Jump to content

  • Curse Sites
Help
- - - - -

Detailed DPS calculations -- need some help

dps power theorycraft precision min/max mechanics

  • Please log in to reply
5 replies to this topic

#1 swarmofseals

swarmofseals

    Fahrar Cub

  • Members
  • 22 posts

Posted 28 August 2012 - 02:22 AM

Does anyone know what the actual damage formulas for GW2 are? The stat system is somewhat more opaque than what I'm used to. I'm a bit of a theorycrafter/min-maxer at heart and I usually like to get an understanding for the way stat systems work in a game. Specifically, I'd like to be able to figure out exactly how Power and Condition Damage affect the overall damage calculation. Do they add a fixed amount of damage? A percentage? Does character level matter like it seems to with Precision?

Any insight would be helpful! I suspect the GW2 wiki covers some of this but the site is down for maintenance and has been down for quite a while. Who knows when it will be back.

#2 swarmofseals

swarmofseals

    Fahrar Cub

  • Members
  • 22 posts

Posted 28 August 2012 - 03:32 AM

It seems that the GW2 Wiki reappeared briefly and I was able to get a plaintext version of the Damage and Power pages. Unfortunately, both commented that the exact formulas are currently unknown.

So I did some very cursory research on my own and this is the data that I collected.

For my level 17 warrior, I checked the damage tooltips for the five greatsword abilities with base power, then with For Great Justice! and then with For Great Justice! and active Signet of Might

My base power was 170. With For Great Justice! (tooltip: +35 power) my power rose to 203. With both active signet and For Great Justice! (total tooltip: +65 power), my power rose to 231. So power increases were 0%, 19.4% (actual)/20.6% (tooltip), and 35.8% (actual)/38.2% (tooltip) respectively.

I did some rounding with the % increases, fwiw. All bonuses are relative to baseline.

Base Case:

Ability 1: 77/77/99 (three attacks)
Ability 2: 488/121
Ability 3: 77
Ability 4: 83
Ability 5: 188

For Great Justice!:

Ability 1: 92/92/119 (+19.5%/+19.5%/+20%)
Ability 2: 576/145 (+18%/+20%)
Ability 3: 92 (+19.5%)
Ability 4: 99 (+19%)
Ability 5: 224 (+19%)

For Great Justice! + Signet of Might (Active):

Ability 1: 105/105/135 (+36%/+36%/+36%)
Ability 2: 656/165 (+34%/+36%)
Ability 3: 105 (+36%)
Ability 4: 112 (+35%)
Ability 5: 255 (+36%)


Conclusions: There do appear to be some minor variations in the bonuses that different skills get from power, but the increases do appear to be fairly direct. An increase to your power of +x% will increase your skill damage tooltips by about the same percent. Now this is hardly conclusive data -- I only ran the numbers for one class and one weapon at a specific level (17). It would be awesome if people could look at similar data for other classes and other weapons and post it. It also might be helpful to run the numbers by manipulating power in different ways (ie: removal of armor) to see if power from armor works differently than power from buffs (I doubt it but you never know)

Diminishing Returns: It immediately jumped out at me that the buffs were not actually increasing my power by the amount advertised. The first +35 power actually gave me +33 power. +65 total power actually gave me +61 power.

My base power at level 17 with gear but no buffs is 143. With Signet of Might (passive) it goes up to 170 (no diminishing returns). With For Great Justice! it goes up to 203 (-2 off tooltip), with Signet of Might (active) and For Great Justice it goes up to 231 (-4 off tooltip). There are clearly some diminishing returns in effect there.

If I remove my gear, I go to 104 power base. That number is exactly what it should be given that my gear advertises +39 power total. With Signet (passive) it is 131. With FGJ! it is 164. With FGJ and Signet (active), it is 192. So the first +27 has no diminishing returns again, the next +35 has the same -2 diminishing returns and the full buff has -5, which is actually more of a loss than when I had my gear equipped!

So based on those observations, I think it's safe to assume that diminishing returns are factored in as a reduction to the stat itself rather than to the effectiveness of the stat. Again it's a little bit of a leap given that I'm only looking at one very specific set of numbers, but I bet more data will back this up.

It also seems that gear, at least at the level that I have it, does not have diminishing returns. It is entirely possible that with better gear and more deliberate stat stacking you will see diminishing returns from gear. There is clearly some sort of diminishing returns from buffs. I do not know if bonuses from trait lines have diminishing returns.


TL;DR Version:
  • All observations are based on a very small set of data, so take it with a big grain of salt. More data would be appreciated if anyone feels like doing the work.
  • Increasing your Power by X% will increase the tooltip damage of your skills by approximately X%
  • Bonuses to Power from gear and/or trait lines may or may not have diminishing returns
  • Bonuses to Power from buffs almost certainly have diminishing returns that are independent of your total amount of stat. The only thing that really seems to matter for buff diminishing returns is the total bonus from buffs.

EDIT:  Gear bonuses to power may have diminishing returns. I hit level 18 and found a greatsword with +20 power. Equipping that on top of my existing gear resulted in an increase of +19 power rather than the expected 20. I removed some of my other gear and the greatsword gave the expected +20 once I was below a certain number. So it seems that there are diminishing returns from gear. Whether those diminishing returns are based on an absolute number or if it has to do with level is unknown.

Edited by swarmofseals, 28 August 2012 - 04:53 AM.


#3 Zinn

Zinn

    Sylvari Specialist

  • Members
  • 621 posts

Posted 28 August 2012 - 06:06 AM

View Postswarmofseals, on 28 August 2012 - 03:32 AM, said:

EDIT:  Gear bonuses to power may have diminishing returns. I hit level 18 and found a greatsword with +20 power. Equipping that on top of my existing gear resulted in an increase of +19 power rather than the expected 20. I removed some of my other gear and the greatsword gave the expected +20 once I was below a certain number. So it seems that there are diminishing returns from gear. Whether those diminishing returns are based on an absolute number or if it has to do with level is unknown.

I'm guessing you were down leveled, just a guess though.

#4 Kaylos21

Kaylos21

    Vanguard Scout

  • Members
  • 232 posts
  • Location:Near Chicago
  • Server:Borlis Pass

Posted 28 August 2012 - 11:53 PM

Another thing to consider, is that it is hard to do exact DPS calculation without know the animation/cast timers on abilities.  For instance on warrior, I know sword hits for less than mace.  But my instinct is telling me that it hits faster than mace.  I am not really sure though and it could be the animation itself throwing me off.

#5 golden_radish

golden_radish

    Vanguard Scout

  • Members
  • 226 posts
  • Location:gw2

Posted 29 August 2012 - 01:44 PM

Yep, what Kaylos21 said.  Without casting times, and millisecond accurate log-to-a-file combat log, including all effects, all attempted effects, casting start, casting end, pet damage, condition damage, etc..  we're just randomly guessing.

I suspect ArenaNet has done this intentionally to prevent people from calling them out on tooltip inaccuracies and class power disparity.

A shame really, considering how well it can be done, that they chose to do it so poorly.

#6 swarmofseals

swarmofseals

    Fahrar Cub

  • Members
  • 22 posts

Posted 29 August 2012 - 06:49 PM

View Postgolden_radish, on 29 August 2012 - 01:44 PM, said:

Yep, what Kaylos21 said.  Without casting times, and millisecond accurate log-to-a-file combat log, including all effects, all attempted effects, casting start, casting end, pet damage, condition damage, etc..  we're just randomly guessing.

I suspect ArenaNet has done this intentionally to prevent people from calling them out on tooltip inaccuracies and class power disparity.

A shame really, considering how well it can be done, that they chose to do it so poorly.

To be honest I'm not actually looking for actual DPS numbers. What I am looking for is the effect of stats on skill damage.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users