Deep Dive on Stats for a Mag Encounter

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Deep Dive on Stats for a Mag Encounter

Postby Coristad » Tue Aug 28, 2007 10:00 am

Hi all,

My guild recently went into Mag's lair. Our raid leader used WoW web stats and collected a ton of useful data. The problem, from my perspective, is doing the analysis and understanding what I need to do to fix some things. My GL is supportive of my tanking, but I know I'm not doing as good of a job as I could do. I don't have anyone in the guild I can crunch with, so I'd like some input from some more senior tankadins.

First, the caveats:

1. I may or may not be logged out with tanking gear. Raid buffed for the encounter I had 14K health, 14.5K armor, 490 Defense.
2. Our raid was collectively undergeared, which we acknowledged. I myself am not uncrushable (3 badges away arrrgh). I'm less worried about the damage in portion of the report versus the dame out portion.

Now the report:
http://www.lossendil.com/wws/?report=vhzwr5pooqlvy&a=15

Now the head scratching parts:

1. White damage was 25% of my total, and the largest single source. I missed 24.6% of the time <!>. Of the 380 Hits, 87 were from reckoning. Yet Seal of Righteousness only hit 205 times with a miss rate of 19.2%. That's effectively 1 in 5. What's up with that?

2. Holy Shield was only 11% of my damage but if I did the math right x2.35 = 25.85% of my threat.

I'm still trying to pick out useful information, but I wanted to throw this out there. I'd love some constructive suggestions on what conclusions to draw from this data, if any.

Thanks!
Coristad
 
Posts: 105
Joined: Wed May 09, 2007 11:46 am
Location: Kul Tiras

Re: Deep Dive on Stats for a Mag Encounter

Postby jere » Tue Aug 28, 2007 10:51 am

Coristad wrote:Hi all,

My guild recently went into Mag's lair. Our raid leader used WoW web stats and collected a ton of useful data. The problem, from my perspective, is doing the analysis and understanding what I need to do to fix some things. My GL is supportive of my tanking, but I know I'm not doing as good of a job as I could do. I don't have anyone in the guild I can crunch with, so I'd like some input from some more senior tankadins.

First, the caveats:

1. I may or may not be logged out with tanking gear. Raid buffed for the encounter I had 14K health, 14.5K armor, 490 Defense.
2. Our raid was collectively undergeared, which we acknowledged. I myself am not uncrushable (3 badges away arrrgh). I'm less worried about the damage in portion of the report versus the dame out portion.

Now the report:
http://www.lossendil.com/wws/?report=vhzwr5pooqlvy&a=15

Now the head scratching parts:

1. White damage was 25% of my total, and the largest single source. I missed 24.6% of the time <!>. Of the 380 Hits, 87 were from reckoning. Yet Seal of Righteousness only hit 205 times with a miss rate of 19.2%. That's effectively 1 in 5. What's up with that?

2. Holy Shield was only 11% of my damage but if I did the math right x2.35 = 25.85% of my threat.

I'm still trying to pick out useful information, but I wanted to throw this out there. I'd love some constructive suggestions on what conclusions to draw from this data, if any.

Thanks!


Just some notes:
1. You definitely need to upgrade your gear as you surmised
2. You are tanking the channelers, which are casters. You will have less Holy Shield charges on them as they often stop melee to cast.
3. Spells have a 17% chance to be resisted against boss mobs (including the channelers)
4. The miss rate you are quoting (25%) is your "All Miss" value, which is the sum of dodge+parry+miss. You have to expand the chart to see the individual. Your miss rate was really 4%

Hopefully that helps with some of the questions.
Image
User avatar
jere
 
Posts: 2960
Joined: Fri May 11, 2007 5:12 pm

Re: Deep Dive on Stats for a Mag Encounter

Postby Coristad » Tue Aug 28, 2007 1:51 pm

jere wrote:
Just some notes:
1. You definitely need to upgrade your gear as you surmised
2. You are tanking the channelers, which are casters. You will have less Holy Shield charges on them as they often stop melee to cast.
3. Spells have a 17% chance to be resisted against boss mobs (including the channelers)
4. The miss rate you are quoting (25%) is your "All Miss" value, which is the sum of dodge+parry+miss. You have to expand the chart to see the individual. Your miss rate was really 4%

Hopefully that helps with some of the questions.


Thanks for the response. Questions lead to more questions...

First, is it worth trying to gear for +spell hit? I would be inclined to say no as threat generation is not really a problem with tankadins, generally speaking.

We're also getting back to the usefulness of Weapons Expertise and tying them to seals. If a seal can only proc on a hit, can you expect a linear increase in the amount of threat generated?

An interesting (albeit costly) experiment would be to do a series of runs, swapping out a point from 1H Spec to Weapons Expertise after each run to see what that did to net threat overall.
Coristad
 
Posts: 105
Joined: Wed May 09, 2007 11:46 am
Location: Kul Tiras

Re: Deep Dive on Stats for a Mag Encounter

Postby DeadMilliken » Tue Aug 28, 2007 3:14 pm

Coristad wrote:
jere wrote:
Just some notes:
1. You definitely need to upgrade your gear as you surmised
2. You are tanking the channelers, which are casters. You will have less Holy Shield charges on them as they often stop melee to cast.
3. Spells have a 17% chance to be resisted against boss mobs (including the channelers)
4. The miss rate you are quoting (25%) is your "All Miss" value, which is the sum of dodge+parry+miss. You have to expand the chart to see the individual. Your miss rate was really 4%

Hopefully that helps with some of the questions.


Thanks for the response. Questions lead to more questions...

First, is it worth trying to gear for +spell hit? I would be inclined to say no as threat generation is not really a problem with tankadins, generally speaking.

We're also getting back to the usefulness of Weapons Expertise and tying them to seals. If a seal can only proc on a hit, can you expect a linear increase in the amount of threat generated?

An interesting (albeit costly) experiment would be to do a series of runs, swapping out a point from 1H Spec to Weapons Expertise after each run to see what that did to net threat overall.


Spell hit effects Judgements, Exorcism, and RD...not much else
Melee hit covers, directly or indirectly, swing procs like SoCommand, SoR, SoB, and SoV.

Also note: +weapon rating/Weapon Ex punches through Parry, Dodge, Block, and miss rates on raid mobs.
User avatar
DeadMilliken
 
Posts: 596
Joined: Tue Aug 28, 2007 7:19 am

Re: Deep Dive on Stats for a Mag Encounter

Postby Fridmarr » Tue Aug 28, 2007 10:54 pm

DeadMilliken wrote:Spell hit effects Judgements, Exorcism, and RD...not much else
Melee hit covers, directly or indirectly, swing procs like SoCommand, SoR, SoB, and SoV.

Also note: +weapon rating/Weapon Ex punches through Parry, Dodge, Block, and miss rates on raid mobs.


Wouldn't spell hit also effect Holy Shield, Blessing of Sanc, Ret Aura, or Consecrate tick resists?
Fridmarr
Global Mod
 
Posts: 9667
Joined: Sun Apr 08, 2007 1:03 am

Re: Deep Dive on Stats for a Mag Encounter

Postby adese » Wed Aug 29, 2007 4:13 am

Fridmarr wrote:
DeadMilliken wrote:Spell hit effects Judgements, Exorcism, and RD...not much else
Melee hit covers, directly or indirectly, swing procs like SoCommand, SoR, SoB, and SoV.

Also note: +weapon rating/Weapon Ex punches through Parry, Dodge, Block, and miss rates on raid mobs.


Wouldn't spell hit also effect Holy Shield, Blessing of Sanc, Ret Aura, or Consecrate tick resists?


I'm pretty sure that when they proc, the person that hits you takes damage, period. Well, I guess I have seen Consecrate resist once or twice - usually only on one mob though, and really only on the first tick at that.

So I'd guess that spell hit could help with Consecrate, but I know that HS, BoSanc, and Ret Aura can't be resisted.
Image
Image
adese
 
Posts: 692
Joined: Wed May 09, 2007 7:33 pm
Location: Aggramar

Re: Deep Dive on Stats for a Mag Encounter

Postby Fridmarr » Wed Aug 29, 2007 7:23 am

adese wrote:
Fridmarr wrote:
DeadMilliken wrote:Spell hit effects Judgements, Exorcism, and RD...not much else
Melee hit covers, directly or indirectly, swing procs like SoCommand, SoR, SoB, and SoV.

Also note: +weapon rating/Weapon Ex punches through Parry, Dodge, Block, and miss rates on raid mobs.


Wouldn't spell hit also effect Holy Shield, Blessing of Sanc, Ret Aura, or Consecrate tick resists?


I'm pretty sure that when they proc, the person that hits you takes damage, period. Well, I guess I have seen Consecrate resist once or twice - usually only on one mob though, and really only on the first tick at that.

So I'd guess that spell hit could help with Consecrate, but I know that HS, BoSanc, and Ret Aura can't be resisted.


I'm 100% certain they all get resisted. Some log parsers don't pick up Holy Shield or BoSanc because they are reported backwards in the combat log ie..

Mob resists your Holy Shield

Instead of the more typical

Your Holy Shield is resisted by Mob

Wow web stats suffers from this problem, I submitted a bug report with them on the issue. Anyhow, they do get resisted and it will be in your log, against a level 73 boss, they seem to get resisted at the normal 17% rate.
Fridmarr
Global Mod
 
Posts: 9667
Joined: Sun Apr 08, 2007 1:03 am

Postby jere » Wed Aug 29, 2007 8:01 am

Holy Shield does indeed get resisted. I see it in my logs all the time when tanking prince. It is always a full resist..I have seen no partial resists yet.
Image
User avatar
jere
 
Posts: 2960
Joined: Fri May 11, 2007 5:12 pm

Postby adese » Wed Aug 29, 2007 9:01 am

I stand corrected, then. I've never really looked to see if it does get resisted or not, but since I've never seen a "Resist" scroll past on my screen when a Holy Shield charge gets used, I just assumed it wasn't possible. I guess I just wasn't looking closely enough.
Image
Image
adese
 
Posts: 692
Joined: Wed May 09, 2007 7:33 pm
Location: Aggramar

Postby jere » Wed Aug 29, 2007 11:35 am

Just as an example, I looked at a log from one of my previous Prince kills. I found 12 instances of:

"Prince Malchezaar resists your Holy Shield"
Image
User avatar
jere
 
Posts: 2960
Joined: Fri May 11, 2007 5:12 pm

Postby Coristad » Wed Aug 29, 2007 12:00 pm

More strangeness datamined from the fight... I'd appreciate some perspective from the experienced tanks on this one. This is also cross posted from my guild's warrior class forum--I'll be curious as to what their response is.

So I'm digging through the charts from our last Mag learning raid, looking for points of weakness that I could change to stop Daigan (GL, prot warrior) from his litany of "Oh, looks like Coristad is dead." I was looking at warrior stats and druid stats, to see if there was anything I could glean from them (realizing that my tanking style is closer to a warrior's than a bear's). I expected to see the warriors mitigating or avoiding a large number of attacks, and no crushes. I was expecting to see the bears getting crushed at a reasonably constant rate, but also having relatively high dodge values. Like I said, I, and the majority of tankadins, are pretty obsessed with achieving and maintaining uncrushability- it's one of our major benchmarks to compare against the more established tanking classes.

Here's what I saw (apologies for the formatting):

Code: Select all
Tank Name      Landed  Crush      %Crush  Crush Avg.      Crush Max
Daigan          267      75       0.13       3926           6192
Coristad        201      20       0.05       3965           5439
Alkurena       126       24       0.09       3785           7256
Savaric(*)      83       19       0.12       3968           5830
Griphon        298       80       0.12       2726           8376
Lureanna       192       48       0.12       3245           7939



Daigan- GL, well geared and experienced MT. Coristad (prot pally), Alkurena (prot warrior), Savaric (prot warrior), Griphon & Lureanna- Drood Tanks.

Notes:
1. Landed = total number of blows.
2. %Crush is expressed in decimal values so .13 = 13%.
3. Although Savaric was not present for the entire fight, he received a statistically significant number of hits, so his data was kept. He could also be excluded.

Questions:
1. Why are the warriors getting crushed at approximately the same rate as the druids, who have no CI?

2. While I am personally fixated on CI, that doesn't necessarily mean it's that important. How important is CI to raid tanks?

3. Related to 2, how important is CI to the healing team?

I'm not sure what to make of this data, so I would appreciate any help on interpreting it. Thanks!
Coristad
 
Posts: 105
Joined: Wed May 09, 2007 11:46 am
Location: Kul Tiras

Postby Ills » Wed Aug 29, 2007 12:03 pm

13% CB rate? Sounds like your warriors seldom have shield block up.

Taking CBs adds to the spike damage that the tanks are taking, making them harder to heal (and eats up more mana).
Every fight is like Sartura, only without the screaming and the dying.
Ills
 
Posts: 102
Joined: Sun Jul 08, 2007 2:57 pm

Postby jere » Wed Aug 29, 2007 12:27 pm

I looked at Daigon specifically. He buffs himself with Shield Block 2 times the entire night. Now many will argue that due to the cooldown < duration, you won't see refreshes in the combat log. Fair enough, but if you look closer, the 2 times that it appears in the combat log are for Attempt 5 and Attempt 6, which means for the trash (which can crush -- he was 14 times) and attempts 1-4, he never used Shield Block at all.

Furthermore, on attempts 5 and 6, he received 9-10 crushes, indicating Shield Block was down for some part of the time, so we should have seen more than one buff gain from shield block in both of those attempts. We only see one for each of those 2 attempts.

In short Daigon isn't doing his job if he is tanking. He doesn't even use it until attempts 5 and 6, and there he apparently isn't using it much as he recieves multiple crushes but only buffs it once (refreshes won't show, but if he gets crushed, then it isn't a refresh, but a buff gain).

Which channeler did each tank have again?
Image
User avatar
jere
 
Posts: 2960
Joined: Fri May 11, 2007 5:12 pm

Postby Coristad » Wed Aug 29, 2007 2:19 pm

jere wrote:I looked at Daigon specifically. He buffs himself with Shield Block 2 times the entire night. Now many will argue that due to the cooldown < duration, you won't see refreshes in the combat log. Fair enough, but if you look closer, the 2 times that it appears in the combat log are for Attempt 5 and Attempt 6, which means for the trash (which can crush -- he was 14 times) and attempts 1-4, he never used Shield Block at all.

Furthermore, on attempts 5 and 6, he received 9-10 crushes, indicating Shield Block was down for some part of the time, so we should have seen more than one buff gain from shield block in both of those attempts. We only see one for each of those 2 attempts.

In short Daigon isn't doing his job if he is tanking. He doesn't even use it until attempts 5 and 6, and there he apparently isn't using it much as he recieves multiple crushes but only buffs it once (refreshes won't show, but if he gets crushed, then it isn't a refresh, but a buff gain).

Which channeler did each tank have again?


You see, that was the immediate conclusion I reached, as well. But to paraphrase Dale Carnegie, seek first to understand, then to be understood.

As it turns out, the tanks were deliberately choosing to allow themselves to be hit (and crushed for a comparatively small amount) by the channelers to prevent rage starvation. I honestly had no idea, so I'm glad I asked instead of going with the "hey look the tanks suck" approach I was initially considering ;-). So a little bit of knowledge really is a dangerous thing...
Coristad
 
Posts: 105
Joined: Wed May 09, 2007 11:46 am
Location: Kul Tiras

Postby jere » Wed Aug 29, 2007 2:30 pm

Coristad wrote:
jere wrote:I looked at Daigon specifically. He buffs himself with Shield Block 2 times the entire night. Now many will argue that due to the cooldown < duration, you won't see refreshes in the combat log. Fair enough, but if you look closer, the 2 times that it appears in the combat log are for Attempt 5 and Attempt 6, which means for the trash (which can crush -- he was 14 times) and attempts 1-4, he never used Shield Block at all.

Furthermore, on attempts 5 and 6, he received 9-10 crushes, indicating Shield Block was down for some part of the time, so we should have seen more than one buff gain from shield block in both of those attempts. We only see one for each of those 2 attempts.

In short Daigon isn't doing his job if he is tanking. He doesn't even use it until attempts 5 and 6, and there he apparently isn't using it much as he recieves multiple crushes but only buffs it once (refreshes won't show, but if he gets crushed, then it isn't a refresh, but a buff gain).

Which channeler did each tank have again?


You see, that was the immediate conclusion I reached, as well. But to paraphrase Dale Carnegie, seek first to understand, then to be understood.

As it turns out, the tanks were deliberately choosing to allow themselves to be hit (and crushed for a comparatively small amount) by the channelers to prevent rage starvation. I honestly had no idea, so I'm glad I asked instead of going with the "hey look the tanks suck" approach I was initially considering ;-). So a little bit of knowledge really is a dangerous thing...


Unless he was tanking the first channeler, then rage starvation shouldn't be an issue. They have no need of extra rage as they aren't fighting dps for threat until thier target is the one being downed. All they have to do is auto attack, shield block every time it is up, and maybe the occasional shield slam, revenge, or sunder to keep threat above healers. They can let themselves build up rage while they wait by not going all out. Once the dps switches to their target, they can then go all out and should already have a decent threat lead to start with.

When I tank my channeler, I do JoL/SoL and merely keep up Holy Shield. Every so often, I throw out a consecrate for good measure. It's how I avoid being mana starved.

Not using shield block on progression content is not a good idea most of the time. They should be using rage pots if they need the extra rage anyways. If they are indeed rage starved, then they aren't taking damage and don't need to use health pots/etc. anyways.
Image
User avatar
jere
 
Posts: 2960
Joined: Fri May 11, 2007 5:12 pm

Next

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: marsbubble and 1 guest

Who is online

In total there are 2 users online :: 1 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 380 on Tue Oct 14, 2008 6:28 pm

Users browsing this forum: marsbubble and 1 guest