Please upgrade your browser for the best possible experience.

Chrome Firefox Internet Explorer

Boss-Adjusted DPS Balance

STAR WARS: The Old Republic > English > Classes
Boss-Adjusted DPS Balance

KeyboardNinja's Avatar

09.25.2013 , 05:16 PM | #1
Now that the set of DPS scalars has been completed, we can merge the theory crafting with the world DPS leaderboard to get some idea of how the specs are balanced against each other. Note that not all specs are represented on the world DPS leaderboard. In order to fill in the gaps, I'll be pulling data from The Ebon Hawk DPS leaderboard, which has the notable property of representing nearly every spec.

The numbers you see below are the boss-adjusted values together with their listed spec, ranked from top to bottom. The best parse for a given spec is chosen and adjusted. All specs (except Sabo/Eng) are represented. This list is not kept up to date or maintained. I'm just putting the data here to facilitate the discussion to follow.
  1. 3597.14 - Gorband - Annihilation Marauder
  2. 3535.06 - Evrydayimsmggln - Hybrid Gunslinger
  3. 3459.89 - Invinc - Lethality Operative
  4. 3433.71 - Handcuff - Telekinetics Sage
  5. 3409.33 - Ranick - Carnage Marauder
  6. 3354.07 - M-nightrider - Balance Shadow
  7. 3341.92 - Pizza'da'hutt - Pyro Mercenary
  8. 3340.79 - Nithnuro - Marksman Sniper
  9. 3322.893 - Tianna - Lethality Sniper
  10. 3247.68 - Mruniverse - Hybrid Vanguard
  11. 3233.87 - Marisi - Arsenal Mercenary
  12. 3210.85 - Pelara - Balance Sage
  13. 3111.18 - A'jantis - Focus Sentinel
  14. 3107.50 - Vepres - Vigilance Guardian
  15. 3049.35 - Aerre - Infiltration Shadow
  16. 3047.994 - Drekkin - Assault Vanguard
  17. 3035.00 - Kishekzun - Focus Guardian
  18. 2780.15 - Ylira - Tactics Vanguard

A couple caveats here. First, data from the world DPS leaderboard is somewhat irregular because they allow parses of arbitrary length, so long as it is at least 300 seconds. This is problematic from the perspective of measuring class balance because different classes have different cooldown timings. 5 minutes exactly is a rare sweet spot that is actually balanced across all classes. 30 seconds longer and we start running into issues with Consulars/Inquisitors of all specs getting an advantage. 30 seconds shorter and we enhance the value of Inspiration/Bloodthirst. The next nearest sweet spot is at 17 minutes, so needless to say, 5 minutes is ideal. The Ebon Hawk DPS leaderboard is normalized to this value, but the world leaderboard is not, thus world leaderboard parses can be slightly inflated relative to other classes and specs.

Another caveat is that these parses are all done with disparate gear. I happen to know the Tactics Vanguard is in full 72s save for the armorings, which are mostly 69s. This contrasts with Evryday, who is in nearly full 75s. Still, it's about as close as we can get.

Additionally, the world leaderboards allow the use of the Double SA bug, which artificially inflates some specs (most notably, Watchman/Annihilation). It's not known when (or if) bioware will be rectifying this issue.

Finally, I don't have a Scrapper/Concealment parse. The best I could find was Kalaen's, which is too old to be reputable. I couldn't get my hands on Invinc's 3.1k Concealment parse, so until we can find something good, I'll leave it off the list.

Overall Observations

Leaving off Tactics (which is getting super-buffed in 2.4) and Assault (which is getting slightly buffed), the gap between the highest DPS parse and the lowest is 15.63%. This is, however, comparing a purely single-target spec with no burst (Watchman/Annihilation) to an AoE spec with very on-demand burst (Focus/Rage). If we compare apples-to-apples, we get a gap of 13.61% between Watchman/Annihilation and Vigilance/Vengeance. Discarding Watchman/Annihilation for the moment due to the double SA issue as well as Evryday's parse due to the gear disparity, we get to compare Invinc's Lethality parse vs Vepres's Vig parse, which gives us a spread of 10.16%. I feel pretty confident in saying that this is the DPS spread in the present game balance where single-target sustained DPS is concerned.

Guardian/Juggernaut DPS

I think it's safe to say that Vigilance/Vengeance is right now under-powered. It does enough damage that people can make it work (the world first clears of both TfB and S&V NiM used a Jugg DPS), but I wouldn't call it "balanced".

In addition to the vast DPS spread between Vig and Lethality, consider the spread between Focus Sentinel and Vig Guardian. Remember, Focus Sentinels are AoE specs with a lot of on-demand burst. They aren't designed to rank highly on the sustained single-target DPS meters. This is in contrast to Vigilance/Vengeance, which is a sustained single-target DPS spec with fairly long ramp-up and minimal burst. And yet, Vig is actually behind Focus Sentinel in terms of single-target DPS (by about 0.1%). That seems very, very broken. It means that, if you're currently running a Vig Guardian and you want to improve your single-target DPS, you can have them roll a Focus Sentinel to improve not only your single-target DPS, but also your sustained AoE DPS and your single-target burst! That seems very unfair.

Note that Focus Guardians are about 2.33% behind Vig Guardians in terms of single-target DPS, which still seems a bit too close (I'd like to see Vig farther ahead), but it's at least balanced in the right direction.

Verdict: Vig/Veng badly needs a buff, and a significant one at that. They should be parsing at least 5-7% above what they are right now. Especially when you consider the enormous RNG that is baked into the spec and the difficulty of achieving consistent results.

Sentinel Specs

Watchman vs Combat is an ever-present debate on the sentinel forums. However, as you can see, Watchman is actually ahead of Combat by a fairly significant margin. Specifically, 5.22%! Now, some of that can be written off as the double SA proc rearing its head, but even with that, it seems that Watchman is at least somewhat ahead of Combat in terms of single-target sustained DPS, exactly as the developers claim. This is significant, because it means that Combat is not the "god spec" of the Sentinel class. Combat has almost no downtime penalties and fantastic burst, while Watchman has nearly no burst and is severely punished by any downtime at all. The single-target sustained damage though appears to make up for it, and I would say that, except on fights where Merciless/Annihilate stacks are likely to fall off due to downtime, Watchman is probably just as viable as Combat from a raid min-maxing point of view.

Focus is the red-headed stepchild of the Sentinel class. Almost nobody plays it in PvE (my guildmaster is an odd exception), and yet it is universally considered to be over-powered in PvP. Interestingly enough, the PvE DPS spread is not as far as is commonly thought. Focus appears to parse about 8.75% behind Combat. However, Ranick is in nearly full 75s, while A'jantis just has a KD relic. If we adjust for the gear disparity, we're probably talking about something closer to 5%. That's a heck of a lot closer than most people think, and it means that Focus is actually a viable option even for bosses with extremely tight enrage timers when the boss has any AoE component whatsoever. In other words, I would have taken a Focus Sentinel to pre-nerf Nightmare Dread Guard and I wouldn't have asked them to respec.

Verdict: Extremely well-balanced. Don't touch.

Mercenary/Commando Specs

I think, much to everyone's surprise, Assault/Pyro spec seems to be doing extremely well these days. It has long been a highly repudiated spec, and rightfully so pre-2.0. However, post-2.0, the numbers look very much in Assault's favor. Specifically, Assault does 3.34% better than Gunnery/Arsenal on a single-target sustained boss. Given Gunnery's low-setup cost and solid burst, this seems like a very well balanced combination. With that said, Assault is still 5.46% behind the best Gunslinger parse in nearly equivalent gear, which seems like too wide of a spread.

Thus, while I think that Gunnery and Assault are well balanced against each other, the class as a whole could use a buff to bring it more in line with the current king of the ranged, Gunslingers.

Verdict: Buff commandos/mercenaries across the board, but only by a small amount. Unless hybrid gunslinger is performing over targets, in which case commandos are just fine.


This is the category which impressed me the most. Telekinetics/Lightning appears to be a very viable spec for a top-tier raid group. It parses only 2.87% behind the best Gunslinger parse, which is a really impressive buff relative to where things were pre-2.0. When you consider the fact that Telekinetics is a very AoE-heavy spec with almost no setup time and decent burst (relative to hybrid gunslinger, which has AoE but very long setup times), that seems like exactly the sort of margin we would want to see. In other words, if I had a choice between taking Handcuff or taking Evryday to a raid, it would come down to what class utilities fit better with the rest of my group. Their DPS is precisely in line. (hilariously, they're both in the same raid group, so…yeah)

Balance/Madness seems to fall distressingly behind though. Even with the double SA bug, Balance is 6.49% behind Telekinetics. That seems very wrong, given that Balance has a lot of setup (it is a very DoT-heavy spec), next to nothing in the way of burst, and comparatively poor AoE options. I would expect Balance to parse better than Telekinetics by a few percentage points, especially with the bugged relics. Thus, while TK is in a really good place relative to Gunslingers, Balance needs a pretty serious buff.

Verdict: Buff Balance by about 7-8%. Telekinetics is perfect.


There is no class in the game more roundly derided as being uncompetitive in top-tier PvE. People generally point to the leaderboards and throw out numbers like the 3.7k Sabo parses vs the 2.7k Balance parses. Fortunately, the boss-adjusted truth seems to be much closer to balanced (though still not quite).

Balance spec for shadows is directly analogous to Watchman spec for sentinels. They're both melee, they both have long ramp-up times, and neither of them have strong AoEs. Balance does have *better* AoE than Watchman does, and it is punished less for being off a boss for a protracted period of time, but these are fairly marginal factors. I would expect Balance to parse about 2-3% behind Watchman on a single-target boss with executes and buffs.

Currently, the margin is 6.76%. That's a lot closer than the ~20% figure that most people would cite, but it's still too far spread. Balance Shadows need a buff to bring them more in line, perhaps 3-4%. Such a buff would still leave them dummy parsing much lower than sentinels, but it would bring them up to the point where a sentinel and a shadow of equal skill would be a roughly even trade for a DPS group, depending on requisite utility.

Infiltration/Deception shadows are in another category all their own. Infiltration/Deception parses 9.09% behind Balance/Madness. Granted, Aerre is in purely 72s, while M-nightrider is in nearly full 75s, but that's still a very broad gap. Narrow that to 5% and it still seems a bit excessive. Infiltration is probably designed to be more of a burst spec. However, their burst is extremely unreliable, and so they generally end up filling a purely sustained DPS role. That, coupled with the fact that their AoE is much weaker than Balance generally means that I would consider the specs relatively balanced if they parsed almost identically. And yet, we're seeing a gear-adjusted 5-7% gap between them. That seems unacceptably wide.

Verdict: Balance needs about 3-4% more damage. Infiltration needs a LOT more burst control, or about 9% more sustained damage.

A Note on Gunslingers

Interesting sidebar: the gap between the best Sharpshooter/Marksman parse and the best Dirty Fighting/Lethality parse is only 0.54%. That's well within rounding error. Now, Sharpshooter has no setup time at all, and so it would probably be more balanced for it to parse slightly behind Lethality (perhaps by about 1%), but even still. Lethality is in a much better place than most people think with respect to boss damage. It is a totally viable spec.

Hybrid, on the other hand, is way out front, and I think that's a little fishy. I have no idea what Nithnuro's gear is like, but a 5.50% spread between Hybrid and Marksman seems a little much, even assuming he's just in 72s. Hybrid seems to be a bit over-powered at present, which is likely stemming from the fact that it's not a spec bioware intended in the first place.

Given the dev's comments on the sniper and scoundrel questions, I'm guessing that they never intended snipers to parse quite as high as they are parsing at present. It's entirely possible that snipers are supposed to parse much closer to where commandos are parsing, which is to say, about 5% behind where sentinels are.

Verdict: Nerf hybrid (or buff sharpshooter and dirty fighting).

Melee vs Ranged

There is a balance theory which asserts that melee DPS should be about 3% ahead of ranged specs (across the board) since they have to deal with a lot more mechanics than ranged DPS do and are generally harder to play. I think that's actually a pretty fair balance goal, given the way that current content significantly penalizes melee over ranged. The current metagame seems to be all over the map in this respect. Sentinels and Scoundrels parse extremely well, and generally about 3-5% better than gunslingers and commandos, if we take the balance goal for gunslingers to be what sharpshooter is parsing (i.e. exclude hybrid spec). This would also mean that Sages are parsing just a bit too high, and should probably come down a tiny bit.

However, Guardians, Vanguards and Shadows are below even the ranged, which means that on average, melee is actually parsing *worse* than the ranged specs are. That seems…very broken, and is probably most easily fixed by bringing up the worst of the melee specs, removing hybrid gunslinger and slightly nerfing TK sage (by "slightly" I mean by about 0.5%).

Verdict: Buff vanguards, shadows and guardians by ~5%. Remove hybrid gunslinger and *slightly* nerf TK Sage by ~0.5%.

General Conclusions

Overall, I'm fairly happy with how the DPS specs are balanced in the current game. There are definitely some outliers, which I tried to point out, but things are not quite as dire as many people assert. Sages are doing great, Commandos are doing great, the three Sentinel specs are absurdly well balanced, and Scoundrels are very competitive with Sentinels. Kudos to Bioware for achieving all of that.

However, some classes are woefully far behind. Vig Guardians are the most egregious example here: Vig needs a buff more than any other DPS spec in the game (post-2.4). Focus should probably be slightly buffed as well, but Vig is the real poster child. I'm guessing that Bioware is assuming Vig can use Dispatch nearly on cooldown below 30%, which just doesn't seem to be supported by the information I've been able to find. Without this natural Dispatch, Vig just doesn't have all that much to gain going from dummy to boss, which puts it substantially behind virtually every other spec in the game.

Shadows and Vanguards both need some slight buffs. With 2.4 coming, I'm very hesitant to draw any conclusions about Vanguards until I have updated parses. I suspect that they'll be in a much better position. Overall though, the melee classes should be parsing better than they are (aside from Scoundrels and Sentinels, which are in a very good place). Melee is just too heavily penalized by current content, and there's really no reason not to run nearly exclusively ranged DPS if you really want to min-max your group. Especially in light of the fact that there are more melee specs than ranged, this seems like a poor bit of game design.

Overall, a DPS spread of 10.16% (between Lethality Op and Vig Guardian) is too much, and this needs to be reigned in. Buffing Guardians will hopefully address this, and slightly buffing the non-sentinel/scoundrel melee classes will essentially shrink the margins to around 5%, which is bioware's stated goal for comparable specs. Things are fairly close to balanced (at least we no longer have classes that are 25-30% behind!), but they still need a little more tweaking. Hopefully we'll see some improvements in these areas coming in 2.5.
Computer Programmer. Theory Crafter. Dilettante on The Ebon Hawk.
Tam (shadow tank) Tov-ren (commando healer) Aveo (retired sentinel) Nimri (ruffian scoundrel)
Averith (marksman sniper) Alish (lightning sorcerer) Aresham (vengeance jugg) Effek (pyro pt)

Vaidinah's Avatar

09.25.2013 , 07:12 PM | #2
Excellent job KeyboardNinja. As you stated, the gear disparity will affect the numbers significantly, but the HM Dread Operations will come out soon to rectify this situation. People will get fully geared (which should be much easier than currently getting full 75s now) in a matter of 2-3 months and this will give us more accurate numbers on every DPS spec in the game.

As for some of the underrepresented specs, I'll talk to a guildie who plays an Assault Vanguard (it's his main spec and he says he has the highest Assault parse) about trying to get numbers for them once we get him and everyone else geared up in the new ops. Invinc would probably be your best bet for getting the top parse for current Concealment numbers as he will be fully geared ASAP.

Unfortunately, I'm not surprised at the low numbers Vengeance/Vigilance currently has and it's clear that many others were previously aware of them as well. There are probably less Juggernaut/Guardian DPS in top guilds than there are Assassin/Shadow ones. Yet the contrast in behavior on the forums couldn't be more clearly different. It's amazing how much difference it makes when people take a more complete look at class differences as opposed to just reading some numbers out of context.
The Harbinger

Guardian/Juggernaut Tank Guide: From Beginner to Master

Emperor-Norton's Avatar

09.25.2013 , 10:59 PM | #3
Interestingly you can still crop Combat Logs that have been deleted like Gorband's, which is now cropped to the 300 second mark. Just find the start time and then go five minutes from that.
plz Fix Watchman

THoK-Zeus's Avatar

09.26.2013 , 12:37 AM | #4
I think invinc's Operative parse is with 75 gear (correct me please if i am wrong here, but from the logs his OS does 10% more damage with the excact same crit rates then a 72 geared operative parse) and so are almost all of the world top parses (except for Vengeance/Rage Juggs/Guardians) are with atleast some 75 gear, so that will probably shrink the difference in overall damage even further.

cs_zoltan's Avatar

09.26.2013 , 07:02 AM | #5
Awesome work and good read!
I wish BW would listen to you, they actually plan to lower the top-end dps of Vigi in favour of less RNG
Melee sustained should be on the top of the leaderboard and ranged burst/aoe on the bottom.
The ability to speak does not make you intelligent. Now get out of here.

The Red Eclipse <PC Gamer Mint Imperials>
Kéja, Kejann, Aemis, Kyrí, Kyra'h, Késsa, Frängit, Lamîa, KalkïMaelin, Morwy

Darth_Dreselus's Avatar

09.26.2013 , 07:08 AM | #6
Pretty much what I had expected.


I'm surprised with this outcome, while I knew it has ways to go (hence the buffs) I have one in progression group and her DPS has never been an issue. I'll see if I can get logs. Then again you say the person had 69 armourings which could be the cause of such a low number.


The 2.4 changes are very strange, they seem to be neither buffs nor nerfs and mostly not needed.

-Moving of AP damage - mostly irrelevant, it's still a weak ability which majority of Assault players avoid.
-Ionised Ignition - this could be said to be a nerf as a 6s DoT is now guaranteed by a 9s cd ability rather than a no cd ability, but it is likely that it will still be up all the time due to the ranged damage and IP spamming.
-IR changes - both a buff and a nerf, lower damage but lower cost which could lead to an IP instead of HS, increasing resulting damage.

Neither of these changes are very clear cut and neither is an outright "yes we needed this" buff.

Shadow/DPS specs

Talking to a former Shadow (now a Slinger) it was really the mind numbingly boring rotation of Balance and the unpredictability (and not much better rotation) of Infiltration that made him drop the class.
No One Important Died Tavernus - Sorc / Taverna - Mara / Pintus - Op / Gintonica - PT
The Red Eclipse. Sweet Dreams to all my Harbinger children.

JMagee's Avatar

09.26.2013 , 08:24 AM | #7
First off, fantastic job KBN. I think this is a very solid effort that can help start illustrating the fundamental disparities between DPS classes currently used in the game. Keep up the good work, and I would love to see this evolve further in the future.

Just a few small thoughts/concerns/questions I would like to add though.

1. I know this may be difficult to do looking only at the parse, if you cannot talk to the parsee directly, but is there anyway you could indicate that talent spec (x/y/z) that some of these parses used. And as an aside to that, am I correct in assuming that the named talented parse (i.e Infiltration/Vengeance etc.) have taken the capstone ability in the tree and are talented a full 36 points into the particular tree?

2. The Focus Guardian and Assault Vanguard parses appear to be switched in some capacity. Currently it shows the Assault parse is higher, but is ranked lower.

3. (This is an aside, and probably more of thought going forward with this) As an Assault VG player, I've personally shied away from full Assault, and Assault Plastique, in favor of a 8/10/28 spec. As has been well discussed on the forums, the talents between Burnout and AP are so...underwhelming, that the points invested in them almost hurt you more than help, considering the low tiered skills in Tactics and Shield Specialist (Frontline Offense, Overcharged Cell Capacitor) that provide tangible DPS boosts) that are available in their place.

While some may consider 8/10/28 as a different sort of hybrid, I think most people would think of that more as the go-to Assault spec than full x/x/36. It's my personal opinion that this spec would probably parse a bit higher in as capable hands (Drekkin) as his full Assault spec, and maybe improve the DPS listing for Assault.

However, I realize that this sort of defeats the purpose of the idea that capstone abilities should be good and this is some preliminary evidence that some are and some aren't, but it could suggest an alternative way to balance the VG class at the expense of a flat damage buff. Simply moving some talents around within the trees to give Assault specced VGs some additional useful talents to spec into on their way to AP.

While my last point doesn't really have a thesis or conclusion, and it's sort of rambling, and I'm not sure if any other class has such an average 6th/7th tier that they avoid talenting into it (without being considered a true hybrid) but it sort of opens a larger discussion of balancing via damage boosts or by adjusting talents. And I know in 2.4 they are adjusting the survivability of Assault, as well as some adjustment of DPS, so the idea of moving skills around probably won't happen, but it is a thought.

So I guess the big picture. open-ended question regarding point 3 is: If BW believed hybrids should not be as strong as full 36pt builds, should the proper course of action be buffing the skills leading to the 36pt builds or simply breaking the hybrids (binding abilities to stances/cylinders where applicable, shifting talents around within a tree to force people to invest more points in a tree to reap the benefits)

namesaretough's Avatar

09.26.2013 , 08:47 AM | #8
A very thorough writeup, although many of the results were what I expected a few of them surprised me (vig guardian in particular).

I looked a little to try and find the gear of the top parses (full meaning BiS, as far as I can tell, nearly full meaning a few pieces short but having MH/OH/at least one relic+ almost all BiS in the rest of the pieces):

-Gorband, full 75s.
-Evryday, nearly full 75s
-Invic, likely nearly full 75s but no AMR I could find
-Handcuff, full 75s
-Ranic, full 75s
-M-nightrider, the first one on this list without nearly full 75s, although he does have a MH/OH.
-Pizza'da'hutt, 3/4 or more 75s including the MH/OH
-I'm reasonably sure Nithnuro is almost full 75s, judging from the timing of the parse.
-Tianna, almost full 75s
-Mruniverse is in nearly full 72s (72s with a couple 69s).
-Marisi: Full 75s
-Pelara: About half 75s, but missing the MH.
-A'jantis, 72s, very few 75s
-Vepres, Mix of 72s and 69s
-Aerre, nearly full 72s/a few 69s
-Kishekzun, full 72s and a DG relic
-Drekkin, mostly 72s, some 69s/pvp armorings
-Ylira, 72s with 69 armorings

Notably, the top half of this board is done with pretty much the same gear level, while the bottom ~6 specs seem to be on average a tier or more behind. I don't know if this changes any of your opinions about class balance.

KeyboardNinja's Avatar

09.26.2013 , 10:11 AM | #9
Quote: Originally Posted by namesaretough View Post
Notably, the top half of this board is done with pretty much the same gear level, while the bottom ~6 specs seem to be on average a tier or more behind. I don't know if this changes any of your opinions about class balance.
I can confirm that Mruniverse was in not quite full 72s when he did that parse (he had some Verpine stuff and I think at least one Arkanian relic). A'jantis is in full 72s with a KD relic and a couple 75 enhancements. Aerre is in full 72s.

So, the top-end hybrid slinger parses jump by 4.7% going from 72s to 75s. This is judging by the difference between your parse (nearly full 75s) and Red'october's parse (full 72s with no 75s). That's right in line with the 5% I was assuming. If everyone where in the same gear, it would shrink the 10% disparity that I quoted at the beginning of the post, but the major issues I outlined are still true even with evenly-matched gear (since I did a 5% hand-wavy adjustment for gear in each case).
Computer Programmer. Theory Crafter. Dilettante on The Ebon Hawk.
Tam (shadow tank) Tov-ren (commando healer) Aveo (retired sentinel) Nimri (ruffian scoundrel)
Averith (marksman sniper) Alish (lightning sorcerer) Aresham (vengeance jugg) Effek (pyro pt)

Darth_Dreselus's Avatar

09.26.2013 , 10:19 AM | #10
I do fully the support the viewpoint that 8/10/28 (and variations of such as 6/8/32 which allow for more burst phases) are generally viewed as VG Assault rather than outright hybrids (Assault with Gut or 22/22). It has been like this even before 2.0 and it is sad that the capstone is still so underwhelming and has little to no synergy with the rest of the tree (the only synergy is the DoT which seems to have been added as an afterthought and does little to improve the ability).

For Mandos I believe that AP does more damage even though I am unsure as to why (most likely the fact that Hyper Barrels is generally a better talent than Riot Augs in terms of which abilities are affected) and indeed Pizza's parse does use Thermal Detonator. The build with Grav Round is probably of more use in PvP but could be better than 36 Assault in PvE as well, the only problem being that GR does less damage than CR in Assault so it's only used to guarantee HiBs and lower armour which is not required if there is a Guardian or Sniper in the raid.
No One Important Died Tavernus - Sorc / Taverna - Mara / Pintus - Op / Gintonica - PT
The Red Eclipse. Sweet Dreams to all my Harbinger children.