Skill Adjustment Discussion thread. The main result from the testing done so far is that the base miss chance is 9% (95% confidence intervals at [8.99,9.13]) or 28% for DW. Weapon skill affects the miss rate nonlinearly but it's not clear exactly how.

◊ Lactose · 08/22/07, 6:05 AM

Originally Posted by Gruntle

Problem is that we actually don't really know how that works yet. Tests are being reported and discussed in the [Rogue/Warrior] Weapon Skill Adjustment Discussion thread. The main result from the testing done so far is that the base miss chance is 9% (95% confidence intervals at [8.99,9.13]) or 28% for DW. Weapon skill affects the miss rate nonlinearly but it's not clear exactly how.

What he means is that, since there is no 'Shield Weapon Skill', abilities that require a shield uses your currently equipped weapon's hit/crit/miss calculations, instead of e.g. [Level*5] Shield Weapon Skill.

alkis · 08/22/07, 11:18 AM

Originally Posted by Gruntle

Problem is that we actually don't really know how that works yet. Tests are being reported and discussed in the [Rogue/Warrior] Weapon Skill Adjustment Discussion thread. The main result from the testing done so far is that the base miss chance is 9% (95% confidence intervals at [8.99,9.13]) or 28% for DW. Weapon skill affects the miss rate nonlinearly but it's not clear exactly how.

You don't have to specify a number. Just the fact that your unarmed skill is < 350 means that when you are disarmed your shield slams and shield bashes are going to hit and crit less often and be parried more often. And also as vice versa: if you have a +skill rating in your gear it will affect your shield slams and shield bashes ([Mallet of the Tides] for example). We may not know the exact formula but we know shield related abilities are affected by it.

Edit: I need to learn to refresh the browser before I submit. Lactose pointed the same thing :-)

Mikari · 08/22/07, 12:47 PM

Originally Posted by Zurgat

Found this bit about hunters & their pets which might be useful.
I don't see any info about warlock pets, but i suspect it follows a similar method.

The blizzard UI files lists pet scaling as follows:

--Pet scaling:
HUNTER_PET_BONUS = {};
HUNTER_PET_BONUS["PET_BONUS_RAP_TO_AP"] = 0.22;
HUNTER_PET_BONUS["PET_BONUS_RAP_TO_SPELLDMG"] = 0.1287;
HUNTER_PET_BONUS["PET_BONUS_STAM"] = 0.3;
HUNTER_PET_BONUS["PET_BONUS_RES"] = 0.4;
HUNTER_PET_BONUS["PET_BONUS_ARMOR"] = 0.35;
HUNTER_PET_BONUS["PET_BONUS_SPELLDMG_TO_SPELLDMG"] = 0.0;
HUNTER_PET_BONUS["PET_BONUS_SPELLDMG_TO_AP"] = 0.0;
HUNTER_PET_BONUS["PET_BONUS_INT"] = 0.0;

WARLOCK_PET_BONUS = {};
WARLOCK_PET_BONUS["PET_BONUS_RAP_TO_AP"] = 0.0;
WARLOCK_PET_BONUS["PET_BONUS_RAP_TO_SPELLDMG"] = 0.0;
WARLOCK_PET_BONUS["PET_BONUS_STAM"] = 0.3;
WARLOCK_PET_BONUS["PET_BONUS_RES"] = 0.4;
WARLOCK_PET_BONUS["PET_BONUS_ARMOR"] = 0.35;
WARLOCK_PET_BONUS["PET_BONUS_SPELLDMG_TO_SPELLDMG"] = 0.15;
WARLOCK_PET_BONUS["PET_BONUS_SPELLDMG_TO_AP"] = 0.57;
WARLOCK_PET_BONUS["PET_BONUS_INT"] = 0.3;

They just need converting to percentages, Hunter ones are basically the same as what you posted though.

Gruntle · 08/23/07, 5:30 AM

Originally Posted by alkis

You don't have to specify a number. Just the fact that your unarmed skill is < 350 means that when you are disarmed your shield slams and shield bashes are going to hit and crit less often and be parried more often. And also as vice versa: if you have a +skill rating in your gear it will affect your shield slams and shield bashes ([Mallet of the Tides] for example). We may not know the exact formula but we know shield related abilities are affected by it.

Edit: I need to learn to refresh the browser before I submit. Lactose pointed the same thing :-)

Yeah, sorry. I need to learn to read a post properly before replying

Athinira · 08/26/07, 11:58 AM

Priests:
Mind Flay is in reality a channeled DoT (Damage over Time) spell and not a direct damage spell. Shaman Totem's are immune to it and it doesn't break a druids Maim either.

koaschten · 08/26/07, 2:46 PM

Originally Posted by Athinira

Priests:
Mind Flay is in reality a channeled DoT (Damage over Time) spell and not a direct damage spell. Shaman Totem's are immune to it and it doesn't break a druids Maim either.

In addition to this, a guildmate of mine figured out that indeed Mindflay channels faster, but over a fixed amount of time will generate the same number of ticks.

Talk about buffing caster haste rating ... more mana ... same dps.