Mage Bomb - DPS analysis

[Deleted]

edit (21/03/2013):

• 3 targets graph corrected
• files updated with 5.2
• result pushed to 110% haste (on my files only)

edit (20/12/2012):

• graph update with 10k sp
• add DPET graph
• add AOE dps graph

-------------------

This post is here to discuss about the mage bomb's DPS in regard of the haste.
All calculations was based on patch 5.1 (13/12/12 hotfixes) with 10k spell power.
All calculations was made on Excel (I will upload it when I have time). Haste is calculated with a resolution of 0.1%.

As pure mathematical calculation, it can be considered that: rotation is perfect, there's no lag, etc.

First: single target

• Frost Bomb is clearly ahead (+18%)
• Nether Tempest show a linear progression because the duration is also fluctuating with a maximum duration for each new ticks.
• Living Bomb show a better DPS just before each new ticks. Because the explosion is equal to 50% of the global damage, the fluctuating duration have a greater impact than added ticks.

Second: 2 targets

• Frost bomb is far behind (-38%) because it can only be cast once.
• Nether tempest take the lead as expected
• Living Bomb bad scaling is increase by the number of target receiving the explosion.

Third: 3 targets

Living bomb is far ahead until 37.5% haste where 3 Nether Tempest is closing the gap.

Damage Per Execute Time - 1 & 3 targets


dps AOE - 5 to 8 targets

[mrgreenthump]

You should prolly redo this with actually calculating the time used to cast the bombs aswell, as frostbomb is a single cast 1-1.5 secs while getting 3x NT will take from 3 to 4.5 secs and getting 3 lb is from 1-1.5 secs as fire to 2-3 secs with fireblast glyph to 3-4.5 secs without fireblast glyph..

[reflection]

Did you take into account that frostbomb can hit multiple targets? You only need 1 GCD to apply it? What about Brainfreeze procs if you are Frost? I think it's not that easy

I don't think frostbomb sucks that much on a 3 target fight. It gets even better with more enemies.

[Deivid]

Something is clearly wrong with your calculations, living bomb's DPS can't get lower as you get new ticks...

Edit: ok, I think I undestand why, you are saying living bomb's total duration gets lower with more haste so you get more explosions overal and when you hit the next tick the duration resets back to the original time and thus less explosions... Seem reasonable in paper but I don't think it would be that big of an issue in real situation. Keep in ming LB is easier to keep a high uptime because you can refresh it at <3s and not lose anything where NT you lose ticks unless you let it run out completely.

Edit2: NVM I was thinking of DPET not DPS...

[Deleted]

Your calculations are always good and trustworthy Nathyiel, but I think you have to take into account multiple cast times, lack of use BF procs (overwriting) due multiple casts and so on.

*Edit* Frost Bomb only 18% ahead single target? That couldn´t be or my calculations are super worng? Do you want to share your excact calculations with me?

[mrgreenthump]

Did you take into account that frostbomb can hit multiple targets? You only need 1 GCD to apply it? What about Brainfreeze procs if you are Frost? I think it's not that easy

I don't think frostbomb sucks that much on a 3 target fight. It gets even better with more enemies.

Well yes it does get better and yes its wrong to compare 1x frostbomb to 3x LB directly unless its a fire mage.. Oh and your post got me thinking aswell that something is wonky here, cause when 3 target cleave it should look like LB= 3x dot damage and 9x explosion damage, NT is 4.5x dot damage, Frost bomb is 2x explosion damage. I cant put my finger on what is wrong from right away but yee its not looking right as LB is doing over 10x the damage with 3 targets(which it shouldnt) compared to single target. Or then the numbers on the left side just arent right..

1 thing what it might be is that on the 3 target LB its both lb and the explosion on all 3 counted and then x6 which would be 2x more damage than it should be, but cant see the function so just guessing here

[Deleted]

For Frost bomb, I have taken into account the cast time.
For multiple cast (NT & LV), the GCD isn't part of the calculation because it's DPS, not DPET (I have it somewhere in the file).
I have take into account tat Frost Bolt can it multiple targets but I have also take into account that the explosion won't damage the primary target.

Living Bomb damage isn't lower after each tick, just the DPS. DPS = damage/duration. The duration fluctuate between 11.5s and 13.5s (more precise duration tomorrow).

I have to recheck before showing AOE (5+ targets) and BF proc, I will relook into 2/3 targets.

edit : it's dps of the bomb, only. No BBF proc or trinket, etc.
When I have finish rechecking some part, I will upload the file so every one can look into it. Before the end of the week.

edit 2 : it isn't damage or DEPT (I will show both later).

[Deleted]

My calculation for Frost bomb and NT single target are:

FB: 3746 / (10s + 1.5s) = 325.73913 DPS
NT: 3180 / (12s + 1.5s) = 235.55555 DPS

325.73913 DPS x 100 / 235.55555 DPS = 138,28548%
235.55555 DPS x 100 / 325.73913 DPS = 72.31417%

Starting from Nether Tempest DPS (100%) FB is about ~38% higher
Starting from Frost Bomb DPS (100%), FB is about ~28% higher

Is there a fault in my calculations maybe? Or did u take Lhiveras numbers? Because his numbers where calculated before the FB buff (late in beta). Frost Bomb where only ~18% higher before the buff

[Deleted]

I have used WoWdb/WoWhead number as I think they are the most up to date.

With 0 haste, we have:

• FB 1x [(3286+246.2%) + (1643+143.1%)]
• NT 12x [(232+17.4%) + (117+8.7%)] = (2784+208.8%) + (1404+104.4%)
• LV 4x [(347+26%) + (1395+104.5%)] = (1388+104%) + (1395+104.5%)

For 10k spell power, total damage on 1 target:

• FB 3286+24620=27906
• NT 2784+20880=23664
• LV 1388+10400+1395+10450=23633

on 2 targets:

• FB 27906+1643+14310=43859
• NT 1.5*23664=35496 (max damage)
• 2xNT 2*35496=70992
• LV 1388+10400+2*(1395+10450)=35478
• 2x LV 2*35478 = 70956

on 3 targets:

• FB 27906+2*(1643+14310)=59812
• NT 35496
• 2NT 70992
• 3NT 106488
• LV 1388+10400+3*(1395+10450)=47323 (max damage)
• 2LV 94646
• 3LV 141969

DPS = damage / duration

• FB: damage/(1.5+10)
• NT: damage/12
• LV: damage/12

For NT and LV, the GCD roll while the DOT is already ticking !!!

The only things that I'm not sure is for SP calculation. Is this like factor (10k*2.462) or percent (10k*(1+2.462)) ?
I think it's more a factor as it give 64.8k for FB witch is what I have on my mage without buff (~25k sp).

[Deleted]

NT dont have the 12s + 1.5, in an ideal world you refresh it in the last <1 second to a perfect match
NT: 3180 / (12s) = 265 DPS

There is a difference between NT and LB in that NT has a normal GCD and LB only a 1 second one is that what you are trying to emulate CitizenPete?

On my naked gnome:
NT does 3240 damage over 12 seconds (1.5 second GCD)
LB does 3240 damage over 12 seconds (1 second GCD)
FB does 3825 damage over 10 seconds (+ GCD cast time for a total of 11.5)

[Deleted]

You can´t refresh NT @ 10,5, because the spell is instant applied. If you play like this, you will lose a tick (probably 2). So NT has to to be 12 + 1,5

*Edit" You get the last Tick at 12 seconds with 0% haste, so rolling with gcd is a loss of 2 ticks.

[Deleted]

This is why I made my calculation in theory, not reality.
I can't calculate human time reaction, lag and other things like that.

[Deleted]

Yeah, but even a robot can´t refresh ealier : D The last tick is @ 12 seconds

[Deleted]

So a refresh right before the last tick between the 11th and 12th will result in a 13 tick new version... the current rolling tick + the fresh NT just like all other DOTs

[mrgreenthump]

I have used WoWdb/WoWhead number as I think they are the most up to date.

With 0 haste, we have:

• FB 1x [(3286+246.2%) + (1643+143.1%)]
• NT 12x [(232+17.4%) + (117+8.7%)] = (2784+208.8%) + (1404+104.4%)
• LV 4x [(347+26%) + (1395+104.5%)] = (1388+104%) + (1395+104.5%)

For 10k spell power, total damage on 1 target:

• FB 3286+24620=27906
• NT 2784+20880=23664
• LV 1388+10400+1395+10450=23633

on 2 targets:

• FB 27906+1643+14310=43859
• NT 1.5*23664=35496 (max damage)
• 2xNT 2*35496=70992
• LV 1388+10400+2*(1395+10450)=35478
• 2x LV 2*35478 = 70956

on 3 targets:

• FB 27906+2*(1643+14310)=59812
• NT 35496
• 2NT 70992
• 3NT 106488
• LV 1388+10400+3*(1395+10450)=47323 (max damage)
• 2LV 94646
• 3LV 141969

DPS = damage / duration

• FB: damage/(1.5+10)
• NT: damage/12
• LV: damage/12

For NT and LV, the GCD roll while the DOT is already ticking !!!

The only things that I'm not sure is for SP calculation. Is this like factor (10k*2.462) or percent (10k*(1+2.462)) ?
I think it's more a factor as it give 64.8k for FB witch is what I have on my mage without buff (~25k sp).

Mm these numbers look a bit more correct than the graphs, as in your graphs the 3 target x3 LvB did 12 times the damage as the single target LvB

[Deleted]

I will recheck my files but it's the same number and formula.
The only difference is here are the damage and the graph show the DpS (for 30k sp). I will redo it with 10k sp for a better comparison.

For the haste, I calculate the expected duration of one ticks first, then the total duration with added tick. And I have made multiple comparison in game for validation.
I was really surprise when I see no plateau or breakpoint with NT.

[thehordemage]

What trinkets were used for this testing? The overall damage increase you get would change if you used LoC and a DoT bomb.

[Deleted]

What trinkets were used for this testing? The overall damage increase you get would change if you used LoC and a DoT bomb.

Please, don't skip the comment: it's a mathematical analysis witch only count the bomb in it self . I have skipped all other factor except spell power and haste.
And the validation of the haste calculation was made with gear only, no proc nor buff.

[Polarthief]

You should prolly redo this with actually calculating the time used to cast the bombs aswell, as frostbomb is a single cast 1-1.5 secs while getting 3x NT will take from 3 to 4.5 secs and getting 3 lb is from 1-1.5 secs as fire to 2-3 secs with fireblast glyph to 3-4.5 secs without fireblast glyph..

LB is actually always 1s "cast" because it has a -.5s GCD, meaning Inferno Blast/Fire Blast only saves you .5-1s

FB: 1-1.5
NT: 3-4.5
LB: 3
LB w/ FB Glyph: 2-2.5


These charts seem... wrong.

Also, you're comparing 1 FB to 2-3 NT/LB. That loss in globals will hurt your DPS overall. One global for a Frostbolt can result in 40-100k extra hit, which would completely eliminate the extra GCDs required for LB/NT. ALSO, if Frost Bomb crits, it instantly doubles. To double LB/NT, you would need all 12 ticks (per target)/all 4 ticks AND explosion (per target, on each) to crit, which is much, much more unlikely. The charts don't really help like, at all, unless you're somehow assuming 0 crit and just the pure Bomb dps without taking into account more/less FFBs for Frost, more/less AM proccs for Arcane, and the loss of globals from addition NT/LB. While I definitely appreciate the theorycrafting, the only useful chart there is the Single target one.

Although, I will say 3 target LB is accurate for Fire only, as IB is a staple of the rotation.

[Tickspoon]

Crit is irrelevent since average DPS is what matters.

For single-target, this graph overvalues Frost Bomb since it doesn't account for increasing the time spent casting.

eg. 6 minute fight, 25% haste
NT Cast Time: 360/12 * 1.5/1.25 = 36 seconds

FB Cast Time: 360/(11.5/1.25) * 1.5/1.25 = 47 seconds

Difference: (47 - 36)/360 = 3% of DPS time lost

Since FB usually is ~16% of total DPS and NT ~13%, the 3% more damage is approximately cancelled out by the extra cast time.