Something that has caught my eye while reading through here and while trying map out a few of my next potential upgrades, using Redcape's stat values I have found (as have others) that some of the tanking gear with mastery on it has exceptionally high STR values as well.

Such examples include the helm off of Valiona (Daybreaker Helm), and our very own T11 prot chest piece. Both of these boast higher STR values than anything designed for ret, and with some reforging become not just viable, but arguably our BiS.

Right now our mastery is valued extremely low, but with the potential boost blizz has already commented on even if it only raises the value of mastery to on par with haste (.16) this still puts the T11 prot chest in the lead, the only real loss to speak of would be managing our set bonuses around it.

But, that brings me to my last point, with the T11 prot chest being a top contender for our BiS chestpiece, is it possible that our 2pcT11 from prot is a good route to take? or that it might be more valuable than the 4pcT11 ret set? These things are outside my scope of capability, however, if one of our math friendly forum posters is a bit more adept at modeling the bonuses this might prove to be very useful information.

By using Redcape's spreadsheet I switched off the 'T11 - 4P - Inq + HP' and increased 'CS Weapon Damage %' by 10 (upto 126.50) and without changing any of the gear or other stats within the spreadsheet the rotation DPS came back -168. So looking at just the two set bonus' we would be at a DPS loss.

I believe that due to the different stats we would have in taking the two prot pieces we would end up stacking some Dodge, Parry or Mastery and I feel this would lead us into a greater overall DPS loss.

Let's please not start presenting ideas for Seal of Truth and other wish-listing. There is also very little to talk about with respect to the new mastery until numbers are provided.

Some preliminary testing for Darkmoon Card: Hurricane

Melee only control


With Seal of Truth, but it fell off halfway as I was AFK.


Normal rotation for 210 melee hits


All testing was done on the 85 normal dummy. It would seem that unlike TAiaJ, Hurricane does not benefit from hidden strikes from seals.

P.S. On a side note, interesting proc from avalanche from a censure tick:

The numbers seem interesting to say the least, further testing, of course would be needed for anything conclusive. If it were possible, a 1k (10k would be ideal) swing log with Hurricane Deck on would, I think, help value the trinket more.

It would have been better if you'd included the number of CS and TV you used in this test.

Assuming you were using your full normal gear, and assuming average proc rates for DP and HoL, in 210 melee swings you would have had 162 CS (if used on CD), 64 Judgement (known figure), 54 TV from natural HoPo generation, 22 TV from Mastery and 28 TV from Divine Purpose HoPo gains. The total number of melee attacks including whites would be 540. 37 procs from 540 is a 6.85% proc rate, consistent with the proc rates of 7.2% and 6.6% observed in the 1000-swing tests.

Barring contradictory findings, it seems safe to say that DMC:Tsunami procs at 7% and can proc from any melee attack including Judgement, but does not proc from spells, seal procs or Censure ticks. (It's not impossible that it can proc from spells, but the variation of the extremes from the average would have to be more than twice as great.)

It may have procced at 7% (or less) for this test, but it's entirely possible (even probable) that it's a PPM mechanic, so unhasted vs hasted weaponspeed is an important factor to determine accurately.

For a full and proper test just run individual ability tests. Examples:
1) Stand off from the boss and hard cast Exo. Go for at least 100 Exo total or first Exo that procs.
2) Repeat for Holy Wrath.
3) Repeat for Consecration.
4) Repeat for Judgement
5) Autoattack with no seal for prolonged period (1k hits). This should give us an accurate baseline PPM or % chance.
6) Autoattack with seal. Stack to 5. Stop and wait. Rinse and repeat. See if any Censure ticks during no-attack periods proc.
7) Find that 90 degree sweet spot where you can CS but autoattack will not occur. Only use CS, no TV, until it procs.
8) Repeat CS test, adding in TV. Do a reasonable run of around 100 TV (Zealotry helps). Number of procs should be higher than CS alone allowing us to verify TV can proc. Alternately build to 3 HP. Wait 6 seconds, use TV. With the TV well separated in time it should be clear if it is what creates a given proc.

The goal is simply to determine which attacks are eligible to proc the card.

Please remember to remove items which proc haste and do not judge during the autoattack test, as the alteration in swingspeed would muddy testing.

Edit: Edit for posting before reading all new threads. Rogues have tested DMC: Hurricane here and are coming to possibly 1 PPM. Benefits from 8% spellpower (?!? Not sure how, this would need heavy testing for my satisfaction) and unsurprisingly crits at 150%.

Just thought I'd bring this to the attention of people here:

World of Warcraft - Retribution Paladin - RetPaladin.com • View topic - censure and inquisition

Just wondering if anyone else here has noticed Censure behaving like Kramoz (first post on the page) is suggesting. Shouldn't make much of a difference for our DPS (!), but if he's right the behaviour does seem odd...

I'm doing some testings with your scheme for DMC'Hurricane, so we can gather more data from different attacks, here are some results, I'll keep this post updating so someone can do the maths with Stardusty's and my data.


1) Stand off from the boss and hard cast Exo. Go for at least 100 Exo total or first Exo that procs.
2) Repeat for Holy Wrath.
3) Repeat for Consecration.
4) Repeat for Judgement

1. No Proccs (obvious)
2. No Proccs (obvious)
3. No Proccs (obvious)
4. Procced



It didn't procc from Censure Ticks but if unleashing SoT proccs it I try to figure out later.

The Trinket itself doesn't have an ICD which you can see here:

[03:50:55.119] Schädel Lightning Strike Raider's Training Dummy 4382
[03:50:56.042] Schädel's The Art of War fades from Schädel
[03:50:56.926] Schädel hits Raider's Training Dummy 1516
[03:50:57.938] Schädel Lightning Strike Raider's Training Dummy 4220

e: narlith wrote this already in his post in the Rogues Section.

And here are more results:

CS only (with 1091 haste)



CS + TV Only (with 1091 haste)



HoW Proccs it: too and you can see that the Lightning Strike ist affected by AV.

DMC: Hurricane
Schädel's 3.8 speed weapon would provide a 6.33% chance per hit on a 1 PPM. His (limited) CS and TV testing come up with 5.6% (CS) and 5.98% (CS+TV). These are reasonable with the sample size involved. If PPM is greater or less than 1, I expect Rogue testing (Huzzah for fast weapons) will let us know.

In general it's looking pretty clean if we say only melee abilities proc. Since Judgement and Hammer of Wrath use specialized melee rules having them proc isn't unreasonable (great and lucky test on HoW!).

With the ruleset of 1 PPM, eligible attacks are autoattack, CS, TV, Judgement, and HoW, and Lightning Charge critting at our spellcrit rate, in my personal gear (far from phenomenal) DMC: Hurricane's proc is worth around 270 DPS (not counting the passive Str). Overall (Str+Proc) it's barely edging out Heroic Epic trinkets in my spreadsheet. It looks to be obsoleted when 4.1 trinkets become available.

Bottom line: Solid trinket, but not worth a large chunk of your total cash. Individually assess your finances and its personal worth.

Reforging
I've built an automatic reforge option into the next release of my spreadsheet. It is not perfect as it's only pursuing two approaches, rather than every possible reforge combination.

All reforging uses the converse of The Price is Right rules - closest over the goal is the winner.

In sequence it:
1) Reforges for Hit cap (it will not consider even 1 point below cap, unless cap is impossible).
2) Any unforged slots are then considered to reforge for Expertise cap.
3) Any leftover Mastery is reforged into Crit or Haste.
4) It determines the total wasted rating over Hit and Exp caps with this setup.
5) Clears all reforging.
6) It reforges for Expertise cap.
7) Any unforged slots are then considered to reforge for Hit Cap.
8) Again tossing leftover Mastery into Crit or Haste.
9) Determines wasted ratings over caps with new setup.
10) Compares the two values and applies the setup with the least wastage.

As I said, this isn't perfect because it is not considering every single stat combination (the billions of combos Chippy referred to last Thurs). It could miss a hybrid solution where neither Hit, nor Expertise are as close to cap as possible, but the overall wastage between them is lower. Example with arbitrarily selected numbers: Option 1) 962 hit, 420 exp = wasted of 30, option 2) 975 hit, 393 exp = waste of 16, uncalculated option 3) 968 hit, 395 exp = waste of 11. Spreadsheet would select option 2, since 3 is never calculated.

It does reach our goal points and helps juggle Mastery (so we don't have to do that manually). Any possible loss by not finding the perfect solution should be quite minimal (such as 5 rating in the above faked example). Manual changes are always valid, so individuals can choose to use other programs or calculate values by hand if desired, then configure in the spreadsheet.

On top of all this, it takes approximately 1 second to run the entire Reforge routine. Somewhat of an improvement over Chippy's 30minute program. Since I rolled the function into the auto-gem and auto-enchant function, it overall takes about 10ish seconds to perform all three tasks and then recalculate the DPS for changed stats (recalc taking the lion's share of time).

I expect a new release sometime this week.

Seems like a reasonable way to do a quick reforging pass. A few notes:

* The biggest issue with this approach I think is going to be the "leftover" items that you have to reforge to haste instead of crit (because the item already has crit on it). This could easily be an extra 200+ haste that could have been crit with an optimal solution which is the equivalent of being an extra 20-40 points over the caps.
* I doesn't sound like you are considering reforing hit -> crit and then crit (or anything else) -> hit on another item to get closer to the cap. I noticed this a lot in my optimized solutions where a lot of hit/exp gear would be reforged to crit so that an item with a slightly lower reforge amount could be used toward the caps to get a few points closer. This also helps ensure that every item with haste/master/parry/dodge can be reforged to crit/hit/exp.
* As you mention, its possible there's a setup that neither minimizes wasted hit or exp, but does minimize the overall waste. This, honestly, seems like it would be the least significant of these 3 effects.

Overall, though, probably a decent way to get a quick set of reasonable reforging values.

I re-implemented the exhaustive algorithm in C# and was able to get nearly a 1000 times speedup. Compiled and optimized code really helps, along with the ability to run the embarrassingly parallel search on all 4 of my CPU cores. I also made some smaller optimizations like packing an items hit, expertise, and overall score values into a single 64-bit int so I can process each item in a single CPU cycle, as well as modifying some of the bookkeeping code to perform fewer memory accesses.

Overall, the computation that took a bit over 30 min in lua took only 2.3 seconds with this new implementation (and found the same results). Still, this is with a lot of PvP gear, so with a full set of gear with 5 valid reforging options each, this will likely take about 10 minutes to chew through. Its also not very user-friendly at this point as I'm running it by using a lua script to dump my current item stats which I then copy+paste into one of the source files and then recompile and run the application, but it was definitely the easiest way to get it up and running.

I'm also tossing around the idea of trying to come up with a method that considers reforging cost that could be run in-game. I'm beginning to realize that I'd rather have a slightly sub-optimal solution that didn't require me to reforge my entire gear set each time I get a minor upgrade than have to shell out 100g+ each time to reforge everything for the optimal setup.

It does indeed consider this solution. Basically it builds a table with values of maximum hit, minimum hit, and if necessary 0 (I'll explain in a sec). For an item with natural hit maximum is full value, minimum is 60% (40% reforged away). The 0 is necessary for items that have 2 non-hit stats such as Crit/Haste. It would consider 40% Crit->Hit, 40% Haste->Hit, but it also needs to consider 0 hit (nothing reforged on that item).

My current optimum, as provided by this routine, actually has multiple juggles of Hit->Crit and Crit->Hit.

Were I to consider every possible reforge the table would grow far wider, the loops grow far more nested, and processing time increase (Excel could even choke on the quantity of iteration required). I limited the routine for these express purposes.

Glad to hear of your systemic improvement. I your solution should always be optimal, but hopefully a large portion of the time my quick routine would match your "closest to perfect" solution.

DMC
HoW with a 100% proc chance on DMC actually raises its value appreciably. Even just 6 more procs at average 5000 damage (ignoring crits being higher) is 30000 damage. In a 5 min fight that's 100 DPS right there. With the low usage of HoW, it may even be something Blizzard finds not worth tracking and correcting, even if unintended. This glitch would push the value higher than all other present trinkets and possibly cause it to remain competitive in 4.1.

I would appreciate it if Schädel (or anyone else with one) could run a quick sanity check of HoW, just to confirm it wasn't a brief glitch or server related (although RNG giving all 15 in a row is less than 1 in a quintillion chance). Should be easy to pop AW and HoW 3 times without using anything else.

Insanity check done, proced straight in a row for all HoWs I did. I remember skimming a post that said it also procs on all Avenger Shield hits including the second and third targets, so HoW proccing all the time may not be unique.

The hotfix rolled today also has fixed GoAK and now properly increases Strength by 20%

Does anyone know if there's still an active maintainer for the Rawr Retribution Model ?
The Rawr Models table says OReubens is the current maintainer, but there's been no check-in since 06/10/2010 and it seems our model is in dire need of an update : Rawr Models

Yay, really! Although it remains to be seen if the Guardian still has problems switching targets, which was odd.

And, the Righteous Fury thing is really somewhat of a nerf for PvP retri: we used it to slightly counter Purge and Spellsteal, Blizzard even said they wanted to nerf them, and now they make this undispellable? This probably means the dispels are going to hit the rest of our other abilities instantly, skipping it.

See
Rawr - Rawr - View Issue #19423: [Ret Optimizer] Showing lower dps after optimization.

I have him on MSN and poked him about it. The situation is still the same. If he can get to work on it at the start of february, the soonest we can see something usable in Rawr is early March.

A guy in the official forums suggested a macro that might work:

Source: Ret Guardian of the Ancient Kings - Forums - World of Warcraft

I'll be testing that today to see if it works as intended.

I have tested this macro and ones similar to it without any luck. I am yet to find a successful command to make the guardian switch target.

Quick update on DMC: Hurricane. As noted the other day in the Rogue mechanics testing thread, the damage proc of DMC: H is affected by Communion. Both the native 2% (for us) bonus, as well as the 3% more from auras update the tooltip when equipped (toggled aura last night just to see). This boosts the proc for us by, naturally, 5%. Since Str is unchanged this does not make the card overall 5% better, just the proc - so overall is probably 2-3% better.

To recap previous findings, based on the fact that the DMC:H will proc on 100% of HoW attacks it becomes our currently BiS trinket. The Str is nice and the procs occur just frequently enough to provide enough DPS boost to beat other trinkets. It is not crazily ahead of equivalents (like DMC: Greatness in first tier Wrath), but still earns the number one title.

When you say that it is BiS, do you mean that it is BiS pre-hardmodes, or are you saying that its just flat out BiS until the next tier of content comes out?

Throughout all of BoT / BWD the trinket was consistently doing between 3-4% of DPS from Lightning Strike alone. Crits based on spell crit. Value would likely be more on longer <Heroic> fights with lengthy 20% phases for higher quantity of HoW throws.

Cho'Gall Example of it doing nearly 283k damage.
World of Logs - Real Time Raid Analysis

I can confirm the data on DMC: Hurricane. It has done 409k (3.7% of total) over 5:12 long fight (Halfus) with 19 hits and 6 crits.
World of Logs - Real Time Raid Analysis

Exemplar's modelling show him that the trinket is somewhat better than the hardmode trinkets.