Retribution: 3.3 PTR Changes - Page 3

Neraya · 10/19/09, 11:42 AM

Ok, but ECD's would be different if you swapped Exo and Cons wouldn't they ?
You're using ECD's here asuming the priority is Cons > Exo.
Swap 'em and the ECD of Exo would be slightly lower while ECD of Cons would be slightly higher.

Trying in Rawr, it doesn't seem to matter (Cons first still is better), but I think your '55' effective is a bit too much in favour of Cons.

• Zurm · 10/19/09, 11:56 AM

Originally Posted by Neraya

Trying in Rawr, it doesn't seem to matter (Cons first still is better), but I think your '55' effective is a bit too much in favour of Cons.

I used logs from this past week to determine the damage, and typical numbers from Rawr for effective CDs. I've never run Rawr for effective cds with Exo > Cons, but really I can't think of a single fight where you'd want to do that in the current content (Ulduar or TotC). Especially considering MOST allow you to hit multiple targets. I encourage you to repost the math in the other priority with Exo first, but I guarantee you won't see an effective CD under 17s, which puts it at a still pathetic 411; since consec has a much lower CD, I'm sure that the effective CD won't even rise to more than 13s, swapping it to 615 dps. There is still a massive margin of difference, and I'm going with the favor of assuming an undead/demon mob. Once you go to a mob without the guaranteed crit (and you experience the fairly low relative crit rate that exorcism has) it becomes even more pathetic. In my mind, exorcism is really not much more than a nice nuke ability for PvP and a filler for PvE.

Sure, there's situations where for a short moment in a fight (Yogg portal room) consecration might be less ideal than exorcism, but you really have to keep in mind that once we lost 2pc T8, exorcism became pretty bad. People underestimate consecration because it doesn't show big numbers on your screen, but it ticks EVERY SECOND and is probably one of the more powerful DoTs in the game.

If you are getting very low consecration uptime, you should reconsider how you drop it/position it. Don't simply hit it when CLCret tells you to; make sure the boss/target is done moving, then hit it. Anticipate movement, and position accordingly. Also, keep in mind that just because the boss doesn't LOOK like they are in the consecrate doesn't mean it's not hitting them; this isn't pre-3.0 where consecrate had to hit the center of the boss's body. Proper positioning of consecrate is one of the things that seperates a mediocre ret pally from a great one, and reflects appropriately in your DPS on movement-heavy fights.

gmedina · 10/19/09, 1:27 PM

Zurm-

Could you provide some hints as to how you use conscrete or any mods you use to help you in assuring you get as much out of consecrate as possible? I know i have tried things like moving in frt of boss for a few sec to drop conscrate to ensure more up time but any other hints or tricks you have learned in the hard modes?

• Zurm · 10/19/09, 1:48 PM

I really can't, as it depends on tank behavior and guild strategy. But things like dropping it in front of the mobile worm in case your tank has to move from cloud during P2 beasts, or holding to drop it on the pull of a boss until it's in position, are what you should look for. Positioning yourself on multi-target fights so you can have it hit multiple targets is huge. It's common sense more than anything else.

Talith · 10/19/09, 3:01 PM

Originally Posted by Neraya

Ok, but ECD's would be different if you swapped Exo and Cons wouldn't they ?
You're using ECD's here asuming the priority is Cons > Exo.
Swap 'em and the ECD of Exo would be slightly lower while ECD of Cons would be slightly higher.

Trying in Rawr, it doesn't seem to matter (Cons first still is better), but I think your '55' effective is a bit too much in favour of Cons.

You have to remember also that Exo is a spell an has a higher hit cap, it still misses often for me. No reason to put exo ahead of cons

• Zurm · 10/19/09, 3:44 PM

Originally Posted by Talith

You have to remember also that Exo is a spell an has a higher hit cap, it still misses often for me. No reason to put exo ahead of cons

/facepalm

They are both spells. Both have full and partial resists, and the same hit is required. The difference is that for consecration the mechanic is determined per tick, instead of on cast.

• frmorrison · 10/19/09, 5:57 PM

Other reasons why Cons is better than Exo when on Undead is how Exo crits at 150%, so the crit bonus isn't as good as with melee attacks. Also we have a glyph that gives two more tics.

Perhaps in 3.3 when you can use Divine Storm a lot and perhaps the bosses will move around a lot Cons will go out of use, but you have to wait until more people go on the PTR and test the new encounters. On the first boss of ICC Cons is outstanding.

Terryn · 10/20/09, 3:10 AM

Biggest issue i see is mana problems popping up with what looks like a lot of DS usage, just may push concecrate off of the rotation some, esspecially if the fight requires even minimal moving.

Rëtribution · 10/20/09, 12:40 PM

and thats exactly why our 2set will most likely be garbage again, coming 3.3

odesius · 10/20/09, 3:11 PM

I am so excited for this new libram, especially for target switching fights.
Let me pick an easy one that everyone here has seen>
TOC Gormok the impaler, using SOV with valiance libram moving to kobalds my dps takes a dive
SO I have macro to swich SOR/pvp libram for add duty, THEN back to SOV/valiance for boss.
With new libram i only have to switch seals, my crusader strike libram is still stacked >thus minimal dps dive

Makes for a happy pally...

Spiteri · 10/21/09, 1:00 AM

Do you still see retribution pallies swapping libram depending on seals or shall we only be sticking with the 3.3 libram since it is not dependent on seal unlike Valiance?

Terryn · 10/21/09, 2:23 AM

Originally Posted by Rëtribution

and thats exactly why our 2set will most likely be garbage again, coming 3.3

Oh no, its not going to be garbage at all. Things i'm seeing is basically moving DS to around a 3-4 sec effective CD if it procs off of ANY melee ability like many trinkets with that wording do. This is HUGE and will just mean you are going to have to pay attention to your mana pool.

Do you still see retribution pallies swapping libram depending on seals or shall we only be sticking with the 3.3 libram since it is not dependent on seal unlike Valiance?

Theres a whole page of discussion on this in just this thread, another 2 or 3 in the 3.2 ret thread.

Soher · 10/21/09, 2:34 AM

T10 2pc will be also insane for pvp. Reseting the cd of DS is just an insane amount of burst we will be capable of, even more than nowdays.
'Say Hello to mah Hammer!'

For PvE personally I don't think that it will be a mana sink. Delaying high mana cost abilities for a DS is probably a gain, instead of a loss in mana.

Thethiala · 10/21/09, 5:47 AM

Should I be putting forth the effort to get Libram of Discord now in case Divine Storm spam turns out to become reality, or will the Libram of Valiance be better anyway? My intuition says it's not worth it.

Neraya · 10/21/09, 5:48 AM

Originally Posted by Zurm

I used logs from this past week to determine the damage, and typical numbers from Rawr for effective CDs. I've never run Rawr for effective cds with Exo > Cons, but really I can't think of a single fight where you'd want to do that in the current content (Ulduar or TotC). Especially considering MOST allow you to hit multiple targets. I encourage you to repost the math in the other priority with Exo first, but I guarantee you won't see an effective CD under 17s, which puts it at a still pathetic 411; since consec has a much lower CD, I'm sure that the effective CD won't even rise to more than 13s, swapping it to 615 dps. There is still a massive margin of difference, and I'm going with the favor of assuming an undead/demon mob. Once you go to a mob without the guaranteed crit (and you experience the fairly low relative crit rate that exorcism has) it becomes even more pathetic. In my mind, exorcism is really not much more than a nice nuke ability for PvP and a filler for PvE.

Sure, there's situations where for a short moment in a fight (Yogg portal room) consecration might be less ideal than exorcism, but you really have to keep in mind that once we lost 2pc T8, exorcism became pretty bad. People underestimate consecration because it doesn't show big numbers on your screen, but it ticks EVERY SECOND and is probably one of the more powerful DoTs in the game.

If you are getting very low consecration uptime, you should reconsider how you drop it/position it. Don't simply hit it when CLCret tells you to; make sure the boss/target is done moving, then hit it. Anticipate movement, and position accordingly. Also, keep in mind that just because the boss doesn't LOOK like they are in the consecrate doesn't mean it's not hitting them; this isn't pre-3.0 where consecrate had to hit the center of the boss's body. Proper positioning of consecrate is one of the things that seperates a mediocre ret pally from a great one, and reflects appropriately in your DPS on movement-heavy fights.

Ok first off... I never said Exo was better than Cons. I even pointed out that "it doesn't matter, cons first iss till better"
I was just pointing out that in the

consecration would need to be less than 55% effective.

this assumed cons>exo order, but if you want to know the efficienfy when to swap them around, you need to calculate with ECD's with them already swapped.

In any case, I made a FCFS simulator program that allows me to simulate the rotation under various rotation priorities GCDs etc, and to do do my 'napkin' projections on the T10 bonus (more on that later). Yes, I have advanced napkins

So doing the numbers again: using perfect rotation with 1.6GCD (to account for lag and less than perfect reflexes) on a 5Min fight:
J>CS>DS>HOW>Cons>Exo:
Cons ECD: 11.53 eDPS: 693
Exo ECD: 17.64 eDPS: 396

Yes the numbers are off, but I can't run numbers calculating what the ECD would be if zurm turned his rotation around

I can't even say why you're getting 19sec ECD on exo. I'm not even sure where you get the 664 EDPS number with 8000 total damage, 83% uptime and 12 ECD on cons :s

Cons and Exo swapped:
Cons ECD: 11.53 eDPS: 693
Exo ECD: 16.66 eDPS: 420
Swapping them around does give Exo a slight benefit ECD wise.

Like I said, Cons remains better than Exo, but the efficiency number posted by zurm (again, sorry, not sure how you got that 55% number out of the numbers you posted) was a tad too much in favour of cons.

Right the big T10 set bonus discussion...
What does it proc off, and what's it going to do...
* We'll loose the T9 bonus so that will likely put us back at a CS>J situation
* I can't run actual DPS numbers since we've got no idea on the amount of Str, Crit etc we'll get on out IC gear.
* This is just to give an idea of how potent the T10 bonus is going to be:

Testing assumes hitting an ability with 1.6 seconds GCD on melee abilities and 1.2 sec GCD on spells (Cons/Exo) in a 5min fight. (1.5 just seems too idealistic to me, even if I do my utmost best on a dummy and try and give myself carpal tunnel syndrome spamming buttons until the server accepts them I can't get perfect 1.5). My current gear gives me 30% haste, so ideal GCD on spells is 1.1 I'm accountinf for some lag/slow reflexes also.
I'm not activating HoW in this, so it doesn't matter where it is in the rotation.
I started off with the J>CS rotation and CS>J
Once the bonus gets activated, I added rotations with DS in front. I also added CS>DS>J to satisfy my personal curiosity in an assumption I had.
Weapon swing speed is assumed to be 2.7 which is slightly slower than what I'm getting in my current gear out of my 3.6 speed weapon, obviously faster weapons will cause slightly more procs, while slower or less haste will cause slightly less procs. Obviously I can't post numbers for every possible weapon/haste combination

The A tests are our current situation.
The B tests are assuming only melee white hits can triger the T10 proc and reset DS
The C tests assume all melee hits can proc the T10 bonus (white, CS, DS, judgement)

The numbers in the CS/J/DS/Cons/Exo columns are the number of times each ability got used during the 5min fight. (Multiply by Damage/abiliy and account for crit and you can get an idea of DPS numbers).
The T10 column shows how many times the Bonus procced and DS got reset.
the GAP column shows how many gaps there are in the rotation with every ability on CD, The number between parens shows the total time of those gaps). Gaps can be 0.1 to 2.5 seconds (would be only 0.5, 1, 1.5, 2 and 2.5 with perfect 1.5 gcd)

Test - Rotation                CS    J   DS Cons  Exo T10 GAP
---------------------------- ---- ---- ---- ---- ---- --- -----
A1) J > CS > DS > Cons > Exo   65   37   27   27   18 n/a 52 (40.20)
A1) CS > J > DS > Cons > Exo   70   35   27   26   18 n/a 35 (37.20)
B1) J > CS > DS > Cons > Exo   67   35   48   21   13  45 21 (20.60)
B2) CS > J > DS > Cons > Exo   68   34   46   21   15  45 19 (20.40)
B3) DS > J > CS > Cons > Exo   56   34   54   24   14  45 23 (24.80)
B4) DS > CS > J > Cons > Exo   59   30   54   24   16  45 23 (23.80)
B5) J > DS > CS > Cons > Exo   55   35   52   25   14  45 21 (26.60)
B6) CS > DS > J > Cons > Exo   67   30   49   23   15  45 21 (20.80)
C1) J > CS > DS > Cons > Exo   65   34   60   19   13 104 9 (7.40)
C2) CS > J > DS > Cons > Exo   66   33   61   18   12 104 7 (8.40)
C3) DS > J > CS > Cons > Exo   54   33   75   19   10 104 7 (6.20)
C4) DS > CS > J > Cons > Exo   59   29   69   21   12 102 14 (10.40)
C5) J > DS > CS > Cons > Exo   59   34   65   19   15 103 9 (7.60)
C6) CS > DS > J > Cons > Exo   67   29   62   21   13 103 10 (7.60)

So some rough conclusions:
- Even if it procs of every melee ability, we still don't get enough abilities to fill all of our fight time with GCD's. The 'good part' of that is that we still end up with a rotation based on largest DPS per ECD gets priority.
- It doesn't matter a whole lot whether it procs on melee white or all melee abilities. If it procs off every melee ability a lot of the procs just end up getting wasted.
- The amount of judgements and CS stays fairly stable so don't worry too much that we'll not be judging enough to get mana returned.
- Even Cons count is fairly stable, it seems the extra DS hits mainly eat into the amount of Exo's.

Have fun crunching some numbers of your own based on this...
If you want extra numbers or more detail, just ask

Jackinthegreen · 10/21/09, 7:53 AM

Neraya, any chance you could do the sim using J>DS>CS>Cons>Exo? Given that 4p is 10% extra seal -and- judgement damage, it still seems more likely that we'll be judging first when available due to damage and 2p taking a little more bite out of our mana. Even though the numbers are fairly stable for mana return, completing the five possible variations of J, CS, and DS would be useful to see where the abilities are going.

Neraya · 10/21/09, 9:38 AM

Done. Added B6 and C6 to the previous post so the numbers would all be together.
Also added the note about weapon swing speed in previous post.

Redcape · 10/21/09, 2:34 PM

Neraya, I have some questions about your models. You have more Exorcism uses than Cons uses in some of your parses even though Consecrate has a lower cooldown and higher priority than Exorcism. How do you explain that? If your model is actually working properly that result is impossible.

Also, our GCD on Consecrate and Exorcism is not 1.5s base. It will be something like 1.3 (depends on individual haste totals, of course) which can change the numbers somewhat. That is probably worth modelling.

Jeh · 10/22/09, 1:05 AM

Hey I wrote a python program that goes through priority lists and gives me some sexy numbers.
It steps through every 0.1 seconds and uses the values below for average damage for each of the abilities.
I've incorporated RV, seal procs, crits etc etc into these average values below and (assuming i'm hit and exp capped) misses for cons and exo.
It checks if the ability is ready and then resets the gcd, i've assumed 1.5 sec normally and about 1.3 for exo and cons.
It also shows me how many spare globals I have and the effective downtime in seconds.
CS = 5795
DS = 7203
J = 9406
EXO = 5935
CONSG = 6650
time = 300sec

Priority                      DPS  Downtime    1.5s   CS  DS  J Cons Exo
CS>J>DS>CONSGlyph>EXO        4049    43.4        8    71  27 35  26  18
CS>DS>J>CONSGlyph>EXO        3994    47.9       13    69  28 33  27  18
DS>CS>J>CONSGlyph>EXO        3958    50.3        8    66  28 33  28  18
DS>J>CS>CONSGlyph>EXO        3975    50.7       17    63  29 34  28  19
J>CS>DS>CONSGlyph>EXO        3991    47.9       17    68  27 35  26  18

5 sec effective DS cd
CS>J>DS-5sec>CONSGlyph>EXO   4319    27.9        9    69  42 34  26  17
CS>DS-5sec>J>CONSGlyph>EXO   4424    21.6        1    67  51 33  25  16
DS-5sec>CS>J>CONSGlyph>EXO   4182    37.7       13    55  56 28  28  14
DS-5sec>J>CS>CONSGlyph>EXO   4446    24.7        2    55  55 36  27  17
J>CS>DS-5sec>CONSGlyph>EXO   4314    29.7        4    63  45 36  24  18

6 sec effective DS cd
CS>J>DS-6sec>CONSGlyph>EXO   4155    37.9       11    70  35 35  23  18
CS>DS-6sec>J>CONSGlyph>EXO   4254    30.9        6    68  41 34  25  17
DS-6sec>CS>J>CONSGlyph>EXO   4222    35.2        7    63  45 33  25  17
DS-6sec>J>CS>CONSGlyph>EXO   4193    36.8        6    61  45 34  24  17
J>CS>DS-6sec>CONSGlyph>EXO   4158    37.3       10    69  35 35  24  18

7 sec effective DS cd
CS>J>DS-7sec>CONSGlyph>EXO   4155    37.9       11    70  35 35  23  18
CS>DS-7sec>J>CONSGlyph>EXO   4155    37.9       11    70  35 35  23  18
DS-7sec>CS>J>CONSGlyph>EXO   4181    37.5        6    65  40 35  24  17
DS-7sec>J>CS>CONSGlyph>EXO   4184    37.7        5    64  40 35  25  17
J>CS>DS-7sec>CONSGlyph>EXO   4213    33.5        0    71  36 36  24  16

As you can see this isn't a sim and I've only considered 3 cases for effective DS cooldowns.
If anyone is interested I can upload the file and you can put in your own average values to see what works for you.
Few cool observations I can make with this, have a look at the normal CS priority and J priority.

Jeh · 10/22/09, 1:42 AM

Interesting numbers without Cons glyph and with the Exo glyph
EXOG = 6600
CONS = 5320

Priority                      DPS  Downtime  1.5s  CS  DS  J Cons Exo
CS>J>DS>CONS>EXOGlyph        4026    39.3    0     70  25 35  34  17
CS>DS>J>CONS>EXOGlyph        3962    43.9    0     68  28 34  34  14
DS>CS>J>CONS>EXOGlyph        3945    44.5    0     68  28 34  33  14
DS>J>CS>CONS>EXOGlyph        4052    39.4    8     63  28 36  36  18
J>CS>DS>CONS>EXOGlyph        4062    38.9    1     66  26 36  36  18

CS>J>DS-5sec>CONS>EXOGlyph   4330    20.7    0     65  48 32  32  16
CS>DS-5sec>J>CONS>EXOGlyph   4466    12.3    0     67  50 33  33  16
DS-5sec>CS>J>CONS>EXOGlyph   4089    37.7    13    55  56 28  28  14
DS-5sec>J>CS>CONS>EXOGlyph   4509    14.1    0     54  55 36  35  18
J>CS>DS-5sec>CONS>EXOGlyph   4418    17.7    0     63  44 36  35  18

CS>J>DS-6sec>CONS>EXOGlyph   4306    21.5    0     70  35 35  35  18
CS>DS-6sec>J>CONS>EXOGlyph   4388    16.7    0     68  43 34  34  17
DS-6sec>CS>J>CONS>EXOGlyph   4234    28.3    1     60  43 34  34  17
DS-6sec>J>CS>CONS>EXOGlyph   4215    29.7    0     59  43 34  34  17
J>CS>DS-6sec>CONS>EXOGlyph   4306    21.1    0     70  35 35  35  18

CS>J>DS-7sec>CONS>EXOGlyph   4306    21.5    0     70  35 35  35  18
CS>DS-7sec>J>CONS>EXOGlyph   4306    21.5    0     70  35 35  35  18
DS-7sec>CS>J>CONS>EXOGlyph   4200    30.1    6     62  40 34  34  17
DS-7sec>J>CS>CONS>EXOGlyph   4200    30.0    5     62  40 34  34  17
J>CS>DS-7sec>CONS>EXOGlyph   4357    18.2    0     71  35 36  35  18

Edit:
Unless my code is horribly wrong; you can squeeze out 1 more J and about 28 more cons ticks for a sacrifice of 2 CS and 1 DS from not using the cons glyph with a J>CS>DS>CONS>EXO priority.

Neraya · 10/22/09, 6:31 AM

Originally Posted by Redcape

Neraya, I have some questions about your models. You have more Exorcism uses than Cons uses in some of your parses even though Consecrate has a lower cooldown and higher priority than Exorcism. How do you explain that? If your model is actually working properly that result is impossible.

Also, our GCD on Consecrate and Exorcism is not 1.5s base. It will be something like 1.3 (depends on individual haste totals, of course) which can change the numbers somewhat. That is probably worth modelling.

Turns out I had Exo and Cons swapped on every test due to an enumerate mismatching with the order in a table. I had noticed that Exo usage seemed higher than I expected and I also found it weird that DS procs ended up eating into the amount of consecrates while leaving exo's fairly untouched which was weird/unexpected. Problem fixed, I'll go crawl in a dark corner now

You're right on the GCD's of Cons and Exo, I didn't initially wanted to model this, but it's easy enough to do so I went ahead and put it in anyway. I have 30% haste myself, so 'ideal' GCD would be closer to 1.1 (so 1.2 accounting for lag and reflexes). I can run the numbers again with diffrent values for swing timer and GCD if you want. My model can currently handle 0.01 intervals for swing timer, melee gcd and spell gcd.

Edited the initial post.

Looks better and my numbers without T10 proc (test A1 and A2) match up with Jeh's results. The slight change is because of different assumptions on GCD's. With T10, numbers are different because of using a different mechanism to trigger procs in the rotation.
I'm calculating a pseudo random number every melee swing (B and C) and every melee ability (C) and checking if it's in the lower 40% percentile of possible values. Each of the tests the PRNG gets reseeded exactly the same. So even though I do account for RNG, the RNG is the same for all tests.

Neraya · 10/22/09, 7:41 AM

Originally Posted by Jeh

Unless my code is horribly wrong; you can squeeze out 1 more J and about 28 more cons ticks for a sacrifice of 2 CS and 1 DS from not using the cons glyph with a J>CS>DS>CONS>EXO priority.

If I punch in 1.5 / 1.3 as melee/spell gcd's I'm getting different numbers for a 300sec fight for that rotation:

CS=68 J=37 DS=26 Cons=34 Exo=17 Down:38.50 (without cons glyph)
CS=69 J=35 DS=27 Cons=26 Exo=18 Down: 47.20 (with cons glyph)

My numbers WITH glyph match up with yours. I get 1 extra CS but the rest is the same. My down time is also 0.7sec less. Do you start your first J at time 0 ?
if you start even a bit later or end a bit earlier that could be the explanation. I get my first CS at 1.50 and the last one just barely makes the 300sec mark at 299.40 Not gettign that last CS in does make the difference of 1 CS and a 0.7sec gap.

My numbers don't seem to match with yours for the 'without cons glyph' scenario. did you use the same numbers for GCD ?

As for why more dps without Cons and With exo glyph... you probably forgot to account for misses and partial resists on exo.

The rotation can give some funky results... with GCD for melee at 1.6 and 1.3 for spells, using J>CS>DS>Cons>Exo, you enter a stable, repeatable pattern at 136.9 seconds. When such a repeating pattern starts happening, results get 'weird'

136.90 Crusader Strike      
138.50 Judgement            
140.10 **GAP 80**			
140.90 Crusader Strike      
142.50 Consecration         
143.80 Exorcism             
145.10 Crusader Strike      
146.70 Judgement            
148.30 DivineStorm          
149.90 Crusader Strike      
151.50 **GAP 100**			
152.50 Consecration         
153.80 **GAP 10**			
153.90 Crusader Strike      
155.50 Judgement            
157.10 **GAP 80**			
157.90 Crusader Strike      
159.50 DivineStorm          
161.10 Exorcism             
162.40 Crusader Strike      
164.00 Judgement            
165.60 Consecration         
166.90 Crusader Strike      
168.50 **GAP 100**			
169.50 DivineStorm

As a human and because of lag, you'd never be able to actually sustain such a pattern though.

• Zurm · 10/22/09, 9:35 AM

Have you done any research into how adding HoW into the mix would affect all this? It would be very interesting to see what, if anything, changes sub-20%.

Also, it was mentioned earlier that the loss of the 2pc bonus would likely revert our rotation back to one with a CS priority. Something else to keep in mind is the sharp loss in the value of crit; using [Inscribed Ametrine] to reach a socket bonus will almost certainly no longer be the ideal choice, and we'll have to go back to getting pure STR gems with one prismatic.

Tobrexa · 10/22/09, 1:32 PM

Hm, now that you mention that, I think, this makes non-plate (or rather non-strength) items less attractive, as these tend to have more crit (for us at least) overall. I say this, because I was surprised that my rawr gave me the 245 reckoning as very close (about 5dps difference) to justicebringer and lothar's edge despite being not as slow, in my current gear setup... (and at the same time the 258 archon glaive ahead of the 258 justicebringer)

Heck · 10/22/09, 1:59 PM

Originally Posted by Tobrexa

Hm, now that you mention that, I think, this makes non-plate (or rather non-strength) items less attractive, as these tend to have more crit (for us at least) overall. I say this, because I was surprised that my rawr gave me the 245 reckoning as very close (about 5dps difference) to justicebringer and lothar's edge despite being not as slow, in my current gear setup... (and at the same time the 258 archon glaive ahead of the 258 justicebringer)

This was my feeling as well.

I've still been taking agi leather/mail upgrades here and there strictly because of the value of crit with our current 2pc bonus. I feel that doing the same in 3.3 wouldn't be advisable for two reasons. The loss of the T9 2pc and not knowing exactly when they're going to roll out the itemization changes they said were coming with Cataclysm (since some of these changes are usually rolled out before new xpacs go live).

Once they take AP off leather and mail gear (since the classes that use them will get all their AP from agility) it's value will be next to zero.