I actually saw a fellow Fury Warrior on my server with one equipped, and despite the tooltip text, the trinket does indeed proc off Rend and Deep Wounds. I certainly think it's worth at least exploring this trinket's potential for DPS Warriors.

For main-hand, maybe. Weapon DPS is not as big of a factor for your off-hand, as its DPS only affects 38% of your white damage, and the average weapon damage only affects 38% of your whirlwind damage.

Passive stats on your offhand, consequently, affects everything else. I'd much prefer an offhand that's 13-26 ilvls lower in damage but with loads of str and crit, over another weapon that only has damage, speed, and stamina on it.

Someone PM'ed me with a problem regarding totaling SEP for items. I totally fudged the AttT contribution where I just added in the AP from AttT directly into the SEP value, which is totally wrong and makes plate items a little better than they actually are. I'm gonna put in a fix that alleviates this soon.


Edit: and... it's up

Well as for possible things to look into (I wont call them bugs/fixes, because its not really a sure thing )

1) There is talk now on rogue forums and here that unlike explicitely stated by blizzard, expertise doesnt actually truncate. That would mean that a lot of current setups basing on chestguard of the recluse (and thus having 18 expertise as opposed to 17 or 19 with other options), are in fact a bit less ahead (although they still are).
Did you considered looking into those tests made? Granted , there isnt really any way to see for sure if 17.9 expertise really reduces more dodges then 17.1, but tests made on these forums included 20k+ hits and look pretty convincing.

2) Slam damage in your sheet seems inflated to me. Probably its a matter of human imperfection in usage. Fact remains - after looking over around 30 parses from top of wowmeters, I think the average for warriors is around 6% damage from slam. Sure there are fights with more like I think Grauls KT fight (almost 11% if i recall) - but they all have disproportionally high crit rate for slam - leading to a conclusion that the higher result is more of a RNG effect, rather then different playstyle. Your spreadsheet shows close to 9% expected bloodsurge damage for me, and its just not what I see in game. It does inflate the value of t7 bonus by a significant amount (We have pretty horrible luck with drops, so im far from BiS gear - but the "manual" reduction of t7 bonus did allow me to change a few things in gear to maybe drop t7 alltogether).

The difference in DPS between the new version and the old (2.304a) is for me ~321 DPS with the same items. How can this be? Perhaps I don't understand the change you made recently...

Shaman Unleashed Rage and Death Knight Abomination's Strength %up cells have no effect from 0 to 90%, then from 90 to 100 it appears to scale roughly linearly.

Hey Landsoul I'm taking a few Excel classes right now, I'm going to talk to some of my teachers to try and find a way to reorganize the gem sorting if you'd let me do that. Maybe color code the drop down box, or just something different.

There is something broken in the new version about the 4-piece T7 bonus. Enabling/disabling it seems to have no effect on the DPS result.

That seems like a pretty solid model of what the actual bonus means in game though

Have you tested all the trinkets?,seems like the spreadsheet is wrong as it rates mirror of truth higher dps than fury of the five flights.And no its not cause I have low crit,my crit is 38 %.Has anyone done any testing?,I think fury of the five flights is better imo,but am not sure anymore as it seems your spreadsheet is almost taken as gospil.

Just a small thing: The uptime for Hyperspeed Accelerators is based off a 2min CD - when they changed them in 3.08 the CD was reduced to 60 seconds. Not a huge thing, but does make some difference for engineers who are unsure as to use Crusher or the tinker

On the left side of the Lookup-Page (in the black panel) is an option whether to "Use Exp Adjustment" or not. In the default settings this is set to "1". Just turn it off and it will use your calculated SEP.

My first post so just wanted to say thanks to everyone who puts in the effort and time to help out their fellow warriors.

My issue is that i cannot access FileFront through my work's firewall. I have looked everywhere for another place to download the spreadsheet but have found none. I may be the only one with this issue i am not sure, but would it be possible to provide a secondary location to download the spreadsheets.

Barring this, would someone be kind enough to email the 97/03 spreadsheet to me, it would be greatly appreciated.

OK I think I have got an idea of how to make gem selection much more user-friendly, and significantly quicker. It will require a few changes. The idea I had was to use forms and make 2 list boxes. The first list box will have you choose the category of gem. For example, you will choose from: STR, AGI, Haste, ArP/hit... etc. Then right next to the first list box will be another list box that will have you choose the exact gem. For example, if you choose STR in the first box, the second box will list: +10 STR, +12 STR, +14 STR, +16 STR, +27 STR. Doing it this will I think will make gem selection much easier. I'll have to restructure the gems in the lookup table. Also, one problem is that there will have to be 2 more columns inserted into the database, and I know that will screw up other data- so a lot of formulas will have to be changed. Landsoul if you're willing to do that I'd be willing to fix the gem sorting, because I don't want to screw around with your forumulas

I am having a problem with the 2/20 Excel 1997/2003 spreadsheet. I am getting different numbers when I swap between Blood Frenzy and Blood Poisoning. Currently, Blood Poisoning is adding an additional 115 dps over Blood Frenzy; both are still 2% damage modifiers in game.

Entering in an extra 2% to the DIM of Phoney Stats will approximately place the dps to that of Blood Poisoning when I have Blood Frenzy activated.

[Shoulderpads of Secret Arts] are incorrectly listed as having 59 hit. I can fix this on my Lookup tab. Just a headsup for the next build and for those like me close enough to the hit cap that it seems an upgrade.

The Berserker uptime has bothered me since Landsoul put in the proper uptime calculation. Massacre should be better in the OH than Berserking, because it should only have a ~25% uptime. I was too lazy to trace the equations though, until now.

The PPM used in uptime is using the base weapon speed, not the hasted weapon speed. This is significantly inflating the value of PPM weapon enchants beyond their actual value. Fixing this brings the uptime back in-line with calculated and observed uptime.

PPM calculations do use base speed, not hasted speed. The values are accurate and not inflated.

PPM used base speed back in Molten Core. Sometime around the end of WoW 1.0 the calculations changed to use hasted speed.

Here, Grayson is saying that the spreadsheet's calculated uptime doesn't match what he observes in game. Changing the calculations to use hasted speed makes it more accurately match reality.

PPM use hasted speed for white hits and weapon speed on specials iirc.
The only special you do with OH is WW. This skews the uptime by a rather low %.

I do like this idea, or something to the general effect. If this were implemented, I would rather a left menu choosing a color (red yellow blue orange purple green, in that order for primary/secondary color, and then descending SEP) and a right menu containing a list that is aggregated from the gems of that color, in descending SEP.

For example:
Where would the quality (green, perfect, blue, epic) selection go? Off of the gear selection area - too much clutter. Instead, a filter would be added on the perimeter of the chart, and the quality selection would commence. From then on, all gems in the chart itself would only display the quality of gem selected once, instead of having to scroll through many gem qualities and flavors in one big honkin' menu. Honestly, people primarily choose one gem uality and stick with it. All perfects, or all rare, or all epic. Rarely does a person PLAN (which is what utilizing Landsoul's spreadsheet is used for, planning, not primarily for modeling)

If the spreadsheet is deemed to be developed around both planning AND modeling, this solution could be tweaked somewhat to allow different caliber gems , but this could also be solved through the exceptions notation described later in this post...


~~~---~~~---~~~---~~~---


This may or may not be doable, but here's my brainstorming...

A "standard" selection where a gemming could be pre-defined (color matching: all red slots use x, all yellow slots use y, all blue slots use z; color-ignore: +8 Str), and further, a user could choose "match all sockets", "ignore non-DPS socket bonuses", "ignore non-DPS increasing socket bonuses", or "ignore all socket bonuses". The actual gear with the gems shown could of course be modified as is the case now, to account for individual gearing choices, say, near the expertise cap, or to fulfill meta requirements.

An simplified example of a potential implementation (assume relative SEP being 2 Str, 1.6 Crit for all examples - actual SEP would be actively drawn from the current sheet):
Alternatively, the sheet could select the sockets that would make the most use out of the required blue gems, as an option to "enfore meta requirements". Also, the sheet could be told to ignore CERTAIN sockets as opposed to ignoring ALL of an item's sockets. The pro to be gained by this would be manual control over specific sockets, and then the sheet could optimize the gemming around this. A user could further use a system like the one Mightybeast described in the manual placing of gems. Manually choosing a gem would automatically flag that socket as being a user-set gem, and it would subsequently be ignored by the auto-gemming. Plus, when a new piece of gear is selected, the gems would automatically be optimized, until a user modifies them, including removing gems that no longer have a socket in them.

Selection of having the JC profession bonus should also be available, whereby the strong +27 Str (or whatever) could be regarded as being the top on the list for hypotheically filling blue sockets, and then yellow sockets, and finally red, should 3 blue and yellow sockets ot be available. Blacksmithing should also be an option, which will allow two more sockets, prismatic, which would be assumed as the highest DPS socket available, and a Dragon's Eye should fill all other sockets before touching the prismatic sockets. Or however it would end up working out.

Automatic enchanting could even be implemented at this point, with little further hassle, by simply using the same logic of drawing the SEP values from a table of enchants for each.



After thinking about this some more, a better way to do implement the "gemming preferences" could possibly be an automatic selection of the highest DPS gems, both of the color-matching and the wildcard, which would automatically pull the SEP for each gem from a gem table and use that to fill in the sockets as it goes along. The user could still force certain certain sockets to be "overlooked" with a manual option being set.

This could be a fair amount of work, but it could turn out to be a very sleek system! Choose your gemming preference, select if you're a BS / JC, let it know what qulaity level of gems you want to use, and you're set! Want to change a gem that was put in by the sheet? Choose what color the socket is, and then choose the particular flavor of gem. The sheet will now ignore that socket. Want to wipe the slate clean? A "reset custom gemming flags" button would remove all user-set gems, and allow the sheet to re-optimize, or simply hitting "optimize gems" would re-optimize the gemming AROUND the user-set ones.

Sorry if this seems rambly, but I really think that this idea could have some legs, if it doesn't come off as too daunting. I've attempted to go back and clean up the post as much as my foggy 6AM brain permitted.

I saw 128 procs out of 2276 attacks, or 5.6% chance on hit with 99 haste(3.02%) and no flurry. I saw 92 procs out of 1500 autoattacks, or 6.1% chance on hit with 218 haste(6.65%) and specced for flurry. I find it difficult to believe haste lowers your chance to proc zerking per hit(by using hasted attack speed) looking at that.