Advertisement
Guest User

DFA Testing

a guest
Sep 6th, 2014
1,313
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 5.18 KB | None | 0 0
  1. Context
  2. -------
  3. I do a lot of work on the Shadowcraft engine, and I haven't raided or been in game in over a year and a half. Previously I was a top tier raider (made some top 10 kills but the only placements of worth were in the top 20 range) who, perhaps too often, would push DPS at the expense of other things. My skills have attrophied quite a bit, but after 1-2 hours of practice most of my skills are back to decent, but still not where they were. Although, keeping track of the BG counter in my head and exploiting Anticipation stacks are a weakness of mine after getting back into the flow.
  4.  
  5.  
  6. The Data
  7. --------
  8. Previous day, approximately 12m damage each
  9. ~13.7k DFA ASAP w/ 5CP
  10. ~14.3k no DFA
  11.  
  12. Logs:
  13. Test 1 - http://www.warcraftlogs.com/reports/cznaQ8pwA4hjbq7B#type=damage-done&boss=0&start=0&end=974819
  14. DFA - Delay DFA for AR&KS
  15. 14,338.6 over 12.24m
  16.  
  17. Test 2 - http://www.warcraftlogs.com/reports/cznaQ8pwA4hjbq7B#type=damage-done&boss=-3&start=1067499&end=2018650
  18. No DFA
  19. 14,151.3 over 12.91m
  20.  
  21. Test 3 - http://www.warcraftlogs.com/reports/cznaQ8pwA4hjbq7B#type=damage-done&boss=-3&start=2839519&end=3748803
  22. Cast DFA as soon as possible at 5CP
  23. 13,983.1 over 12.19m
  24.  
  25. Test 4 - http://www.warcraftlogs.com/reports/cznaQ8pwA4hjbq7B#type=damage-done&boss=-3&start=4625127&end=5245624
  26. No DFA - Talent selected
  27. 13,795.2 over 8.18m
  28.  
  29. Test 5 - Restarted game to load up a CP mod and forgot to re-enable logging, oops
  30. No DFA
  31. ~14401.8 over 12.76m
  32.  
  33. Test 6 - http://www.warcraftlogs.com/reports/cznaQ8pwA4hjbq7B#type=damage-done&boss=-3&start=8495972&end=9559743
  34. DFA - Delay DFA for AR&KS
  35. 13,943.1 over 14.26m
  36.  
  37.  
  38. Performance, excluding previous day testing:
  39. No DFA -> 14158.99613 = 5397.14277696 + 3333.67019202 + 5428.183161 over 33.85
  40. DFA w/ delay -> 14125.7762264 = 6622.80996226 + 7502.96626415 over 26.5m
  41. DFA ASAP - > 13.983.1 over 12.19m
  42.  
  43. Stats:
  44. i660 lvl 100 premade
  45. 3810 agi
  46. 1123 crit
  47. 662 haste
  48. 938 mastery
  49. 359 multistrike
  50. 561-1043 2.6s mace Mark of Warsong(both hands)
  51. i660 Glad Insignia of Conquest (734 agi for 20s proc)
  52.  
  53.  
  54. Thoughts
  55. --------
  56. With a few hours of practice and testing. I'm not sure if DFA is even a DPS increase. This data still isn't all encompassing, and you can easily see enough variation in damage (and perhaps performance) to realize this isn't conclusive. But it's over a large enough spread to ackowledge that even if DFA was undervalued, it just does not discernably help single target DPS. It is, at best, a glorified Shadow Step. I highly recommend revisiting the talent, particularly for combat rogues.
  57.  
  58. If I may be so bold as to make a suggestion, consider making the finisher deal Shadow damage instead of whatever it would otherwise deal. This can actually fix a number of problems at once. This allows you to buff Combat's usage without buffing Assassination's, and only slightly impacts Sub's usage. This gives you more leverage to fine tune the bonus 50% finisher damage so you don't have to make it +80% or something more drastic (Sub burst needs to be considered of course). This also gives Combat rogues an additional avenue of mitigating armor in PvP, which has been cut down significantly in Warlords.
  59.  
  60. There are several other thoughts and concerns I've encountered during these tests. Combo points are attrocious. It's not a disappearing thing, it's not a "CPs are still on the rogue" whine. The problem is that CPs mislead the player during finishers. The default UI current misrepresents how many CPs you have at a given time. It will flash 3 CPs in an unorganized way (like 1, 3 and 5 will light up, or 1 and 4) for 1/4 of a second after a finisher, then remove them to display them correctly. This, combined with the unintuitive Anticipation interface makes the CP UI misleading, and bad. I doubt this problem arrises for non-anticipation specs, but it could be a problem with some of the tier bonuses.
  61.  
  62. The second thing I want to point out is that the swing reset during DFA is actually somewhat substantial. This isn't a Vanish swing reset, this is somewhere between 1 and 1.5s of downtime reset. Combat currently relies significantly on it's white attacks to generate damage. This isn't a passive vs active thing, this is an issue where both the damage and the energy and mechanics from those autoattacks (like combat potency) contribute a -massive- chunk of Combat's DPS. It's like halting 90% of your DPS for 1s for a 50% stronger finisher, when you're doing 14k DPS and you lose 90% of it for 1s, you lose 12.6k damage. The finisher therefore needs to hit for more than 25k, at this 14k level, in order to break even in this trade. Do not underestimate just how passive rogues are. This downtime is a big deal. The fact that DFA during AR is distinctly worse than delaying it should be ample evidence. DFA on CD did not get the same treatment as the other two approaches, but after looking at the charts and other testing, I don't think it will get any better than where it is.
  63.  
  64. I'm told from others that DFA was potentially still bugged, and that DFA would occasionally fail to deal damage, while still consuming energy/CPs/whatever. I did not catch it while testing if it was still happening, but if evidence suggests it did then let me know what it gets fixed to re-examine the logs.
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement