Nite_Moogle

Updated to reflect the recent Mage changes.

 

Toskk

In regards to glancing blows.. I would very much like to clear up the "~25% glancing rate" estimate with a hard formula on how glancing blow chance is calculated. In particular, I've seen suggestions that the glancing blow rate might well be calculated at 25.00% of landed blows (anything that isn't miss/dodge/parry) rather than 25.00% of all attacks. Does anyone happen to have any really good data on glancing blow rate, or would be willing to put it to the test?

Along those same lines, I have heard a number of suggestions that recorded crit rates for yellow and/or white attacks are falling considerably lower than the character sheet values indicate.. and again I do not have solid enough data on this topic.

Unfortunately, an accurate test on either of these requires:

1). only data against raid level bosses (no adds can be counted).
2). an extreme duration of the fight.. on the order of 30 straight minutes.
3). recorded from behind the target (to eliminate parries, which are another unknown chance)
4). a static hit, crit, and expertise value. These need to remain constant through the entire fight recording time.

Does anyone happen to know how to engineer such a fight, or better yet have data?

 

• Aldriana

I suspect the way to manufacture such a thing involves doing something akin to what was done in a prior weapon skill thread: take a tank, a melee dps, and a healer, and beat away at the snake boss in ZG - he heals himself, so won't die quickly (eventually he runs out of mana, but I believe the pallies took almost 50 minutes to get to that point).

However, I'd be utterly shocked if it turns out to be 25% of hits, as that flies in the face of the well established and demonstrated 1-roll theory that we've been using for well over a year.

 

Toskk

Yes, I agree that a 2roll mechanic for white attacks and glances is perhaps a bit improbable, however I've definitely seen enough strangeness (even statistically significant strangeness) in combat logs to warrant some testing. I'd heard that the snake boss in ZG was probably the best boss to test against I don't suppose those pallies left their combat log online somewhere?

 

• Aldriana

Thread is [Rogue/Warrior] Weapon Skill Adjustment Discussion and the summary of paladin results is in post 64. Data isn't exactly what you were looking for, but it's something to look at for starters.

 

Drundia

My own experiments on this issue currently indicate that spell requests stay in short queue server-side for about 0.2-0.5 seconds. For me such "no delay" between spells averaged to about 6 ms. Of course sometimes such chaincasting fails and between those spells we have pre-2.3-like situation, this however happens very rarely and I'm button spammer (this is subjective as cast time and latency play a big role here, whether you will usually send request at good time, or it will expire right before spell actually completes).

Due to presence of this queue it should fairly easy to determine a good point to press button for next spell, something around 0.25+latency sec before standard bar fills should be good enough (considering that request is registered when you release button). Quartz can assist in determining this point, but it should be fairly easy to determine through some practice even without having cast bar timer.

 

◊ Wraithlin

To pick up a question based on a BB thread.
Is there any listed testing on the armor equation ; particularly with mobs of a higher level.
Ive done a few forum serachs but I cannot find on these forums any actual testing. Are we still basing the equation off paper-doll numbers ?

 

Nite_Moogle

This has been established for so long that there has been no recent testing on it, but all number-running using the formula has proven it to be correct. I'm reasonably sure this was nailed down in the Patchwerk era when people determined how much AC was needed to eat hatefuls in combination with a certain gear set.

Even a bear with 30,000 AC will take damage from level 1 mobs in newbie zones which means you cannot achieve damage immunity from AC, so an AC cap must exist.

 

darrenan

Isn't it more accurate to say 2/45/11+3 or 10/45/3+3? Technically speaking, Blazing Speed and Flame Throwing don't factor into DPS. I realize that taking points in Flame Throwing may result in less movement, thus higher DPS, but I'm guessing that most theorycrafting doesn't take that into account. I actually have an unorthodox 48, where I only have 1 point into Flame Throwing, and 2 points into Imp. Flamestrike. I may respec out of that at some point, but right now there are enough AoE pulls in SSC, TK and MH to make those 2 points worthwhile, IMO.

 

◊ Wraithlin

The only establishment Ive ever seen is where people made graphs from values in the paperdoll; certainly the first time I ever encountered the armor equation someone had plotted paperdoll values and then determined the equation.
Which would mean there are at least three if not more valid interpretations; enemy level, player level, squareroot(player level^2+enemylevel^2). Ive tried digging around the last few days but I just cant nail down any testing, only examples where people took the paperdolly information and fitted the data for level X against level X.

My google foo is weak, but its nagging at me that I cannot find any numbers where people tested against various mob levels at various player levels.

[e]
To pick up on your last point, why hasnt anyone taken a 30k+ armor bear and found out how much a level 20 mob hits them for, if its 1 point on a hit then the amor cap would be well above 75% against a level 20.

[ee]
Im going to pester some of my guildies and try to gather some numbers over the next week or two.

 

Nite_Moogle

The values from the paper doll have matched up with the reduction amounts seen vs equal level NPCs, but the reduction percentages do not match when you are dealing with Boss mobs taking hits with known values (like Hateful Strike). Again, this goes back to Patchwerk era where people swore up and down that they had the HP to survive a hit with their current AC and they did not because they had not accounted for the level discrepancy. It can therefore be concluded that the AC value that you have is scaled against the level of the NPC, and the forumlas have been subsequently derived or datamined.

They have. It's 75%. You can do this to satisfy your own curiosity but I an assure you that you will end up with 75% for an AC cap and you will end up with the same values for damage reduction.

 

Drundia

I believe Blazing Speed isn't in that build, however what you say makes some sence. We can easily drop points from Blast Wave. If we also drop points from Flame Throwing we come to 43 points in Fire, or we can drop Improved Scorch (relying on other Mages to keep it...) coming to 42 points in Fire and then we can put as many as 2 or 3 points in Arcane Concentration, leading to 7/43/11 or 8/42/11 which might be better than that "2/47/11+1". Time to visit Mage thread?

 

Hate Monkey

In the same breath we can also say that Pyroblast and Blast Wave are not required for dps boosting.

But for all realistic purposes if you're in a raiding guild are are progressing past the SSC/TK level, the 2 points into arcane become worthless as well as a 0/47/14 has all the same dps talents that 2/47/11+1 has but it picks up Imp Blizzard+Frostbite+2/3 Permafrost for extra control in Hyjal and beyond.

 

tholex

DPS aka Ret paladins receive about as much of a benefit from WF as 2H warriors/ DW Fury warriors, and should probably be considered to be 2x for the calculation of when to throw windfury totem under the totem section. See the ret pally thread for evidence.

 

• Binkenstein

Moogle, may want to check the non-removable spell resist part.
Using formulae from wowwiki, I get a 3% mitigation figure overall, which is approx the figure I get on WWS.

 

Roywyn

Hm, it seems to be around 5-6% for me.
Did you account for the odd display of partial resists, that is doesn't account for damage debuffs or crits?

I.e. if you have a lightning bolt that hits for 2k and you get a 50% partial, it may look like this:
2000 * 0.5 (partial) * 2.09 (crit) * 1.2 (stormstrike) * 1.05 (misery) = 2633 damage.

The log will display "Your Lightning Bolt crits Mob X for 2633 (1000 resisted).", which is a 1000/(1000+2633) = 27.5% partial resists in WWS, although it actually is a 50% resist.
Reports for shaman with high crit rates at 209% crit damage would be particularly skewed against WWS logs.

Having a level 65 wand down Dr. Boom or having a holy priest put SW:P on Venoxis seem to be the more reasonable methods of testing I guess ...

 

clavarnway

I didn't see it when I read it...2.4, spell haste reduces spell global cooldown, down to 1 second if possible.

http://sigs.guildlaunch.net/sig.php/1424974uXyta.png

 

Nite_Moogle

Did the topic name change to say 2.4? No? Okay. 2.4 changes will be added in at some point soon. The patch isn't even up on the European servers.

 

Allev

Question about the expertise cap.

The current first-post has the value of expertise rating cap (melee chance to be dodged) at 23 expertise (aka 5.75% dodge). I've been looking for an authoritative answer on this value for awhile; rogues have it at 6.5%, warriors have it at 5.6%, shamans have 6.25%. Is there an authoritative accepted value? Does this need its own thread?

 

Chew

Regarding Mage - Polymorph and Spell Hit Rating.

This is what I have read. Please correct or confirm this.

Spell Hit is taken into the equation when the Polymorph spell is cast upon the target.
Then, if successful, every 2 seconds a new check is performed. (Which again the Mages Spell hit is taken into account.)
To see if the target will break or stay sheeped.

Thanks.