1. The trinket issues are because you have to manually change the stats. Changing the trinket will only change the trinket proc, the static trinket stats need to be manually changed in the stats portion, that's why you're getting nonsense results.
    Don't use it if you don't wanna use it, but so far I haven't seen a good argument why it's inaccurate. The problem I have with static gear lists is that your stat weights change with every piece of gear you have. Spreadsheets take this into account while a gear list simply gives an EP value for one item, but if you were to hit the expertise cap for example, the EP value on the list for any item with expertise will be miles off. It's easy when gearing your character to know you shouldn't get more expertise past the cap obviously but the balance between crit and AP is a lot harder (like when should you gem for crit, and when for Strength?), and I would argue that a spreadsheet calculates this more accurately.

  2. The trinket issues are because you have to manually change the stats. Changing the trinket will only change the trinket proc, the static trinket stats need to be manually changed in the stats portion, that's why you're getting nonsense results.
    Don't use it if you don't wanna use it, but so far I haven't seen a good argument why it's inaccurate. The problem I have with static gear lists is that your stat weights change with every piece of gear you have. Spreadsheets take this into account while a gear list simply gives an EP value for one item, but if you were to hit the expertise cap for example, the EP value on the list for any item with expertise will be miles off. It's easy when gearing your character to know you shouldn't get more expertise past the cap obviously but the balance between crit and AP is a lot harder (like when should you gem for crit, and when for Strength?), and I would argue that a spreadsheet calculates this more accurately.
    You're missing the point, a static list can give you a strong approximation of item individual strength that is very accurate since while stat weights do change on a gear to gear basis, it's a very organic progression and putting aside the obvious Hit/EXP caps there are no relevant breakpoints with TBC Fury Warrior that could cause the numbers to shift significantly. That means that for example we know that hit rating is weak after the soft cap, we know that crit rating is very strong and remains strong, we know that AP is always valuable and we know that armor ignore has exponential/increasing gains, while we also know Haste becomes very gradually better with improved gear, but also scales backwards if overstacked.

    So it's easy for us to put pieces side by side and calculate them based on how many stats they have and come up with an accurate result, very simple to stat count an item. But what happens if you're not basing your gearing on what is known, but rather on a flawed mathematical solution the simulates the game? The problem is that now you're basing your gearing choices on the sheets flawed interpretation of the game, the spreadsheet is counting those stats in a flawed manner from the beginning because the model does not follow the same one we see in game.

    That's why like I said the sheet is overvaluing Crit and Haste considerably, it has no working models for certain procs at all and the proc difference between some of the trinkets is off. Do you really honestly think that the difference between an identical 340 AP proc and a 300 AP is merely 3-4 dps? That's nonsensical to begin with, nevermind if it had included stats. If you just take the Romulu's Poison Vial average damage proc (not including crits/increased damage debuffs or anything related) it would provide 15.7 dps, that's something I can figure out by a basic PPM x damage / time, like I said the sheet is nonsensical.

    Use the dev of the gdocs list this forum?
    I have some suggestions to add and some questions (BT trash hammer best in offhand slot after glaive and s3 for t6 post za?)
    Anyway, the lists are almost the same what the most players declare as best in slot for fury.
    The sheet (gear ranking one) doesn't value offhand weapon speed that highly in the rating, so you're mostly getting a comparison of weapon dps + weapon stats. So if you use the comparison of Rising Tide + Swiftsteel Bludgeon, you have almost identical dps with Swiftsteel being 0.1dps higher but then

    Rising Tide = 21 hit + 44 AP
    Swiftsteel Bludgeon = 19 hit + 27 haste + 40 ap

    So the only differential advantage that Rising Tide has is the 2.6 weapon speed vs 1.5. Now unless you weigh that weapon speed very highly you're always going to get a superior value from the faster weapon because it's far superior in stats, and with the offhand stat sticks often outweigh the differential, if that were not the case then the offhand Warglaive would be outlcassed by much lesser weapons.

    In any case weapon dps/speed calculations are still a WIP on the sheet, it's fairly close but there is more work to be done on that front.
    Edited: August 5, 2017

  3. On the subject of offhand weapon speed though the primary concern of course is Whirlwind offhand damage. Just looking at Whirlwind damage for a second let's compare 2x Rising Tide with Rising Tide + Swiftsteel Bludgeon, at 3000 AP.

    Swiftsteel Bludgeon = 105-196 damage = 150.5 average
    Rising Tide = 208-313 damage = 260.5 average

    So we calculate the MH + Offhand whirlwind damage between both setups at 3000ap.

    Rising Tide MH = 260.5 + (2.4 x 3000/14) = 774.78
    Rising Tide OH = 38.5% decrease = 476.49

    Rising Tide x2 = 1251.27

    Rising Tide MH = 260.5 + (2.4 x 3000/14) = 774.78
    Swiftsteel Bludgeon OH = 150.5 + (2.4 x3000/14) = 664.78 - 38.5% = 408.84

    Rising Tide + Swiftsteel Bludgeon = 1183.62



    So we have a difference of 67.65 damage on Whirlwind at these numbers, of course that number gets multiplied between additional targets but on single target it's an incredibly small difference, still an advantage of course but then you consider the superior stat breakdown on the faster weapon and the more consistent Execute phases (though one could argue you may swap to fast weapons for Execute phase anyway). During the execute phase AP is devalued slightly while Hit/Haste/Crit (outside of reckless) and especially Armor Pen have a relative increase in value, which also plays in the favour of Swiftsteel due to the Haste on the weapon

    With that in mind from a single target point of view it's not unreasonable to think that Swiftsteel Bludgeon here is at least equal or slightly superior, but with Rising Tide taking the advantage as targets increase.

    For comparison - Warglaive of Azzinoth offhand WW damage.

    1.4 speed, 109.3 dps, 107-199 damage (153 avg), 21 agi, 23 crit, 44ap.

    153 + (2.4 x 3000/14) = 667.28 - 38.5% = 410

    So as we can see the WW damage here is nearly identical with Swiftsteel Bludgeon, the primary difference in damage contribution comes with the autoattack damage per second.
    Edited: August 5, 2017

  4. Going to add a Dagger comparison though too, to show the larger difference created by weapon type attack power normalisation as apposed to base weapon speed on the Whirlwnd damage calculation. Still using Rising Tide as the main hand reference, and 3000 attack power.

    Boundless Agony (Dagger) - 1.8 speed, 100.3 dps, 144-217 damage (180.5 avg) - 24 crit, 210 ignore armor

    Rising Tide x2 = 1251.27

    vs

    Rising Tide MH = 260.5 + (2.4 x 3000/14) = 774.78
    Boundless Agony = 180.5 + (1.7 x 3000/14) = 544.78 - 38.5% = 335.04

    = 1109.82

    A loss of 141.45 damage compared to 2x Rising Tide, or a loss of 73.8 vs Rising Tide + Swiftsteel Bludgeon. So despite the weapon being slower and thus having a higher base damage than Swiftsteel Bludgeon, the dagger attack power normalisation means there is a more significant loss in Whirlwind offhand damage.

  5. No, I got your point, you think that a static gear sheet represents the models in the game better than a spreadsheet. A static sheet also misrepresents the models in the game AND is based on a spreadsheet to begin with. Which do you think is more accurate?

    Yes, the different between 340 and 300 AP proc is probably around 3-4 DPS. Why? Uptime for both is maximally 20%, 20% * 40 AP = 8 AP (in practice it will be a lower uptime, and so the AP difference will be even lower). 8 AP is around 3-4 DPS.

    Romulo's Poison Vial proc DPS: 2.5% (procrate) * attacks/sec * 277 * 0.83 (17% spell miss chance) * (1+0.05 (misery)).
    Attacks per second for a Fury Warr wielding 2x Dragonstrike with 160 hit and 3/3 precision: 0.4 (MH) + 0.4 (OH) + 0.28 (MH instants) + 0.11 (WW OH) = 1.19
    Then proc DPS = 0.025 * 1.19 * 277 * 0.83 * 1.05 = 7.19. Show me how you get your 15.7 DPS because that is nonsense. When I use the empty Romulo proc in my spreadsheet it says the DPS addition is 4.3, which is a much better approximation to my calculation than your 15.7 DPS. A closer look to the spreadsheet shows that indeed, it is a bad spreadsheet for Fury (I'm guessing it was designed for Arms and later Fury was added) because it calculates the Romulo poison proc for 2H, not DW. If I adapt the formula to also count offhand attacks it gives almost the exact same result (the spreadsheet does not include Misery so it's 5% off). Like I said, it's not perfect but these small errors are easy enough to spot and repair. If your gear sheet is telling you that Romulo's proc is 15.7 DPS then yes, this spreadsheet is a lot more accurate.

    I still have not seen any good arguments why this spreadsheet is so bad! With some simple maths you can easily figure out it's very accurate.

  6. No, I got your point, you think that a static gear sheet represents the models in the game better than a spreadsheet. A static sheet also misrepresents the models in the game AND is based on a spreadsheet to begin with. Which do you think is more accurate?

    Yes, the different between 340 and 300 AP proc is probably around 3-4 DPS. Why? Uptime for both is maximally 20%, 20% * 40 AP = 8 AP (in practice it will be a lower uptime, and so the AP difference will be even lower). 8 AP is around 3-4 DPS.

    Romulo's Poison Vial proc DPS: 2.5% (procrate) * attacks/sec * 277 * 0.83 (17% spell miss chance) * (1+0.05 (misery)).
    Attacks per second for a Fury Warr wielding 2x Dragonstrike with 160 hit and 3/3 precision: 0.4 (MH) + 0.4 (OH) + 0.28 (MH instants) + 0.11 (WW OH) = 1.19
    Then proc DPS = 0.025 * 1.19 * 277 * 0.83 * 1.05 = 7.19. Show me how you get your 15.7 DPS because that is nonsense. When I use the empty Romulo proc in my spreadsheet it says the DPS addition is 4.3, which is a much better approximation to my calculation than your 15.7 DPS. A closer look to the spreadsheet shows that indeed, it is a bad spreadsheet for Fury (I'm guessing it was designed for Arms and later Fury was added) because it calculates the Romulo poison proc for 2H, not DW. If I adapt the formula to also count offhand attacks it gives almost the exact same result (the spreadsheet does not include Misery so it's 5% off). Like I said, it's not perfect but these small errors are easy enough to spot and repair. If your gear sheet is telling you that Romulo's proc is 15.7 DPS then yes, this spreadsheet is a lot more accurate.

    I still have not seen any good arguments why this spreadsheet is so bad! With some simple maths you can easily figure out it's very accurate.
    1. Of course I think the static gear sheet is more accurate, because it's using stat weights from an accurate sheet to rank gear and you're comparing it with an inaccurate sheet. That's the point I'm making that you've seemed to miss every time. If you have flaws all the way through the process then why do you trust the end result? Purely because a process is occuring? The only uncertain part of the gearlist sheet is the weapon list, because reliable weights of weapon dps/speed from Landsouls sheet were not available.

    2. Procrate on Poisen Vial at 3.4ppm average tested is 15.7 dps, yes that is without accounting for miss chance but it's also without accounting for crit chance, the sheet has it a 3 dps increase for me, 3 dps. My example wasn't intended to consider all aspects, it was intended to show in very rough terms that the spreadsheet is wrong. My numbers werent 100% correct or comprehensive, they weren't intended to be, but then neither are yours. All you've done while pointing out that my back of the hand math is wrong is to put in the extra effort to prove sheet wrong yourself, great stuff.

    Of course my sheet isn't telling that Romulo's is 15.7dps. For a start it's not even listed because it's a terrible trinket, and secondly my sheet isn't a gameplay model as has been said countless times, it's a pawn value sheet using stat weights from Landsouls sheet. It's a stat counting sheet. That's what you're seeming to miss here.

    3. The correct way to determine stat proc trinket strength is to determine the effective dps gain while it's up and then calculate that over the fight duration, so that for example you're accounting for 340 AP while Dragonstrike / Bloodlust / Other are active and then calculating the overall dps gain. Of course that is oversimplifying it and you have to account for many possibilities here, but they are extremely relevant factors.

    You're calculating trinket "dps" value based on an average of proc's stat value (8 AP more than Hourglass at 20% uptime) over an entire fight, which means you're undervaluing the effective contribution of the trinket through combined synergy (which is the primary reason that proc trinkets have often far outweighed passive stat boosters). The trinket in practical use has (tested) approximately 12-16% uptime depending on the fight, depending on group setup, depending on fight length, with a 45s ICD and a 10% chance to proc on Crit.

    4. If you have the ability to spot and repair errors that's fine, you can improve the sheet. But from a user point of view we are getting the sheet as is, so no matter how many fixes you put in place, that doesn't take away from the original sheet being bad. Do you have an answer to the sheet over/undervaluing certain stats? If so then can you identify and fix that? Or are you going to sit there and claim "it's fine" and then base your gearing around that very sheet? Are you going to start gemming pure crit as fury and searching for haste gear? Are you gonna sit and parrot that Rod of the Sun king or Syphon procs do not generate any dps gain?


    Like I've said, I cannot fix the sheet I can only spot the inconsistencies. Landsouls sheet is the only old sheet worth using for Fury, until someone brings that back or creates something better (like modern day Simcraft) then I'm going to stick to my guns. A pawn gear list created from Landsouls sheet SEP is better than some whacky half baked sheet that nobody used (for Fury) from 2007/2008.
    Edited: August 5, 2017

  7. If anyone is interested, this is the oldest version of Landsouls sheet I currently have

    FuryDPS2.000bBeta - https://mega.nz/#!tvBXgbyJ!0uyDO9chj...3xothU1_vJuWHE

    Do not use it for gearing advice, it would need to be modified. It's also giving me a lot of issues to get it working properly in the latest version of Open Office, or with Google Sheets. I haven't tried it in Excel.
    Edited: August 5, 2017

  8. 1. You're saying the sheet is inaccurate while I'm saying it's not.

    2. Procrate for Poison Vial is 2.5%, I've taken it from the EJ Rogue DPS spreadsheet which is the actual equivalent of the old Landsoul. That one is still on the internet luckily. I'm trusting that over anything else because that's the spreadsheed everybody on EJ used back in the day.

    3. Yes, because on average that is what the DPS addition is. Bloodlust and haste procs and all that **** average out to a certain haste value over the duration of a fight. Spreadsheets are to be used while using your brains, everybody knows that the value of procs and cooldowns goes up as the fights get shorter than a multiple of their cooldowns, and you need to keep in mind that everything is an average. If you're going to argue for a full model that accounts for all possibilities then you should give a distribution of all possible DPS outcomes, which you cannot give. You would have to make a simulationcraft for BC to do that. Landsoul didn't calculate a distribution for the procs either, Landsoul calculated the uptime based on how often you would hit and would then give an average uptime, not a distribution of uptimes where the proc would be calculated against the distributions of uptimes of other procs and cooldowns. Every TBC spreadsheet is based on averages.

    4. True, but like I said spreadsheets have to be used with brains. Any spreadsheet I use, I check the maths before I use it because I want to know what the numbers I get mean. Take this sheet, I used it for Arms and so I checked the formulas it used and it was mostly fine (one thing that disturbed me was that Blood Frenzy was not enabled by default for Arms specs so for my personal copy I have Blood Frenzy hardcoded in the pages where I use Arms). Anybody who doesn't bother checking what the spreadsheet is actually calculating deserves to be misinformed IMO.

    If your argument is that this sheet is bad because it uses averages, then Landsoul is also bad and your EP sheet is also bad. One more thing to think about, you think this would've been accepted on Elitist Jerks if it were some whacky half baked sheet? Food for thought.

  9. @SagaRally, thanks a lot for the Landsoul sheet. I'll fire it up in Excel just now and let you know. I wont use it in its current form, like you suggested tho.

    Questions; would it be worth it to swap to quick weapons for Execute? Where does the breakpoint go as to how good these weapons need to be compared to the normally used weapons.

    Also, what changes do I need to do in the BiS list you posted earlier (google doc) for it to work with Humans expertise taken into consideration (I see Axes has an added Pawn note thing). Also, T4 2set and T5 4set, how much do these weigh?

  10. If your argument is that this sheet is bad because it uses averages, then Landsoul is also bad and your EP sheet is also bad. One more thing to think about, you think this would've been accepted on Elitist Jerks if it were some whacky half baked sheet? Food for thought.
    Clearly you haven't read my posts and you're jumping to conclusions from your imagination, considering I actually said earlier in our discussion that Landsouls sheet works on averages, what is the point in this discussion if you're not even on the same page and following what is written? The context has long been lost at this point and this is going down the ****ter.

    And you hold EJ too highly, the place was/is full of nonsense, just go read the first few pages of any of the compendiums, there is a lot of unknown information simply being speculated, placeholder text that never got finished because it was never figured out, lots of wonky theorycrafting, lots of "unknowns" and very few guys who truly had a good understanding, Landsoul was one of few on what was a public forum with discussion contributed by many. And this sheet never was "accepted" by the Fury community, it was just posted on the site, nobody used it for Fury because everybody used Landsouls sheet and your defense of it being "accurate" seems to rest on the fact that it can be fixed.

    Well go ahead and fix it then, come back to me when it posts reliable and predictable results.

    @SagaRally, thanks a lot for the Landsoul sheet. I'll fire it up in Excel just now and let you know. I wont use it in its current form, like you suggested tho.

    Questions; would it be worth it to swap to quick weapons for Execute? Where does the breakpoint go as to how good these weapons need to be compared to the normally used weapons.

    Also, what changes do I need to do in the BiS list you posted earlier (google doc) for it to work with Humans expertise taken into consideration (I see Axes has an added Pawn note thing). Also, T4 2set and T5 4set, how much do these weigh?
    That's debatable, you could switch to 2x Boundless Agony daggers in T6 since Execute does not benefit from AP directly the only way to improve Execute dps is to improve rage generation and through Armor Ignore/Crit. Faster weapons give you smoother rage generation (making it more likely you can fit in execute on every GCD without delay) but they also consume more Flurry procs, effectively reducing Flurry uptime, but you're talking small margins.

    I'm not sure it's anything that's going to give you any particular big gains, if you do go for fast Execute weapons though try to keep those points in mind. Boundless Agony was popular for this because they have 210 armor ignore. There aren't really any good weapons worth using to pull this off in T4/T5 anyway so I would put it on the backburner for now, though personally I'm not going to bother.


    As for Human racial. Edit : Actually what I posted is wrong and not how the orc calculations are done, you have to account for main hand + offhand contribution of Expertise before doing this. Logging into a Human confirms at least sword is showing 5 expertise, so copy the Orc Axe expertise contribution numbers from the weapon notes over and add them to the pawn value.
    Edited: August 5, 2017

  11. I see. The sheet kinda worked in Excel, there was some warnings and I'm actually not sure how it's supposed to look. But like you said, a lot of work is needed to make it TBC compatible. Something I don't have the knowledge for sadly.

    As for the second question, what would I change in the googledoc in order to make it more accurate towards Human users in terms of expertise.

    Racial expertise isn't taken into account on the sheet as it's just a generic, it could be put into the sheet though. The axes do have notes to account for orc-racial though.
    Oh, and where do you play? Outland? H/A?

  12. I see. The sheet kinda worked in Excel, there was some warnings and I'm actually not sure how it's supposed to look. But like you said, a lot of work is needed to make it TBC compatible. Something I don't have the knowledge for sadly.

    As for the second question, what would I change in the googledoc in order to make it more accurate towards Human users in terms of expertise.



    Oh, and where do you play? Outland? H/A?
    I play Outland Horde. I haven't even looked into Alliance, was it 3 or 5 expertise they gain from Swords/Maces? If it's 5 like orcs then you can just take the number from the notes of the Axes for the Orc Racial and apply it to the Swords/Maces for the Human Racial.

  13. FuryDPS1.320b.ods should be the correct Landsoul spreadsheet version which we are looking for.
    http://web.archive.org/web/200810141...adsheet_2_4_a/

    Sadly the Filefront DL aren't available anymore.

  14. I feel like I should be ploughing through the last of old hard drives to see if It's still living on one of them, but the chances are unlikely. The sheets I have also don't work with Google Sheets due to no circular support and the Fury 2.0 beta one doesn't work in Open Office for some reason, it's likely that even if we got hold of it we might have trouble getting it to work.

    I've found 3 different versions of the sheet so far by scouring old hard drives (WOTLK prepatch, WOTLK launch, and 3.3.2) and also a combat rogue sheet from TBC but this one has evaded me. It amazes me that with all the people who used it back then it hasn't been preserved at all.

  15. Clearly you haven't read my posts and you're jumping to conclusions from your imagination, considering I actually said earlier in our discussion that Landsouls sheet works on averages, what is the point in this discussion if you're not even on the same page and following what is written? The context has long been lost at this point and this is going down the ****ter.
    All I'm asking for you is an argument why this spreadsheet is as garbage as you make it out to be. The only real flaw (apart from the Romulo Poison Vial which is a ****ty trinket and shouldn't be used anyway) is that it's not updated to 2.4 (offhand crits refreshing Flurry) and all that needs to be done is change 1 cell. You can't give an argument as to why the maths in this spreadsheet are worse than Landsoul. So tell me, why are the maths in Landsoul better and why should this sheet not be used?

First 1234 ... Last

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •