Jump to content


Photo

Moonkin DPS Spreadsheet


  • This topic is locked This topic is locked
1925 replies to this topic

#21 Lorewanderer

Lorewanderer

    Moof.

  •  Patrons
  • 2,360 posts

Posted 16 May 2007 - 07:10 AM

My latency isn't that bad usually. I try to take a little breather and get a tic of spirit regen in the middle of the first SF cast.

So it goes

00 IS
03- SF
06- SF
09- SF
12 IS ...

with 6 seconds (minus lag, plus .5 s if the previous one crit) between IS casting and paying mana for the fist in the next cycle, allowing a slight bit of regen if I'm lucky. Or am I off base?

#22 Boevis

Boevis

    Bald Bull

  • Members
  • 1,284 posts

Posted 16 May 2007 - 02:20 PM

My latency isn't that bad usually. I try to take a little breather and get a tic of spirit regen in the middle of the first SF cast.

So it goes

00 IS
03- SF
06- SF
09- SF
12 IS ...

with 6 seconds (minus lag, plus .5 s if the previous one crit) between IS casting and paying mana for the fist in the next cycle, allowing a slight bit of regen if I'm lucky. Or am I off base?


That's actually a pretty interesting idea...

IS/MF cast, initiating the FSR, but the GCD lasts another 1.5+Lag seconds, and a talented-but-not-NG'd Starfire is 3 seconds, so 4.5 seconds + lag before the server would register you casting another spell and restart the FSR. If I understand you correctly, you wait another ~1 second after the IS/MF to get a normal mana regen tic in before the starfire actually casts.

This is highly conditional though, if your regen tic doesn't occur soon enough, you end up waiting far too long and lose a lot of DPS. There's also the issue that if all you are doing is IS SFx3 every 12 seconds, this is a pretty poor DPS rotation by itself unless you take advantage of NG to add a wrath or MF, which you can't really do here because you are taking a good chunk of time waiting for the regen tic. There's also the issue that this cycle is really easy on the mana, as such waiting for the regen really shouldn't be necessary even with low end gear ... maybe in Kara where you aren't likely to get more than 3 useful buffs (Int, BoW, DS or VT)

#23 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 16 May 2007 - 04:41 PM

That actually doesn't seem that bad. Probably similar to my no-consumables and/or crappy buffs rotation of IS:SF(rank6)x4 in terms of total DPS & mana efficiency, with the advantage of never missing IS ticks (and the debuff for the sake of the tank/tank's healers) but the disadvantage of relying on manual timing of the starfire following the IS. I'll look at trying to model it.

#24 Bezu

Bezu

    Glass Joe

  • Members
  • 2 posts

Posted 16 May 2007 - 04:59 PM

It seems weird that spell hit has almost no effect on DPS, in some cases more +hit gives you less dps =S

If this is true I will take the 2 points out of balance of power and change all my spell hit gear for crit =P

#25 Boevis

Boevis

    Bald Bull

  • Members
  • 1,284 posts

Posted 16 May 2007 - 05:04 PM

It seems weird that spell hit has almost no effect on DPS, in some cases more +hit gives you less dps =S

If this is true I will take the 2 points out of balance of power and change all my spell hit gear for crit =P


That can't be correct, more +hit will never reduce your DPS, at worst it will return a 0% change when you reach the 16% cap vs level +3 mobs. There's an error with the spreadsheet if you're ever getting a reduction from increasing stats.

#26 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 16 May 2007 - 05:50 PM

Please provide screenshots of what you're seeing here, so I can troubleshoot it.

#27 Bezu

Bezu

    Glass Joe

  • Members
  • 2 posts

Posted 16 May 2007 - 06:04 PM

Check the "starfire only" box

Posted Image

With 0% hit, no buffs and "infite mana"

Posted Image

With insane amount of hit%

#28 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 16 May 2007 - 06:16 PM

I blame Caliane.

But seriously, it seems to be an error that increases the estimated cast times as your non-crit, non-miss increases for both Starfire & Wrath. This is because of my attempts to model the belief that crit% is static, and not a fraction of spells that hit, but a fraction of all spells--which I still haven't seen data to support one way or another. But, I digress.

I'm not properly penalizing spell miss in the cast time calculations. I'll get that sorted out & repost soon.

#29 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 16 May 2007 - 06:40 PM

Version 0.67

Changed calculations of estimated cast times to properly penalize spell miss. Most settings will show a SLIGHT DPS DROP as a result of these changes.
Cloaks now available on gear select.

#30 Nobbynob Littlun

Nobbynob Littlun

    Von Kaiser

  • Members
  • 57 posts

Posted 19 May 2007 - 11:53 PM

Apologies for sidetracking the thread, but it seemed as good a place as any to ask this.

Suppose moonkin druids are typically limited by threat rather than mana. It is said that only 50% of amount-healed is converted into threat,.


My question:

-Assuming it is a single-target encounter, I hypothesize it would be viable to "trim" your threat by spending some time healing raid members. Perhaps this would allow a healer to throw down a nuke or two, but let's not assume that.

-How would you theorize this hypothesis? That is, if you are nuking at n% of your potential DPS due to a threat limit, how much time spent healing will reduce your potential DPS to the point that it matches your actual DPS? More trickily, how would you tie this proportion of DPS-to-Healing into the amount of +spellpower and/or +spellcrit you have?

If this question doesnt make sense, let me know.

#31 Lorewanderer

Lorewanderer

    Moof.

  •  Patrons
  • 2,360 posts

Posted 20 May 2007 - 04:09 AM

I haven't actually done the math to back it up, but I'm going to say with the upcoming patch we'll be decidedly less threat limited and more mana limited with the changes to shadow priests and the addition of balance spells to subtlety.

#32 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 20 May 2007 - 04:34 AM

I haven't actually done the math to back it up, but I'm going to say with the upcoming patch we'll be decidedly less threat limited and more mana limited with the changes to shadow priests and the addition of balance spells to subtlety.

The consumables change is also going to be huge in making mana a limiting factor moreso/again. Very few Balance Druids are actually going to "require" 5 pts (or more than 2, for that matter) in Subtlety, unless they can't get BoSalv and/or are putting up with a poor tank.

My question:

-Assuming it is a single-target encounter, I hypothesize it would be viable to "trim" your threat by spending some time healing raid members. Perhaps this would allow a healer to throw down a nuke or two, but let's not assume that.

-How would you theorize this hypothesis? That is, if you are nuking at n% of your potential DPS due to a threat limit, how much time spent healing will reduce your potential DPS to the point that it matches your actual DPS? More trickily, how would you tie this proportion of DPS-to-Healing into the amount of +spellpower and/or +spellcrit you have?

It's a pretty complex question that you can consider this way:

1) What's the most TPS I can put out before I surpass the tank, or risk surpassing the tank with a string of crits? (Note: If you're ~50 TPS below the tank, by 2 minutes into the fight you'll be practically incapable of pulling aggro on a crit chain.)
2) Generating my maximum TPS (wait 10 sec, tank's TPS - 50 for 2 minutes, then my TPS exactly the same as the tank) how much mana would I end the fight with?
3) How much healing can I do with that mana?

As mentioned, the consumables, shadow priest, and subtlety changes are likely going to skew things in favor of DPSing 100% of the time.

#33 Caliane

Caliane

    Piston Honda

  • Members
  • 105 posts

Posted 20 May 2007 - 02:28 PM

Finally got a copy of excel to check the spreadsheet out.

Im no expert on spreadsheets.
But for modeling t4 2pc, its based on offensive casts, so the mana regen would have to be tied to the cast rotation. 5% per cast of the rotation.


Btw, its been a long long time since I was in high school or college statistics.
With 95% chance to hit with spells, what is the chances of not missing out of 100 casts?


Oh, and, What exactly am I suppsed to enter for mp/5?
Mp/5 from gear? or mp/5 while casting as shown on the char sheet which includes intensity mp/5?

#34 Caliane

Caliane

    Piston Honda

  • Members
  • 105 posts

Posted 20 May 2007 - 02:33 PM

Apologies for sidetracking the thread, but it seemed as good a place as any to ask this.

Suppose moonkin druids are typically limited by threat rather than mana. It is said that only 50% of amount-healed is converted into threat,.


My question:

-Assuming it is a single-target encounter, I hypothesize it would be viable to "trim" your threat by spending some time healing raid members. Perhaps this would allow a healer to throw down a nuke or two, but let's not assume that.

-How would you theorize this hypothesis? That is, if you are nuking at n% of your potential DPS due to a threat limit, how much time spent healing will reduce your potential DPS to the point that it matches your actual DPS? More trickily, how would you tie this proportion of DPS-to-Healing into the amount of +spellpower and/or +spellcrit you have?

If this question doesnt make sense, let me know.



The threat issues balance face are less overall aggro issues, and more aggro during sensetive times.

Things like nightbane where he drops aggro, or the first 30 seconds of tghe fight where chain crits will cause you to pull.

I haven't looked at any warrior threat spreadsheets, but from experiance I would suspect 1100 threat per second is probably normal?

Under normal dps conditions I dont really have to hold back. its just the first few seconds, if I crit too much, or on deaggro fights.
Although I might have to reconsider that statement under the effects of extra mana regen from priests, manatide. Unlimited mana supply might make aggro and issue that Im not really seeing atm.

#35 Boevis

Boevis

    Bald Bull

  • Members
  • 1,284 posts

Posted 20 May 2007 - 06:16 PM

The only way a warrior can maintain over 1k TPS is as a fury DPS warrior, Prot tanks really can't manage that right now, 900 or lower is a better estimate.

The chance to never miss out of 100 casts with 95% hit/crit is going to be .95^100 = .59% That's in the realm of nearly impossible.

#36 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 20 May 2007 - 10:56 PM

Im no expert on spreadsheets.
But for modeling t4 2pc, its based on offensive casts, so the mana regen would have to be tied to the cast rotation. 5% per cast of the rotation.

Oh, and, What exactly am I suppsed to enter for mp/5?
Mp/5 from gear? or mp/5 while casting as shown on the char sheet which includes intensity mp/5?

Yeah, both JoW & T4 2 pc. (and the mana restore meta gem for that matter) are going to be rotation specific. I (possibly) need to redesign the spreadsheet to model them entirely accurately. To a lesser extent, spellsurge & Bangle procs also fall into the "rotation specific effect" category.

The mp5 is supposed to be what your character sheet shows. It then checks intensity & dreamstate talents to figure out how much those two talents are contributing (which will change with your buffs) & assumes the rest is from your gear.

#37 Caliane

Caliane

    Piston Honda

  • Members
  • 105 posts

Posted 22 May 2007 - 07:14 PM

The only way a warrior can maintain over 1k TPS is as a fury DPS warrior, Prot tanks really can't manage that right now, 900 or lower is a better estimate.

The chance to never miss out of 100 casts with 95% hit/crit is going to be .95^100 = .59% That's in the realm of nearly impossible.


Right of course. I zoned out when considering the 1100 or so dps I needed to do to pull and forgot to consdier the 130% bubble for ranged dps.
Which would have left that number around 850 threat per second.

that .59% doesnt sound right..

#38 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 23 May 2007 - 04:56 AM

that .59% doesnt sound right..

Consider a 20 sided die (D20). Roll it once (or choose any number from 1-20). Now roll it 100 more times without that number coming up again. That's your odds of going 100 casts without a miss with only 95% hit.

Mathematically, it's expressed as 0.95^100 (or 95 to the 100th power), which works out to 0.0059205292203340254829249648824032, or ~0.592%.

#39 Efejel

Efejel

    Piston Honda

  • Members
  • 246 posts

Posted 25 May 2007 - 04:28 AM

Version 0.68 Now Available

Updated items to reflect 2.1 changes
Added Neck & Rings
Added Nature/Arcane specific items (including Spellfire)
Added new items & gems -- SPELL HASTE ITEMS HAVE HASTE IGNORED
NO MECHANICS CHANGES - If you don't want to use gear select, you don't need to download this version



Additional to do:
Figure out how to properly incorporate spell haste.

It's starting to look like the massive spell rotation set-up needs to be done to get everything working properly. I'm not sure I care enough at this point. Anyone interested in taking this up should contact me by PM for suggestions!

#40 Dexter

Dexter

    Glass Joe

  • Members
  • 8 posts

Posted 25 May 2007 - 04:20 PM

1st off - awesome spreadsheet! Very nice work indeed!

Seems that whenever I try to use the Windscale Hood on the 'Gear Select' tab, it errors out and replaces all values with #N/A.


-Dex




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users


    Google (1)