Reeshet

I'm not sure you've looked at the formulas in the spreadsheet to call this a trivial solution. On the Gear list tab, every item has a calculated AEP. That value for each item comes from a formula that is derived from the stats weighing at the top of the page. The value of those stat weights vary based on your current gear. To do what you're talking about you would need unique stat weights for every item (your current gear except without that item).

I do agree that what you're suggesting would correctly update the total AEP, but I don't think it would be trivial at all to implement.

Or am I misunderstanding what you're saying? Maybe there is a trivial & elegant solution. If so, just make changes on a spreadsheet with the new formula and post it.

 

todemax

Regarding defence, wouldn't it be feasible just to remove the crit reduction from defense in the EP calcs? Or to set the base defence at the cap in the EP calcs. Normally one would manually increase the value of defence if you're under the cap, and knowing how much defence is actually worth under the cap isn't of that much use.

 

ZorbaTHut

This is true, and this is clearly the problem. The solution is that I want to provide some sort of Quality Point system, where a set of gear can be objectively measured and given a total quality rating (obviously depending on the specific boss, and on buffs, etc etc etc, but we'll ignore that.) Once that is done, it's easy enough to provide delta-quality-point values, and quality-point-per-item-of-gear values, and everything just sort of works.

This is exactly what I am going to attempt to do. If it were trivial, I wouldn't be interested in it, so I'm glad it's not

It's worth pointing out that the "burst" math currently triples the chance of being critted and that turns out to easily push Defense into the top stat if you care at all about burst. Also, in the example I gave, I only used Defense as a simple easily-understandable value - you can replace it with absolutely anything that has a soft or hard cap and apply the same logic to it. (It's even more blatant when you have a straight hard cap, since it becomes 0.)

 

Gruntle

Oh, that's bad, good catch. Maybe the Burst statistic should just be something similar to what Resheet was suggesting earlier. So, just calculate the number of hits you can survive (i.e. no dodge/parry/blocks at all, just armor mitigation and health involved).

 

todemax

But then you should look at the max damage per hit from the boss, and not use the average, and that would require a remake of the boss sheet.

And as a comment to the sheet, Devotion aura and stoneskin stacked the last time i checked.

 

ZorbaTHut

This is pretty much what it already is as well No dodge/parry/block, tripled crit chance. If it's not already I'm also going to get in there and triple the chance of being parried, so that parry haste on parry-haste-able bosses becomes a major problem.

Yeah, that is pretty much true. I may just multiply the average damage by 1.1 or 1.2 as a quick fix.

I'm not sure I believe Devo Aura and Stoneskin stack. If they do, they're probably not meant to. I'll check once there's a shaman and paladin guildie on at the same time.

 

Thorgrim

They do stack, for some reason. Stoneskin is in its own special category for stacking purposes, unsure why.

 

bryn88

Is it possible to move this into a new thread for easy access to the spreadsheet since Zorba took over?

 

Gruntle

Hmm, I went and checked the calculations of the burst statistic. Yes, the crit chance is tripled, but this does not affect anything when you're def capped in the sheet. It's not the base crit chance for the boss that is tripled but the resulting crit chance that is left after defense anticrit is taken away. You can see this easily by looking at row 35 in the stats sheet. The worst case crit value (that is used to compute the burst time to live) is given there and when you're def capped this value is 0.00 at all times. So I think the statistic is ok, extra parry haste would be a nice addition to it I think.

 

todemax

Adding parry haste just seems to be a smaller factor, than bosses having special abilities that can land at the same time as auto swings. Adding the possibility for a special1, special2 etc. with damage and periodicity would be nice, but if it was possible to parse combatlogs to find a distrubution of "how long it takes the boss to give damage greater than the tank health" throughout the fight, then the burst calculations could be based on a 95 percentile time to live.

 

ZorbaTHut

I am hesitant to do this before it's been proven that I'm not a slacker. If I'm still keeping it maintained in a week or two, then I'll start my own thread.

I like the idea, but keep in mind that I've gotten info for precisely one boss post-3.0 so far. I'm kind of unexcited about requiring more data from bosses, considering that I can't accumulate the data I need already.

 

kahalm

It might be better placed at the simple answers/simple questions (or the other way around) threat but blockrating/value is nothing really to aim for. Unfortunately your current gear at the armory is your pvpgear so any furher advice would be a guess.

Block might be useful if you are on adds duty anubhero, and I do not value the t9 4-piece bonus high for other fights so it might be a good idea to change your pants and there might be an upgrade for your helm but maybe I underestimate the bonus.

 

todemax

Zorba have stated that the boss profiles in the spreadsheet are outdated. You're posting in a spreadsheet thread, so use the spreadsheet to try to answer any questions you have. You can input boss stats yourself and see what effect different gear pieces have, however since the profiles are currently outdated, you have to put in an effort yourself.

The questions you're raising are of personal preference, so I would highly suggest you to post in the simple questions thread, or create a thread on the topic of gearing for high end tanking, if you feel that it's a matter currently not covered by the warrior forums.

 

kahalm

If I understand WotLK Diminishing Returns - Avoidance - TankSpot correctly, you should aim for a balance between dodge and parry and the closer you are to this balance (1.88) the better defense becomes. With gear from Toc25/10 you are normally above the cap (I have currently 566) and adding a defense gem comes out slightly ahead of adding a dodgegem in terms of parry+dodge+miss

 

Reeshet

I double checked this last night before our guild Ulduar run. The armor bonus from Stoneskin and Devotion aura DO stack. Not sure if they're supposed to or not, but currently they DO.

 

ZorbaTHut

Alrighty then, I'll update the spreadsheet

fuck it, making a maintenance/bugfix release

 

ZorbaTHut

New version!

Mostly a maintenance release.

- Fix gem requirements (I hope)
- Update the dodge% change in 3.2
- Opened up second gem slot for Boreal Guard
- Split Devo Aura and Stoneskin into separate stacking buffs, 'cause they seem to be
- Simplified some equations

 

Llilyth

FYI, The Leviathan's Coil's now has 448 Armor, not 882. Thanks ZorbaTHut for updating this spreadsheet!

 

janaka

I haven't used the spreadsheet for a while, but for what it's worth, these are the values I use with Rawr atm (ToC25, only melee damage):

 

hefti

Thx for updating.

the blacksmithing Shield Enchant "Titanium Plating" have BV 81 instead of 40

 

Gromgoth

Could we get TOC and TOGC bosses in the next update? I tried to change it in my copy but things keep breaking. Using naxx 10 -25 bosses is giving you all kinds of wrong information if you look for upgrades in the tier 9 range.

Also, but less important: could you fix race change?

Thanks again for keeping the wonderful spreadsheet alive.

 

Gruntle

Just thought I'd post my updated version of Taliafear's sheet as well, could be good to compare the two at some point.
It's available here: wotlkwar20091020.ods

v. 091020
- Corrected mechanics for 3.2.2, talent changes, dodge/parry/ArP changes etc
- Added gear up to 3.2.2 (some might be missing, some probably have faulty/incomplete descriptions on where to find them, stats should be correct)
- Implemented a cap to BV contribution to Shield Slam, based on stuff on tankspot, hardcap at ~3100. Linear interpolation in empirical data.
- New Blade ward simple "model", assuming one stack and an uptime of 20%
- New modeling of hp/sec stat and similar trinket/enchant, AEP based on the value of unmodified BV
- Added some boss profiles for ToC25, need more input on them (weightings? magic dps? parry?)
- Trinket modeling for The Black heart, Eitrigg's oath (based on the discussions earlier in this thread)
- Split Devotion aura and Stoneskin totem

The main thing to add to this sheet are the set bonuses and better boss profiles. Taking set bonus into account for the computed stats with the current equipment is easy enough, getting AEP values for the bonuses is harder (but doable). Boss models need more input, I have used the dps and speeds given above, but some more details would be good. Any magic dps? Do the bosses parry? Suggestions for weights, how important is time-to-live, burst survival and threat/dps for each of them? I put in some guesses, but since I'm not raiding at that level, it would be good if someone who has experience with ToC bosses could comment.

edit: Also, Blood draining is very unfavorably modeled in this sheet, take the AEP for enchants with a grain of salt. It's tricky to model these things, a better Blood draining model is something that should be done.

 

janaka

I don't have any data on magical dps currently, maybe I can hand them in by the weekend.
All bosses do parry, but I suppose you mean parry haste. I don't have reliable information on that, all I know for sure is that Gormok on hardmode doesn't parry haste. I think he neither does on normal, and I'd suspect Anub to also have the parry flag turned on.
Weightings do of course strongly depend on personal situation, but I can give you the values I use with Rawr (roughly):

Also note that the damage values I posted are raw damage, not DPS. If you need DPS, just divide them by the speed. I will re-check the speed values on the weekend, it's possible that they are off by about 0.1s. So take them with a grain of salt atm.

 

Reeshet

So, while I don't have any specific problems with the weightings you've listed out, can you give any reason why those should be deemed correct other than the fact you use them?

For instance on Jaraxxus, having a lot of the dps killing adds makes threat per second way down on my list of priorities. It goes slightly higher if we're just tanking the mistresses for the achievement but even then survival and mitigation is a higher priority to give the healers breathing room.

For the Twins, I'd again rank survival higher (or at least as high) as TPS considering all the raid-wide damage that healers are going to have to heal through.


I'm not saying your numbers are wrong so much as asking if there's any mathematical reason they should be considered "right".

 

Jayde

One thing I would mention is that the Twin Valkyr hit extremely weak and when your Mages are pushing out 14k DPS TPS becomes a lot more important than survival.

The MTs takes virtually no damage in that encounter, other than when the opposing shield is up--which can easily be cooldowned. We have never had an early MT death in any attempt on any difficulty of Twins that I can remember.