Moonkin DPS Spreadsheet - Page 90

mader · 05/01/08, 5:13 PM

yah imp ff is a required talent in my opinion. i am spoiled in the fact that I almost always have a spriest, but even if i didn't i should be able to sustain on any fight with mana pots, using my own innervate, and drums of resto if absolutely necessary. generally, if i am running oom the fight is lasting longer than it should anyways.

spi · 05/01/08, 6:08 PM

Well according to the WWS on brut we had, imp FF only would have increased raid dps by 70 which I make up more than enough by not casting it. I don't see why this would be a "requirement" unless tanks had threat issues. To get the required calculation I went back and added all the dps that would have increased had imp FF would be up for all the rogues enh shamans and hunter pets for the entire night for large enough sample. On average it was less than 1% dps increase for rogues/shaman/warrior and not very significant for the hunter pets.

You can check my math on spreadsheet here
imp_FF_calculation.xls - FileFront.com

Edit: link fixed.

Lazi · 05/01/08, 6:33 PM

I'm not downloading your spreadsheet because I just don't understand how what you are saying makes any sense. With Imp FF, all the Physical DPSers except maybe a Fury Warrior would have been wearing different gear, so how do the stats in WWS with a different gearset help you to calculate the benefit of iFF?

spi · 05/01/08, 6:48 PM

Originally Posted by Lazi

How do you work out how much Imp. FF would have added from WWS exactly?

First of all, melee are on a 1-roll system. Therefore, even if you improve chance to hit by 3%, those attacks that can now be hit are not crittable. There may be things that I have glanced over so please correct me if you think I'm wrong.

Now what you do is, you go through WWS and look for the breakdown on melee (rogue/warrior/enh shaman). We are only concerned with swings that can be improved by IFF, which will be rogue white dmg, warrior white damage and MS, and shaman white damage. Everything else is at capped hit with the case of my particular WWS so i ignored it. Take the number of hits the melee had and take the average damage of their swings. Let number of hits be x and average damage y. Their number of hits would have been x/0.97 with IFF. If you multiply this number by y, you will get the total damage they would have done with white hits with imp FF. If you do y*(x/.97- x) it will let you get the damage difference that imp FF has made.

spi · 05/01/08, 6:49 PM

Originally Posted by Lazi

I'm not downloading your spreadsheet because I just don't understand how what you are saying makes any sense. With Imp FF, all the Physical DPSers except maybe a Fury Warrior would have been wearing different gear, so how do the stats in WWS with a different gearset help you to calculate the benefit of iFF?

All of the rogues have at least 4% chance to miss on their whites. Also, none of them will be wearing low enough hit to actually have misses on their specials even if they were wearing different set with imp FF. However, To make this simple you can say that I'm assuming no one make a gear change due to IFF. I'm just doing a quick case study on our 20+ wipe brut WWS and how much imp FF would have made the difference keeping everything else equal.

mader · 05/01/08, 6:50 PM

Originally Posted by spi

Well according to the WWS on brut we had, imp FF only would have increased raid dps by 70 which I make up more than enough by not casting it. I don't see why this would be a "requirement" unless tanks had threat issues. To get the required calculation I went back and added all the dps that would have increased had imp FF would be up for all the rogues enh shamans and hunter pets for the entire night for large enough sample. On average it was less than 1% dps increase for rogues/shaman/warrior and not very significant for the hunter pets.

You can check my math on spreadsheet here

imp_FF_calculation.xls - FileFront.com

link doesn't work.

besides, all i really mean is dreamstate is largely useless for me while raiding. the melee compenent of any raid generally shifts a lot which would make imp ff less/more effective depending on the fight and how your group is going about killing it.

spi · 05/01/08, 6:58 PM

Originally Posted by mader

link doesn't work.

besides, all i really mean is dreamstate is largely useless for me while raiding. the melee compenent of any raid generally shifts a lot which would make imp ff less/more effective depending on the fight and how your group is going about killing it.

Sorry the link should be fixed. Please look it over with the WWS I posted earlier to see if I had made any errors.

P.S. I've assumed that it would add 3% hit to both MH and OH attacks. However, all of our rogues have enough hit that imp FF will only affect their OH damage, which makes the dps increase from imp FF significantly lower than I stated above. So it would give our raid at most 70 dps, and much lower in reality. ** edit: MH/OH attacks both receive DW penalty as mader pointed out below. Please don't be confused.

mader · 05/01/08, 7:42 PM

Originally Posted by spi

Sorry the link should be fixed. Please look it over with the WWS I posted earlier to see if I had made any errors.

P.S. I've assumed that it would add 3% hit to both MH and OH attacks. However, all of our rogues have enough hit that imp FF will only affect their OH damage, which makes the dps increase from imp FF significantly lower than I stated above. So it would give our raid at most 70 dps, and much lower in reality.

dual-weild penalty applies to all white attacks, regardless of which hand. specials are still capped ~9% tho.

also after a quick look, your number of attacks is wrong based off incorrect reading of wws.

Landed Norm Dot Crit Glanc Crush All Miss Miss Resist Block Parry Dodge
Nb 4990 3195 2378 1795 359 274 19 66
% 64 % 41 % 30 % 23 % 4.6 % 3 %

if you mouse over the 4990 on wws, it will specifically read as non crits, therefore you need to add crits into that so it would be 4990 + 2378 to get the total number of attacks landed (glancing and normal are added together for the initial 4990). repeating this for the other chars in your spreadsheet yeilds > 100 raid dps gain but it still seems off a bit. and also you ignore your tanks damage/threat as well.

also, your raid is extremely caster heavy so yea def not gonna get max benefit from imp ff with that setup.

Lazi · 05/01/08, 8:01 PM

I won't lie and say that I understand any maths past basic arithmetic but I downloaded your spreadsheet and saw you had it setup for 3 Rogues, a Warrior & an Enhancement Shaman which is by no means even close to all the physical damage in a non-stacked raid. Hunters? Feral Druid, either DPS'ing or Tanking? Prot Warrior, damage is quantifiable even if the extra threat is less so. Still not convinced that iFF is anything but a requirement for a raiding Moonkin.

spi · 05/01/08, 8:08 PM

There are no dpsing feral druids and hunters are all hit capped for that specific WWS. Like I said this is not a perfect model, but I accounted for all significant physical dps that benefits from IFF. For each rogues the dps increase was about 16dps, arms warrior 11dps, and enh shaman 8dps. I think it's needless to say that bear druid, pets, and prot warrior will be much less so. I'm sorry for not accounting for all physical damage and i'm not claiming this to be a general case for every raid. I'm just pointing at the specific raid that I was in that required extensive amount of steady dps that also had large amount of data that I could work with to be reliable.

This was a case-study. Not a generalization by any means.

spi · 05/01/08, 8:09 PM

Originally Posted by mader

dual-weild penalty applies to all white attacks, regardless of which hand. specials are still capped ~9% tho.

also after a quick look, your number of attacks is wrong based off incorrect reading of wws.

Landed Norm Dot Crit Glanc Crush All Miss Miss Resist Block Parry Dodge
Nb 4990 3195 2378 1795 359 274 19 66
% 64 % 41 % 30 % 23 % 4.6 % 3 %

if you mouse over the 4990 on wws, it will specifically read as non crits, therefore you need to add crits into that so it would be 4990 + 2378 to get the total number of attacks landed (glancing and normal are added together for the initial 4990). repeating this for the other chars in your spreadsheet yeilds > 100 raid dps gain but it still seems off a bit. and also you ignore your tanks damage/threat as well.

I see what I missed. Thank you.

Saraya · 05/02/08, 12:04 AM

Originally Posted by spi

According to earlier discussion of the trinket in this thread, you cannot approximate the value of the trinket by its uptime since it is proced AFTER starfire has been fired. Therefore, you need to subtract at least 2.5s everytime it procs, on top of that you need to subtract the time between the last starfire fired and the time the buff runs out. Just doing a rough approximation, you'd have to subtract 15x2.5(avg starfire cast time)=37.5 and about 15x.5=7.5 .5s per proc (the time between last starfire cast during the proc and the proc running out in middle of the starfire cast) as an effective uptime making the effective trinket up time 177.4s. This is just an approximation. However, I am very interested in finding out exact effectiveness of the trinket so someone please correct me if i'm being wrong. I suppose I can go through the entire combat log to find out every occurance of the proc and get the exact times that the starfire isn't affected by the proc...

So i'm proposing (177.4/353)*150= 75.4dmg as effective dmg from the trinket procs according to the log given.

Reread the trinket discussion you are referencing, you missed several key posts. You might want to start with http://elitistjerks.com/708795-post1161.html

There's no reason to subtract 2.5 seconds for every proc. Only the time left over at the end is wasted, none of it is wasted from the beginning. It works just like every single other effect in the game, just think about how the clicky from Icon of the Silver Crescent is calculated. As an extreme example: Assuming your starfire is 2.5 seconds cast, if you had a click trinket that gave you +1000 damage for 2.5 seconds, by your calculations, the trinket is worth 0 damage.

spi · 05/02/08, 12:17 AM

Originally Posted by Saraya

Reread the trinket discussion you are referencing, you missed several key posts. You might want to start with http://elitistjerks.com/708795-post1161.html

There's no reason to subtract 2.5 seconds for every proc. Only the time left over at the end is wasted, none of it is wasted from the beginning. It works just like every single other effect in the game, just think about how the clicky from Icon of the Silver Crescent is calculated. As an extreme example: Assuming your starfire is 2.5 seconds cast, if you had a click trinket that gave you +1000 damage for 2.5 seconds, by your calculations, the trinket is worth 0 damage.

And precisely it would be worth 0 damage if your starfire cast time is more than 2.5 seconds. Spell damage is applied at the end of the cast so unless you moonfired right after clicking on the trinket, that theoretical trinket that you discussed above would be worth 0 spell damage. I suppose I didn't interpret the way nakedduck did, but I believe you should subtract starfire cast time off beginning of the procs regardless. Like you pointed out, +10000 damage click trinket that buffs for 1.5s is worth 0 spell damage on average. I kind of understand logical leap i'm making because everyone doesn't subtract 2.5s off all the trinket effects, but i think it's clearer now that if you have a 1 minute cooldown trinket with +10000 dmg click with 1.5s cooldown it would be worth 0 spell damage instead of averaged out 250.

Radixx · 05/02/08, 12:27 AM

Even then, it's not quite as simple as removing 2.5s from each proc. Suppose you have zero haste, then if you refresh Moonfire either immediately after the proc, or after the two subsequent Starfires, then it should still be affected by the proc, as 2xSF and 1xMF would only be 7.5s worth of cast times + gcd.

With 0 haste and a 1xMF, 4xSF rotation then you have a 75% chance of the MF dot expiring (and thus being refreshed) within the 8s proc duration.

spi · 05/02/08, 12:29 AM

Originally Posted by Radixx

Even then, it's not quite as simple as removing 2.5s from each proc. Suppose you have zero haste, then if you refresh Moonfire either immediately after the proc, or after the two subsequent Starfires, then it should still be affected by the proc, as 2xSF and 1xMF would only be 7.5s worth of cast times + gcd.

With 0 haste and a 1xMF, 4xSF rotation then you have a 75% chance of the MF dot expiring (and thus being refreshed) within the 8s proc duration.

You're definately right. What i'm saying is obviously only the case of starfire. I suppose I oversimplified again.