Are you sure Normalisation is working as intended?
Switching from harbinger of doom to iblis (same attack speed) I get almost the same DPR values despite one being a dagger and the other a sword which should result in very different normalizations (1,7 for daggers and 2,4 for swords).
Currently with the values selected for my character i get 19.70 DPR with harbinger and 19.65 with Iblis.
Redoing the calculation but changing the normalization to its correct value yields.
15.37 DPR for Harbinger which is a substantial DPS loss.
For Iblis the recalculated value is 19.6 which is the same as in the spreadsheet.
This leads me to believe that in the latest version of the spreadsheet weapons are not subject to their proper normalization speeds which skews the DPS results.
Normalization isn't working for daggers, I only have it working for 2H and non-dagger 1H. It would require abit of extra work to get that, but since I'm probably going to do that for specific weapon skill, it will eventually go in.
With all the (fairly redundant) gear and spec related questions that have been popping up on these boards the last couple weeks, I'd really like more people to use this spreadsheet as a reference and framework for discussion.
Any chance to look at the divide by 0 errors that seem to pop up with 2H weapons in the latest version? I haven't been able to trace them yet.
Vanick, sorry about that, I copy and pasted the previous cell that was "ExeCD", and the name went with it. For now, you can select E58, go to Insert -> Name -> Define put in "ExeCD", and hit enter, should work then.
I posted that in the previous page. It'll be fixed in the next release, sorry for the goof up.
Added Big Execute field, Execute Spam Field, and DPS while executing field(thanks celandro)
Added calculation for using excess rage on HS(thanks celandro)
Added Threat Per Second column for each DPS stage
Fixed Item Haste calcs(they were additive, should be multiplicative)
That #REF error SHOULD be fixed. Technically, there is a circular reference in my spreadsheet, but I'm having trouble figuring out how to fix it. YellowConnects relies on ability cooldowns, Crusader relies on YellowConnects, and just about everything else in the sheet relies on strength. Let me know if you're having problems.
Celandro, on your suggestion for Total DPS, I decided not to implement that exactly. The way you had it caused a slight discrephancy. It basically didn't count for the 1.5s that the big execute takes.
Things I'm working on for next release:
Specific weapon skill(should fix normalization for daggers as well)
Weapon switch for execute spam
Modeling activated abilities for optimum usage(i.e. DW + Recklessness during execute spam)
Oh, question for anyone. Do procs from HoJ/WF/SS eat flurry charges? I seem to remember reading that somewhere. Similarly, I assume they can proc flurry if the extra attack crits? Just noticed that I'm currently not including those.
Nice changes deathwing.
What my numbers imply is that the fight ended due to your execute killing the mob. Your numbers assume the fight ends just before your next execute(or maybe halfway in between). Both numbers are valid.
I was working on some if statements for the big execute that would allow people to turn it off since its so hard not to waste rage saving up for that big execute but its going to get tricky if you are going to model DW + recklessness. Also I believe DW and recklessness both trigger the global cooldown so the first execute happens 3s after deathwish.
Would you be interested in having me work on threat per second for tank gear + abilities + stance + tank spec? Its quite a bit of work adding in the tank gear and the various new abilities to spend rage on.
I was looking over the spreadsheet and playing around with it and noticed something that seemed unusual to me.
Basiaclly what I noticed was an increase in DPS with edgemaster's and iblis over maladath. I would think that +11 weapon skill would offer greater dps increase (both white and overall) as damage reduction from glancing at that point is ~0% whereas having Iblis OH would only give 7 to swords giving ~9% damage reduction from glancing.
Am I missing something?
"Doubt is the thief that often makes us fear to tread where we might have won.” —William Shakespeare
I was looking over the spreadsheet and playing around with it and noticed something that seemed unusual to me.
Basiaclly what I noticed was an increase in DPS with edgemaster's and iblis over maladath. I would think that +11 weapon skill would offer greater dps increase (both white and overall) as damage reduction from glancing at that point is ~0% whereas having Iblis OH would only give 7 to swords giving ~9% damage reduction from glancing.
Am I missing something?
Some possible explanations:
Well maladath is only giving you +3skill if you are using edgemasters when it comes to glancing blows which is very very roughly 3% total dps. Also as your heroic strike usage goes up, your glancing reduction on your MH attacks goes down since yellow attacks dont glance.
The higher dps, faster speed, +1hit, +1crit, and +26ap is giving you more than a 3% boost.
Also if you are using another 1.6 speed weapon you gain a bit on flurry with matching weapon speeds and if you are using TF or another procing weapon, the damage lost from glancing blows is lower. If you are using a slow MH the Iblis OH is increasing your average execute speed as well since sometimes that OH swing (or 2 OH swings) will let you execute even though your MH hasnt hit.
I was looking over the spreadsheet and playing around with it and noticed something that seemed unusual to me.
Basiaclly what I noticed was an increase in DPS with edgemaster's and iblis over maladath. I would think that +11 weapon skill would offer greater dps increase (both white and overall) as damage reduction from glancing at that point is ~0% whereas having Iblis OH would only give 7 to swords giving ~9% damage reduction from glancing.
Am I missing something?
Some possible explanations:
Well maladath is only giving you +3skill if you are using edgemasters when it comes to glancing blows which is very very roughly 3% total dps. Also as your heroic strike usage goes up, your glancing reduction on your MH attacks goes down since yellow attacks dont glance.
The higher dps, faster speed, +1hit, +1crit, and +26ap is giving you more than a 3% boost.
Also if you are using another 1.6 speed weapon you gain a bit on flurry with matching weapon speeds and if you are using TF or another procing weapon, the damage lost from glancing blows is lower. If you are using a slow MH the Iblis OH is increasing your average execute speed as well since sometimes that OH swing (or 2 OH swings) will let you execute even though your MH hasnt hit.
So then that would explain why I was getting a higher increase in dps over iblis without edgemaster's as I was getting more out of the sword skill, correct?
"Doubt is the thief that often makes us fear to tread where we might have won.” —William Shakespeare
So then that would explain why I was getting a higher increase in dps over iblis without edgemaster's as I was getting more out of the sword skill, correct?
Yes, about 30% more (you still gain some hit and crit and less dodge/miss/parry from the 1 extra skill)
So then that would explain why I was getting a higher increase in dps over iblis without edgemaster's as I was getting more out of the sword skill, correct?
Yes, about 30% more (you still gain some hit and crit and less dodge/miss/parry from the 1 extra skill)
Cool, thanks for the help!
"Doubt is the thief that often makes us fear to tread where we might have won.” —William Shakespeare
Nice changes deathwing.
What my numbers imply is that the fight ended due to your execute killing the mob. Your numbers assume the fight ends just before your next execute(or maybe halfway in between). Both numbers are valid.
I was working on some if statements for the big execute that would allow people to turn it off since its so hard not to waste rage saving up for that big execute but its going to get tricky if you are going to model DW + recklessness. Also I believe DW and recklessness both trigger the global cooldown so the first execute happens 3s after deathwish.
Would you be interested in having me work on threat per second for tank gear + abilities + stance + tank spec? Its quite a bit of work adding in the tank gear and the various new abilities to spend rage on.
Didn't think of that scenario, which do you think is more accurate?
I don't think the first execute happening 3s later is a problem, I usually just time my DW around 22% or 21%.
Can you expand on this a bit more? I can add som tank gear and some tank abilties, and use Kenco's mod for the threat values, that part is easy. The problem with calculating a tank's threat is that it's so dependant on the incoming damage. I really don't tank much, is there a safe pattern of abilities to assume, regardless of boss encounter?
Your discussion about HS reminded me of one thing. I don't think I calculate the added benefit of using HS over a white attack(no glancing, no DW hit penalty), just the extra damage. I'll need to incorporate that into the next release.
Awesome spreadsheet. I've had some fun playing with it the last 3 days at work haha. Not that we didn't know that +weapon skill owns but it's crazy to see that the hakkar sword set as human beats everything until Naxx. And the huge boost being orc/human gives you.
Nice changes deathwing.
What my numbers imply is that the fight ended due to your execute killing the mob. Your numbers assume the fight ends just before your next execute(or maybe halfway in between). Both numbers are valid.
I was working on some if statements for the big execute that would allow people to turn it off since its so hard not to waste rage saving up for that big execute but its going to get tricky if you are going to model DW + recklessness. Also I believe DW and recklessness both trigger the global cooldown so the first execute happens 3s after deathwish.
Would you be interested in having me work on threat per second for tank gear + abilities + stance + tank spec? Its quite a bit of work adding in the tank gear and the various new abilities to spend rage on.
Didn't think of that scenario, which do you think is more accurate?
I don't think the first execute happening 3s later is a problem, I usually just time my DW around 22% or 21%.
Can you expand on this a bit more? I can add som tank gear and some tank abilties, and use Kenco's mod for the threat values, that part is easy. The problem with calculating a tank's threat is that it's so dependant on the incoming damage. I really don't tank much, is there a safe pattern of abilities to assume, regardless of boss encounter?
Your discussion about HS reminded me of one thing. I don't I calculate the added benefit of using HS over a white attack(no glancing, no DW hit penalty), just the extra damage. I'll need to incorporate that into the next release.
Yes the issue is modelling deathwish + recklessness so that 3s of deathwish is before 20% and 1.5s of recklessness is before 20%. Also you can sometimes use deathwish 2 or 3 times in a fight (patchwork can be 3 but to not pull agro you should fetish it)
As far as tanking stuff, you can assume spamming shield block then a 3 cycle rotation that is revenge every 5s and a 30 point ability + a sunder, then revenge + sunder + 30point, then revenge + sunder + sunder. Then add an incoming damage per second input which would be added to the total rage per second available (this could be used for regular dps too) and use HS to use all the rest of the rage. Due to the way revenge works, its almost always better to just wait for it than try and squeeze in more than 2 other abilities in between revenges.
As far as HS, take a look at my spreadsheet, I think its accurate. You may already be handling it correctly but Im not sure.
For the trinket cooldown field, I'm assuming it is meant to be changed manually (ie. when selecting Earthstrike it manually needs to be changed to 120)?
I was trying to decide between Earthstrike and HoJ, and they were reletively even until I changed the cooldown for ES to 120, in which case HoJ pulled ahead by a fair margin, with my gear anyway.
It seems you broke the 2H field.
Last version you could enter a arbitrary value for speed of the offhand to remove any div/0 errors but that is no longer possible.
It seems you broke the 2H field.
Last version you could enter a arbitrary value for speed of the offhand to remove any div/0 errors but that is no longer possible.
Yep. Everything else is pretty damn accurate. I plugged in everything as best I could to simulate a Patchwerk fight, then killed him on Tuesday and the spreadsheet predicted my DPS to within three-tenths of a point of the actual value I got ingame (I took screenshots, will post them later).
Very impressive job. Thank you for all the effort you've put into this.
Plugged in exactly what buffs were used, exact fight time, talent setup, and gear. I'd be satisfied with a DPS estimate +/- 10 of an actual value, but holy shit.
Just curious about the CD column for the list of abilities right under RagePS. Are we supposed to manually enter data into those? Foe example you have an equation for Heroic Strike, but shouldn't that just be a look up to the cell with the speed for the MH?
"Doubt is the thief that often makes us fear to tread where we might have won.” —William Shakespeare
It seems you broke the 2H field.
Last version you could enter a arbitrary value for speed of the offhand to remove any div/0 errors but that is no longer possible.
Yep. Everything else is pretty damn accurate. I plugged in everything as best I could to simulate a Patchwerk fight, then killed him on Tuesday and the spreadsheet predicted my DPS to within three-tenths of a point of the actual value I got ingame (I took screenshots, will post them later).
Very impressive job. Thank you for all the effort you've put into this.
Plugged in exactly what buffs were used, exact fight time, talent setup, and gear. I'd be satisfied with a DPS estimate +/- 10 of an actual value, but holy shit.
Thats very impressive, kudos to Deathwing and everyone else thats contributed. Sadly my DPS always comes out higher on the sheet than in practice :P but then again, most fights arent like Patchwerk. WTB BoS.
It seems you broke the 2H field.
Last version you could enter a arbitrary value for speed of the offhand to remove any div/0 errors but that is no longer possible.
Damnit, again? I guess I'll have to put in those if statements ruro suggested.
Yes, polocabbit, any column or row with a green shaded header is user input. I put equations in HS and Execute Spam because it allows the user to easier compare gear. I'm assuming people will want to use only excess rage on HS, and spam execute as fast as possible. You're welcome to change the values, just copy the equations elsewhere to save them(don't copy the cell, go in and highlight equation, that way you don't move the name of the cell).
I'm glad my spreadsheet is that accurate. Patchwerk seems to be about the only fight though where you could expect that accuracy. Agro, mobility, server performance, and human reponse time are mitigating factors in most other fights.
Celandro, I thought more about the HS...perhaps it is modeled correctly. That tank ability pattern seems easy enough to program in, but still, what do you assume for incoming damage? Does rage gained from monster damage count as agro, btw?
08/07/06, 7:33 AM
Similar Threads
