Sunday, March 15, 2009

RIP Rogues in 3.1?

GC words that Rogues are in a good place, HFB and other rogue mechanics are rather broken, but it's too risky to try and fix us has hit the community rather hard. Here is a consolidated lists of issues still facing the various specs and how to fix them. While these fixes are more focused around improving things for PvE, there is an eye on not making the changes over powered for PvP.

1) Assassination

Instant Poison: Baseline instant is still weaker than baseline wounding, even with a 50% envenom buff up time. Baseline IP needs a boost in base damage, AP scaling is on par with wound. Until the Improved Poison bug is fixed it'll be impossible to fully test Assassination's DPS, and with the patch being scheduel for early April it is starting to feel like the WotLK Beta all over again with rogue unable to fully test unbugged specs before release.

HfB: Pages and Pages and pages have been written about how utterly unfun, uninspider, and needlessly powerful this talent has become. The problems with this talent have been hashed out over and over again, the real issue with the new implimentation is that it is only useful in 25 man raids, and the usefulness deminishes as the group size gets smaller and smaller. It also worses the specs ramp up to peak DPS.

Rotation Ramp Up: Longest longest lead time in game to reach peak DPS, HfB change has made this worse. Spec becomes less and less powerful the more target switching is in place and the single target DPS gain over combat will cause this spec to become a relic.

My Fixes:

  • Deadly Brew - Give back IP procing Deadly. Yes this is a massive DPS gain, probably too good for the tier level and talent points as each point is worth about 5% DPS, but it has no effect on PvP play and it's better than 15% DPS for a single talent. Swap it with MP if needed to put it out of reach of non Assassination builds (however this messes with 41/5/25 builds - having to give up DD or HS).
  • Master Poisoner - 33/66/100% chance not to clear Deadly Poison with Envenom. The increased chance is basically the same thing, the only difference is that having envenom clear the stacks forces the rogue to watch a DP tick timer in order to release Envenom as just the right instant. Timer watching isn't fun or needed in this case.
  • Overkill - Rework Talent. This talent is a PvP bane and causes a rogues already high out of stealth burst to be all that much worse. Here is what I would do: Each combo move critical strike buffs the rogue with the ability to Overkill, stacking up to 10 times. When the Overkill ability is used all energy moves cost 10 less energy for 1 second per stack consumed. 60 second Cooldown. This talent provides a much needed on demand cooldown for the assassination rogue. The stacking mechanism prevents it's use during the out of stealth burst.
  • CttC - Evis/Envenom Refresh SnD to 5 CP max or Initiates SnD at the number of CP uses. This change is needed to reduce the lead in time to reach max DPS.
  • HfB - Damage is increased by 3-5% (Glyped to +1%) against targets that are bleeding. Other DPS gains in the tree means that this talent can be greatly toned down and no longer needs to be a button that has to be pushed. The talent could also be: Poison damage is increased by 10-15% (Glyphed +3%) against targets that are bleeding to make it 100% totally unattractive to PvP rogues.
2) Combat

In general the combat tree is in the best spot of all the trees. The talent tree does flow, through there is only a single spare talent point to put into flavor talents, and duel weapon specing is no longer possible. Cutting all the weapon specs down to 3 talent points for the same effect and allow for more specing into flavor talents (for people who single weapon spec) would be a nice change, but not required.

Killing Spree - The loss of character control with this ability isn't acceptable. If there is only a single target in range of the ability and the player is already in melee there should be no ninja jumping. This would make the talent safe to use on most bosses.

Mace Spec - If haste is indeed added to this talent it will probably become too good, forcing rogues to use Maces (YUCK!).

Daggers - Combat daggers really isn't a viable spec. This is made worse with the PPM poison change as SS with a 2.5-2.7 speed sword will almost always proc wound, while a 1.7-1.8 speed dagger will not (~30% less chance to proc the poison). To make matters worse back stab is performed less than SS. When you add in poison damage with the PPM system SS has a higher DPE than back stab, and this is even before you take into consideration the number of combo points generated. BS either needs to scale at a strong rate, cost less energy, or award 2 combo points to make up the difference.

No real Fixes are needed here, only some minor Tweaks.

3) Subtlety

This tree is a mess and it is hard to fix without knowing if the Honor Among Thieves talent will stay in it's current form. Personally I'd like to see the talent changed to "Whenever anyone in your party or raid group critically hits with a damage or healing spell or ability your next ability costs 10-20 less energy." This would be much easier to balance around (especially for raiding) as you could assume that the buff is always active in a raid enviorment without the need to stack a perfect group.

Saturday, December 13, 2008

The Problems With Poisons

Problem: Instant Poison is Less Effective than Wounding Poison at all Attack Power Levels and requires an investment of 4 talent points just to being it's effectiveness inline.

How many of you rogues out there know that Wounding Poison (Our utility anti healing poison) is a better than our baseline Instant Poison? In fact it takes 4/5 talent point into Improved Poisons just to bring the two up to parity in a raid situation. This means that the assassination rogue is spending 4 worthless talent points for that fifth point that finally makes Instant Poison a better source of damage than Wounding Poison. Below is the chart that proves this fact as Instant Poison damage for the various application percentages is subtracted by wound poison damage (The Zero Line).



As you can clearly see only 4/5 and 5/5 in Improved Poisons makes Instant Poison more viable than Wounding Poison, and 4/5 takes around 4000 Attack Power before it crosses the zero line (Wound Poison Damage). This happened from the massive poison buffs and subsequent nerfs to assassination DPS. When Deadly Brew allowed rogues to double up on instant poisons with a 40% proc rate and still get Deadly Poison ticking the poison damage was making up nearly half of a rogues damage. Deadly Brew and the application percentage was nerfed and this brought rogue poison damage back to where it belonged. However, in a later beta patch wound poison's proc rate was upped to 50% causing the problem you see with instant and improved poisons above.

Solution: Simple, make Instant Poison a 30% baseline application rate, with Improved Poisons Bringing it up to 40%. This has no effect on PvP while PvE rogues get a slightly more effective DPS poison at the cost of utility and Assassination Rogues get about a 5-7% (33% More IP Damage) overall DPS buff (and no longer have to waste 4 talent points just to make instant as effective as Wounding). If the 5-7% DPS gain is too high for 5 talent points, the talent could be reduced to only 3 talent points increasing the chance to 36% fully talented. This would result in a 3-4% (20% More IP Damage) DPS gain for the 3 talent points invested.




Problem: The fixed damage and application rate of poisons greatly favors fast weapons over slow weapons.

At first glance this doesn't appear to be that great of a problem and is largely a "duh" moment. Of course if the damage is fixed and the proc rate is fixed the faster you swing the more damage is going to be done. The problem comes from the nature of the scaling as the chart below shows that damage scales much more quickly the more and more you increase weapon speed. The chart below shows the DPS provided by baseline poisons at the various weapons speeds. The Effective weapon speed includes having all raid haste buffs and 5% Haste on gear. I've included Wounding, Instant (@30% and @40% per my Recommendation Above), and Deadly Poison Ticking at a 5 stack for a point of reference.



You can clearly see the parabolic nature of the scaling. To further prove my point check out the DPS gained by added 15% haste to your gear.



This makes the rogue highly dependent on finding a fast weapon to place their Instant/Wounding Poison on and will cause gear scaling issues as World of Warcraft continues into the future. Weapon speed becomes the single most important stat to the point of a 1.3s blue becoming a better weapon than 1.5s epics that offer 10-30 DPS more in addition to better stats.

Solution: Change Poisons Damage to scale off Weapon Damage and not attack power. I've balanced the percentages to provide DPS between a 1.3 and 1.4s weapon under the current system.

Instant Poison: (Weapon DPS + AP/14) * 120% * Weapon Speed (Tool Tip: Each strike has a 30% chance of poisoning the enemy which instantly inflicts 120% Weapon Damage as Nature Damage)

Wound Poison: (Weapon DPS + AP/14) * 65% * Weapon Speed (Tool Tip: Each strike has a 50% chance of poisoning the enemy, causing 65% Weapon Damage as Nature damage and reducing all healing effects used on them by 50% for 15 sec.)

Here are some graphs that show the viability of this change. The first Chart shows current DPS scaling with weapon speed vs. the change to poison damage based upon weapon damage. The other charts show Haste and AP scaling of the change to poisons based upon weapon damage as a proof of concept that such a change would eliminate the current parabolic scaling from faster weapons. You'll notice that weapon damage poisons are scaling a little better with AP than current Poisons as I used 1.4s weapons and thus damage is scaling slightly better with AP. Even slower weapons would result in even more AP scaling, while faster weapons would result in less scaling, however the DPS provided by the poisons remains a constant.







The problem with this solution is that is devalues Deadly Poison a bit as now all poisons deal equal DPS over all weapon speeds, and would likely require a slight damage buff to deadly poison to keep it viable in a raid situation. Increasing Deadly Poison Damage by 10% would be enough and give all rogues a slight (~1% DPS boost in PvE).

I would also like to see Deadly Brew allow an assassination rogue to stack Deadly Poison up to 6/7 Stacks (for the 2 talent points). This would be a 40% increase in Deadly Poison Damage (about a 4% DPS boost overall), but would help out the rotation and quality of life as Assassination quiet a bit. There are more pressing reasons for this change, but they are outside the scope of a discussion on poisons alone. However, without this change Assassination rogues will continue to favor double instant poisons.


Solution Summry:
  • Increase Baseline Instant Poison Application Rate to 30%
  • Change Instant Poison Damage to 120% of Weapon Damage
  • Change Wounding Poison Damage to 65% of Weapon Damage
  • Increase Deadly Poison Damage by 5-10%
  • Add a Secondary Effect to Deadly Brew that allows Deadly Poison to Stack up to 6/7.

Friday, December 12, 2008

Rogue PvE DPS

It shouldn't come as any surprise to Rogues that our PvE DPS is currently not competitive with other pure DPS classes, and even many hybrid classes. In the best cases rogues are middle of the pack DPS and in worst cases are bottom of the pack, but this blog isn't about QQing over the state of rogues; it's about identifying the problems and looking at possible solutions.

This post is going to outline the primary causes of Rogue under performance and I'll go into greater detail with math and charts proving the points. For those of you that might question my roguehood you can check me out on the armory. I've cleared all TBC content and all of the current Wrath content. My DPS has always been within a percent or two of what the Rogue Spread Sheet reported DPS.

The Rogue Restrictions - The core rogue mechanics (combo points, energy, SnD, etc) are extremly restrictive and only allow the rogue to perform at peak efficency on a very small number of fights. All classes suffer from class mechanice restrictions, but no other class suffers as badly or as frequently as rogues do.

The Energy Factor - Energy was probably the single most overpowered class mechanic when the original WoW was released. The power of energy changed to balanced in TBC, but it has now become underpowered in Wrath. Energy, and energy gained per second, has remained a realative constant since the original release of WoW. The problem lies in the fact that as talent trees expanded for the other classes they have gained new ways to turn their resource into damage and that has caused energy to be devauled over time.

Poisons - Poisons are what the developers used to balance out rogue DPS, but the changes to poisons have ended up causing more harm than good. Poisons are going to have to be reworked if the rogue class has any hope of ever being balanced out.

Gear Scaling and the PPM Changes - Everyone is used to rogues being the "kings" of gear scaling. That is old news however because as of right now Rogues, and Enh Shaman, face some of the worst gear scaling issues in the game.