What factors cause real-world DPS to be lower than paper DPS
By
Arcanaville,
Posted on: 2011-03-05 16:49 in Scrappers
The level shift does not apply to the pylon.
As you guessed, the level shift doesn't apply to the pylon.
You list yourself most of the differences between "paper" DPS and real DPS - although I think most of these things can be calculated in paper DPS with minimal effort.
I've found Aid Self to reduce DPS more than slightly, with each use you're spending 4.5 (? I think interruption time is included in the animation, but I'll admit I never bothered to test and time) seconds doing absolutely no damage.
One thing you didn't mention (and that might not necessarily apply to you, but can to some others) is lag. Input lag, everyone gets some, and it's best to press clicks half a second before they actually go off, during the previous attack animation, in order to have everything going off as it should - but then, there's situations in which you can't always do that. Then there's lag spikes, which can lower DPS considerably, but are of course more obvious.
I don't believe the level shift applies to the pylon. I *think* it's like a giant monster in that regard. I could be wrong. I'm not thinking of anything other than the factors you mentioned that reduce (or raise) DPS. The 95% to hit sometimes works in more subtle ways, such as in the reduction of the utility of Achilles' Heel, combos and Follow Up type powers, but I don't think that matters on Fire.
Edit: LOL, no answer for two hours, then three of us answer in three minutes.
"That's because Werner can't do maths." - BunnyAnomaly
"Four hours in, and I was no longer making mistakes, no longer detoggling. I was a machine." - Werner
Videos of Other Stupid Scrapper Tricks
Thanks for the replies guys. My suspicion was that the level shift wasn't applying too, but it's good to have it confirmed.
Another small thing that might be affecting it is the arcanatime of the attacks.
I won't go into the big explanation of it here, but she points out that unless your next attack is queued, you're going to lose a server tick while the game is checking to make sure you're ready to use the next attack. I'd imagine over the course of the fight, that's going to change your output slightly.
Level 50 is a journey, not a destination.
▲Scrapper Issues List - Going Rogue Edition▲
Anytime you use a power a that isn't an attack is going to lose DPS.
Usually the on paper DPS calculations don't account for missing.
Lag will make a factor, even a little bit (when I ran the pylon test on my DB/WP, I noticed less lag when using a costume without a cape for instance).
BrandX Future Staff Fighter
The BrandX Collection
Quote:
I account for arcanatime in calculations and keep attacks queued so I'm not losing damage waiting to click a power. |
Originally Posted by Stupid_Fanboy
Another small thing that might be affecting it is the arcanatime of the attacks.
I won't go into the big explanation of it here, but she points out that unless your next attack is queued, you're going to lose a server tick while the game is checking to make sure you're ready to use the next attack. I'd imagine over the course of the fight, that's going to change your output slightly. |
On another (but similar) note, is there a way in Mids to have it show the global enhancement effect of a T3 or T4 incarnate boost without showing the effects of the level shift?
Quote:
On damage? Mids' reports damage for fighting even-level enemies. In other words, it's already showing you the numbers ignoring the level shift from your T3 or T4 abilities.
|
Originally Posted by firespray
On another (but similar) note, is there a way in Mids to have it show the global enhancement effect of a T3 or T4 incarnate boost without showing the effects of the level shift?
|
"That's because Werner can't do maths." - BunnyAnomaly
"Four hours in, and I was no longer making mistakes, no longer detoggling. I was a machine." - Werner
Videos of Other Stupid Scrapper Tricks
Quote:
If that's the case then I'm still confused as to why my DPS is so much higher than on paper. I'm calculating DPS from the mids numbers and not doing anything myself to account for the level shift. I had assumed that if you were level shifted, then mids gave damage out as if you were attacking a -1 enemy instead of an even-level.
|
Originally Posted by Werner
On damage? Mids' reports damage for fighting even-level enemies. In other words, it's already showing you the numbers ignoring the level shift from your T3 or T4 abilities.
|
Gaussian proc on, maybe?
Quote:
Lol, thanks for reminding me of that, but not this time. It's in build up and I don't calculate DPS with build-up running.
|
Originally Posted by Nihilii
Gaussian proc on, maybe?
|
Quote:
What is your damage setting in the effects & maths window of mids (options -> configuration -> effects and maths)? The default is for mids to calculate average damage for any attack - which means that for any attack that has less than a 100% chance for a DoT to go off it will calculate in some percentage of the damage. As every attack in fire melee but incinerate has bonus DoT damage that is on an 80% chance per tick, the damage mids shows you is going to be an average at best and I am not sure it is a good average. |
Originally Posted by firespray
If that's the case then I'm still confused as to why my DPS is so much higher than on paper. I'm calculating DPS from the mids numbers and not doing anything myself to account for the level shift. I had assumed that if you were level shifted, then mids gave damage out as if you were attacking a -1 enemy instead of an even-level.
|
For example, mids shows 57.81+5.005x3 damage for scorch (total 72.82). They arrived at this by multiplying the dot damage of 6.26 by 0.8 - however I am fairly sure that the fire melee dots stop ticking after a failure is rolled - which means that you don't get a full 80% of the dot damage, even on the average.
I could be incorrectly remembering how FM DoT's work - but the numbers in Mids are pretty clearly a simple average which is unlikely to show up in the game except over a VERY long run of attacks.
Globals: @Midnight Mystique/@Magik13
Quote:
An interesting point. I do have Mids show me the average damage for an attack, so the effect of procs and DoTs should be averaged out over time. I would think that soloing a pylon would take long enough that my damage would be fairly well averaged out over it. It took me about 9 minutes. My attack chain is 3 attacks, and takes about 6 seconds to execute, so that would be something like 250 attacks or so. That's not a huge statistical sample, but it is large enough that my DPS shouldn't drop 25 points just from procs and DoTs not going off.
|
Originally Posted by EricHough
What is your damage setting in the effects & maths window of mids (options -> configuration -> effects and maths)? The default is for mids to calculate average damage for any attack - which means that for any attack that has less than a 100% chance for a DoT to go off it will calculate in some percentage of the damage. As every attack in fire melee but incinerate has bonus DoT damage that is on an 80% chance per tick, the damage mids shows you is going to be an average at best and I am not sure it is a good average.
For example, mids shows 57.81+5.005x3 damage for scorch (total 72.82). They arrived at this by multiplying the dot damage of 6.26 by 0.8 - however I am fairly sure that the fire melee dots stop ticking after a failure is rolled - which means that you don't get a full 80% of the dot damage, even on the average. I could be incorrectly remembering how FM DoT's work - but the numbers in Mids are pretty clearly a simple average which is unlikely to show up in the game except over a VERY long run of attacks. |
What EricHough was getting at is that the DoT damage for fire attacks in Mids is shown to be above the true average damage.
Additionally, you mentioned yourself you didn't consider 95% maximum tohitchance, Aid Self, Hasten, PB, BU and critical hits. That's a lot of factors that can account for that 25 DPS difference, not just the DoTs.
The quick calc on tohit chance is that you'd lose 5%... 5% of 225 DPS is 12.5DPS... It's actually more then that for claws because some of those misses could be on follow-up, which reduces the +damage on other attacks (one approximation of this could be to add a .95 multiplier to the +damage you add for follow up stacking).
Regarding the other clicks... if you think of time lost in DPS-ing ... over a 9 minute fight, if your other clicks add up to just 25 seconds (3 aid selfs is 12 seconds), thats another 4.6% reduction (10.3DPS) not counting unstacking of follow-up accross the delay.
I guess what I'm saying is that with your assumptions, 225 DPS on paper gives around 200 DPS real world.... sounds like your calculations are pretty good.
I gotta make pain. I gotta make things right. I gotta stop what's comin'. 'Least I gotta try.
Quote:
Aid self accounts for about 15 seconds of additional time in the above calculations, PB about 6 seconds, Hasten about 4 seconds. That's 25 seconds of delay in the calculations. Build up is tricky because it depends on how you calculate around it, but in the above example Build Up accounted for over 16 seconds of time you were not dealing damage. That's 36 seconds of time with no damage.|
Originally Posted by firespray
I just recently got my Musculature Core Paragon boost on my Fire/SR scrapper and decided to see what my new DPS was like by soloing a pylon. I came up with just a hair over 200 DPS running a chain of GFS/Incinerate/Cremate.
Calculating the DPS on paper for the same build, same character, etc, I'm coming up with about 225 DPS. I'm wondering where the difference is coming from. My calculations are simple, and don't take into account the 95% to-hit cap or critical hits. They do include the level shift (I think), which may be what's causing the difference if the level shift doesn't apply to the pylon. The only other powers that I'm using which I didn't account for in my DPS calculations are: Hasten every 2 minutes Build up every 25 seconds or so Practiced Brawler every 2 minutes Aid self was used 3 times over the course of the 9-minute or so fight. I would think Hasten, Aid self, and practiced brawler would lower my DPS slightly, while Build Up would increase it. Any other thoughts on what causes the differences between real-world DPS and paper DPS? |
That plus the 95% tohit chance can theoretically account for about 17 dps of the 25 dps discrepancy, depending on how the calculations for Build Up were incorporated. The difference is just 8 dps, which is often accountable by the occasional lag glitch and by not queuing attacks with perfect mechanical consistency.
[Guide to Defense] [Scrapper Secondaries Comparison] [Archetype Popularity Analysis]
In one little corner of the universe, there's nothing more irritating than a misfile...
(Please support the best webcomic about a cosmic universal realignment by impaired angelic interference resulting in identity crisis angst. Or I release the pigmy water thieves.)
So it sounds like there's not really one major factor that's causing my DPS to drop, it's probably a combination of several small factors, most of which there's very little I can do to mitigate.
I appreciate all the help guys.
I just recently got my Musculature Core Paragon boost on my Fire/SR scrapper and decided to see what my new DPS was like by soloing a pylon. I came up with just a hair over 200 DPS running a chain of GFS/Incinerate/Cremate.
Calculating the DPS on paper for the same build, same character, etc, I'm coming up with about 225 DPS. I'm wondering where the difference is coming from. My calculations are simple, and don't take into account the 95% to-hit cap or critical hits. They do include the level shift (I think), which may be what's causing the difference if the level shift doesn't apply to the pylon.
The only other powers that I'm using which I didn't account for in my DPS calculations are:
Hasten every 2 minutes
Build up every 25 seconds or so
Practiced Brawler every 2 minutes
Aid self was used 3 times over the course of the 9-minute or so fight.
I would think Hasten, Aid self, and practiced brawler would lower my DPS slightly, while Build Up would increase it.
Any other thoughts on what causes the differences between real-world DPS and paper DPS?