Powertech/Vanguard DPS Compendium

[Keren]

(03-02-2012 10:40 AM)blott Wrote:  @Deschaine: I would say power over all, adept mods are by far the best for us in terms of long run dps gains as we have crit increasing talents and we have more then enough aim on to increase crit as well. Unlike other classes all our mods are heavy main stat, and we get no heavy endurance/tertiary stat ones.

@thrakkemarn: what really is the point of doing this kind of test? Just like you admit in the post itself, "Whether these numbers are indicative of DPS vs level 50 enemies or what use they might be, I don't know, so take with a pinch of salt." So why talk about them? We have no real way of know how much armor the world bosses have, or how much raid bosses have still. I mean we have guesses atm but its all still very much theory and no way to prove anything still.

As for SR being negligible, with raid buffs (rakata stim / 5% aim) I gain 261 aim over unbuffed values. I don't understand how you could think that is negligible. I am still missing my rakata mh/pants and one last reflex 25 armoring for my bracers.

Just checked your statements with my numbers and Kray's sheet. Roughly 15% in favour for pyro (27/28) > ap (31). Also it seems very indeferent if you pick Steely Resolve, Prototype Cylinders or BurnOut, cause they differ only in few (< 10 dps). Though everyone should spec in a way, to best support their operation.

Moreover the sheet approved that power is after aim our best secondary stat, too.

[Cosmic Osmo]

(03-02-2012 10:16 PM)Keren Wrote:  Just checked your statements with my numbers and Kray's sheet. Roughly 15% in favour for pyro (27/28) > ap (31). Also it seems very indeferent if you pick Steely Resolve, Prototype Cylinders or BurnOut, cause they differ only in few (< 10 dps). Though everyone should spec in a way, to best support their operation.

Moreover the sheet approved that power is after aim our best secondary stat, too.

Let's all not forget that the game takes place in the game and a simulator takes place on paper. I think it's well established at this point that pyro/assault does more dps on paper and prototype/tactics is easier to execute in-game to match its paper dps.

There's little point arguing about whether tactics needs a buff until we have a combat log and people can truly compare for themselves which one works better for them in the game. By the time we get the combat log, we're also going to get more UI customizability to make pyro/assault less of a chore.

Personally, my main spec is tank. When I'm not needed to tank and we're short dps, I run tactics, because it plays almost exactly like tank spec, and allows for greater awareness. The dps checks aren't very hard to overcome in this game, so the raid is more likely to win if I have greater awareness than if I had 5-10% more dps.

[thrakkemarn]

(03-02-2012 10:19 AM)frmorrison Wrote:  In your original post you said 1.5% dps difference, but the above is a ~11% difference, which is close to my observations and Kray's simulation. When you saw the 1%, that indicated to me that something was wrong. My mistake (looks like the 1.5% in your OP was a typo), you appear to be a great player and theorycrafter. Thanks for the testing. AP needs some help in the talent tree.

Yes sorry. 11.5%. Brain fart, my bad >.<

(03-02-2012 10:40 AM)blott Wrote:  @thrakkemarn: what really is the point of doing this kind of test? Just like you admit in the post itself, "Whether these numbers are indicative of DPS vs level 50 enemies or what use they might be, I don't know, so take with a pinch of salt." So why talk about them? We have no real way of know how much armor the world bosses have, or how much raid bosses have still. I mean we have guesses atm but its all still very much theory and no way to prove anything still.

As for SR being negligible, with raid buffs (rakata stim / 5% aim) I gain 261 aim over unbuffed values. I don't understand how you could think that is negligible. I am still missing my rakata mh/pants and one last reflex 25 armoring for my bracers.

First of all, i'm just talking about 1 point in SR not all 3. So it's 3% Aim ... which I doubt is giving you 261 Aim. If you mean all 3 points give you 261, by giving up 1 point of SR you would lose 87 Aim. On my gear it's about 53 Aim ... which I consider negligible compared to 1% tech crit and extra damage below 30%. Maybe I'm wrong, but lets keep the numbers straight.

Now, the point of the test is that it's actual in game numbers from doing real dps on a real target in the game, as opposed to Simulations. I mean, you might as well ask "what is the point of theorycrafting or discussing things at all?" Frankly I don't know why anyone would not want more information to fill the gap until in-game combat logs. If you don't want to theorycraft, I suspect you're in the wrong place...

I posted the numbers because someone asked me to. While I don't think it is reliable for acquiring an absolute number for DPS, I do feel it is a fairly good indicator for comparing different specs and rotations. When one spec/rotation significantly outperforms another on the exact same target with the exact same gear etc. that is still more information than I had before. And as I said there are many world bosses so if I continue to do these tests or someone else does them we can check results across several tests.

[blott]

(03-03-2012 07:19 AM)thrakkemarn Wrote:  First of all, i'm just talking about 1 point in SR not all 3. So it's 3% Aim ... which I doubt is giving you 261 Aim. If you mean all 3 points give you 261, by giving up 1 point of SR you would lose 87 Aim. On my gear it's about 53 Aim ... which I consider negligible compared to 1% tech crit and extra damage below 30%. Maybe I'm wrong, but lets keep the numbers straight.

AHHHH sorry sorry I dunno why I thought you were talking about a 31 pyro spec. Not the 7/6/28! Depending on how far down the crit rating DR you are you are really valuing about ~30/35 crit rating over the 50/55 aim. While the aim wont make up for the amount of Crit in it is worth much much more in power.

(03-03-2012 07:19 AM)thrakkemarn Wrote:  Now, the point of the test is that it's actual in game numbers from doing real dps on a real target in the game, as opposed to Simulations. I mean, you might as well ask "what is the point of theorycrafting or discussing things at all?" Frankly I don't know why anyone would not want more information to fill the gap until in-game combat logs. If you don't want to theorycraft, I suspect you're in the wrong place...

In game observations are great when they are for things we can easily observe and repeat. I feel like damage done to a target is not one of those things that is easily done. I feel like lots of people have done these kind of tests and then use them as fuel against findings from sims. I'm not trying to say STOP WHAT you are doing, not at all! I guess I kinda came off that, sorry for that! I just feel like its REALLY subjective data, where I feel like the simulation data is at least based on factual game code.

While doing more in-game observational tests are an interesting idea I feel like we are mostly just testing our DPS as players vs the DPS differences in specific builds. Once again sorry if I came off a bit snarky there.

[thrakkemarn]

(03-03-2012 10:06 AM)blott Wrote:  In game observations are great when they are for things we can easily observe and repeat. I feel like damage done to a target is not one of those things that is easily done. I feel like lots of people have done these kind of tests and then use them as fuel against findings from sims. I'm not trying to say STOP WHAT you are doing, not at all! I guess I kinda came off that, sorry for that! I just feel like its REALLY subjective data, where I feel like the simulation data is at least based on factual game code.

While doing more in-game observational tests are an interesting idea I feel like we are mostly just testing our DPS as players vs the DPS differences in specific builds. Once again sorry if I came off a bit snarky there.

Yeah. I may have come off a bit defensive because I felt you were being abrasive, sorry for that. I fully welcome honest criticism or debate. Both sims and in-game tests have a place IMO. While I appreciate simulations a lot, I simply don't have the aptitude to break them down and know whether they are properly constructed. Which is why I like to have some sort of verification. While I admit that in-game tests can be highly subjective, a simulation is also subjective in that it assumes a perfect rotation and makes assumptions about target armor/dodge values etc. I am approaching this from the point of view of trying to apply theorycrafting and sims to the actual game, even if just for my own edification. And in this case, the simulation and my in-game dps tests lined up almost exactly, both supporting Pyro being ~11% more DPS than AP. Win!

[Keren]

Is there someone who wants to make a new guide? Seems that Sol isnt any longer active, thus we need someone else who can put our informations together. Most of the guide, like stas and so on doesnt changed.. I would appretiate an new up to date guideline for further discussions

[Turbopuschel]

(03-03-2012 06:54 PM)Keren Wrote:  Is there someone who wants to make a new guide? Seems that Sol isnt any longer active, thus we need someone else who can put our informations together. Most of the guide, like stas and so on doesnt changed.. I would appretiate an new up to date guideline for further discussions

It's already under way: http://sithwarrior.com/forums/Thread-Pow...9#pid15879

[kray]

(03-03-2012 12:04 PM)thrakkemarn Wrote:  Yeah. I may have come off a bit defensive because I felt you were being abrasive, sorry for that. I fully welcome honest criticism or debate. Both sims and in-game tests have a place IMO. While I appreciate simulations a lot, I simply don't have the aptitude to break them down and know whether they are properly constructed. Which is why I like to have some sort of verification. While I admit that in-game tests can be highly subjective, a simulation is also subjective in that it assumes a perfect rotation and makes assumptions about target armor/dodge values etc. I am approaching this from the point of view of trying to apply theorycrafting and sims to the actual game, even if just for my own edification. And in this case, the simulation and my in-game dps tests lined up almost exactly, both supporting Pyro being ~11% more DPS than AP. Win!

I'll take partial credit for that win, thankyouverymuch xD

But seriously, in-game testing is an incredibly useful thing to be doing. Data, no matter how statistically relevant, is still data -- I welcome anyone who's not on the road all month (damnit!) to let me know about any tests you've run so I can run them by my simulator.

Just some random bits of info from my sim working with Pyrotech. (probably going to post this version tonight, keep an eye on that thread):
Exotech Stim adds about 80 DPS (no surprise there)
Explosive Fuel adds about 40 DPS
Exotech Adrenal adds about 30 DPS
Bloodthirst adds about 20 DPS.

Now let's take a look here! I was super concerned and went over my code for Bloodthirst like a hawk, and couldn't find a problem. Finally I sat down and thought it over:

25% Crit from Explosive Fuel is huge, especially with the close-to-cap Surge rating that I was simming with, so that number makes sense to me.

The Exotech Attack Adrenal, according to my stat weights, will end up giving an additional ~500 DPS for those 15 seconds, but over a 5 minute fight, it can only be used twice. 500*(30/300) = 50 DPS, which is still high, but the sub-30% burn phase means our damage isn't distributed evenly, so that's where the discrepancy comes from. Additionally, CGC which makes up about a fourth of our damage, has a relatively low power coefficient, so it doesn't scale as well with the extra Power.

Now, Bloodthirst being so low, we have to look at it comparatively. Not only can it only be used once during any given fight, but that 15% damage is definitely not as good as Explosive Fuel, and I'd wager that it's around the same quality as an Adrenal. Clearly we can't be expecting Bloodthirst to be a huge personal DPS gain.

And as soon as I stop going AFK in the chat room I'll talk to GWARRR/Sol about a new post.

[kray]

So, I noticed something just now. Thermal Detonator isn't terribad, it's just not good yet.

I spent some time fooling around with stats on the sim in prep for writing up a new post, and noticed that Thermal Detonator has a gigantic coefficient on it. How much? 2.42. That's huge. Really huge. Every point of Aim or Power you get affects TD twice as much as most of your other abilities. With an Aim level of about 2400, 31pt Pyro only falls behind by 20 DPS, a 1.1% DPS loss.

Obviously this doesn't change anything we're looking at right now, but in the next tier of ops, assuming we don't see any modifications to the tree, we could inch pretty close to TD being a good attack! It scales OBSCENELY well.

Food for thought.

[goat4dinner]

just as a heads up according to swtor forums - raid bosses have 5% resist meaning stacking 105% accuracy is must for PVE content at least.

or am i wrong?