Page 1 of 2
1
2
LastLast
  1. #1
    Warchief Akraen's Avatar
    10+ Year Old Account
    Join Date
    Dec 2011
    Location
    Tjøtta, Norway
    Posts
    2,149

    PSA: Simulationcraft fixed for MoP Mages

    For the past expansion or two, SimulationCraft has had a lot of trouble giving accurate numbers for mages. After a quick round of fixes to frost at the start of MoP, the mage module was left without a dedicated maintainer and left in shambles. Many top mages have been quoted saying that SimC is useless, and as a result it was largely abandoned for most purposes.

    We're happy to announce that the SimC mage module now has an active maintenance community hosted at Altered Time (http://www.altered-time.com). Many of the problems have been fixed, and simulations accurately represent what mages of each spec do with their rotation. In fact, if you hop over to the latest sample charts (http://www.simulationcraft.org/reports/Raid_T16H.html), you'll notice that all 3 mages are within the top 6 of the 27 DPS specs. In particular, Fire spec used to be in the bottom 5, and is now ranked 3rd or 4th among all specs.

    Some of the notable fixes are as below:
    - Changed all specs to react to PBoI procs properly, especially with Alter Time
    - Fixed a bug where Frost used to lose their icicles after using Alter Time
    - Fixed Arcane where it used to have ~88% LB uptime, now ~99%
    - Fixed Arcane where mastery used to affect cloak damage
    - Fixed a lot of issues with Fire's rotation: ignite building with Pyroblasts before Combustion, POM-Alter Time Combustion combos, Pyro camping instead of firing instant pyros ASAP, 2T16 camping, reacting to trinket procs, avoiding munching Pyro procs
    - Too many other small optimizations to list out

    Just to whet your appetite, here are some heroic trinket comparisons generated with the new SimC mage module:
    Arcane: http://altered-time.com/forum/viewtopic.php?f=3&t=202
    Fire: http://altered-time.com/forum/viewtopic.php?f=4&t=196
    Frost: http://altered-time.com/forum/viewtopic.php?f=5&t=197

    All of these are now live on the release version of SimulationCraft, and we'd like to invite you to try it out youself. Head over to http://www.simulationcraft.org/download.html to try it out yourself. If you see any issues with it, or if you wish to make suggestions for improvement, don't hesitate to register at Altered Time (http://www.altered-time.com) and post about it in the forums!

    For more info on what we did, check out the news post on AT: http://altered-time.com/forum/viewtopic.php?f=23&t=205

    (signed)
    Akraen of the House Natural Order, the First of His Name, King of Argent Dawn, King of Ysera, Stormrage, Steamwheedle Cartel, and Frost Mages. Co-Regent of Altered Time, Protector of the Theorycraft Realm, Archmage of Kalimdor, Breaker of Fears and Roots, Father of Water Elementals

    P.S. All 3 Mage specs are above all 3 Warlock specs. As it should be!

  2. #2
    The Lightbringer LocNess's Avatar
    10+ Year Old Account
    Join Date
    Oct 2011
    Location
    Indiana, United States
    Posts
    3,623
    Woo awesome! Would have been nice like a year ago, but at least there are eyes attending to it for sure.

  3. #3
    Deleted
    Quote Originally Posted by LocNess View Post
    Woo awesome! Would have been nice like a year ago, but at least there are eyes attending to it for sure.
    Yup. At least there is a group of people working on it now, and it'll definitely be ready for WoD, most likely in time for us to use it to find out what to do rather than build SimC based on what we do!

  4. #4
    Deleted
    So now we can use SimC refferences to prove our points?
    Woeeeheoeoeoeoe

  5. #5
    High Overlord Huevos's Avatar
    10+ Year Old Account
    Join Date
    Aug 2012
    Location
    Rochester, NY
    Posts
    175
    That's awesome news guys, thanks so much for your hard work on this. SimC is a tool I've missed this expansion, pretty excited to get it back!

  6. #6
    Wait... really? I thought it was never going to get fixed. Great news for the mage community, +1 internets to the SimC mage team.

  7. #7
    Herald of the Titans Shangalar's Avatar
    10+ Year Old Account
    Join Date
    Aug 2011
    Location
    Rijeka, Croatia
    Posts
    2,641
    Very well done to all who participated! I tried to follow the discussions for a very limited time and it didn't work out for us
    My magic will tear you apart.

  8. #8
    Bloodsail Admiral WarpKnight's Avatar
    10+ Year Old Account
    Join Date
    Aug 2011
    Location
    Straya
    Posts
    1,193
    Awesomeballs! Good work guys!

    Can't wait to get home and test it out.

  9. #9
    Deleted
    but it still shows only what is in theory better - playtesting is what realy matters, but it can push you to the right way. pls just see difference between theory and real situation, because you can learn much more from playing your spec

    i'd be glad to have this for WoD to push us to the "right" direction at least for first few weeks - after that it doesn't matter

  10. #10
    Deleted
    Quote Originally Posted by Quetesh View Post
    but it still shows only what is in theory better - playtesting is what realy matters, but it can push you to the right way. pls just see difference between theory and real situation, because you can learn much more from playing your spec

    i'd be glad to have this for WoD to push us to the "right" direction at least for first few weeks - after that it doesn't matter
    For sure! Simcraft is a tool to test theory and get technical ideas out of. The user has to understand this to make the fullest use out of it.

  11. #11
    Good job! What is the reason for you to change pyroblast.react heating_up.react to pyroblast.up heating_up.up? Isn't react checking if player can react in time?

  12. #12
    Fluffy Kitten Wilderness's Avatar
    10+ Year Old Account
    Join Date
    Dec 2010
    Location
    Maryland
    Posts
    6,799
    Quote Originally Posted by tolgag View Post
    Good job! What is the reason for you to change pyroblast.react heating_up.react to pyroblast.up heating_up.up? Isn't react checking if player can react in time?
    I could be mistaken, but I think that's the change that makes it so the Sim holds a Pyro! until you also have a heating up proc as well.

  13. #13
    Deleted
    Quote Originally Posted by Vegas82 View Post
    I'll give it a look. Hopefully it can actually calculate stat weights properly with these changes, getting really tired of explaining to people how much SimC, and by extension Noxx rankings and AMR, suck for mages.
    I just tried it out, and as it seems it actually calculates the difference between Mastery and Haste very good atleast. Have to try out more then 10k iterations to see if it actually Simulates it properly.

  14. #14
    Deleted
    Quote Originally Posted by Liveet View Post
    I just tried it out, and as it seems it actually calculates the difference between Mastery and Haste very good atleast. Have to try out more then 10k iterations to see if it actually Simulates it properly.
    Well we have been running extensive simulations of Haste vs Mastery for Fire, and we are seeing a trend very much like we see people reporting on live. I think they linked the specific threads in the OP, but we are trying to explore the secondary stats now.

  15. #15
    Deleted
    Quote Originally Posted by Liveet View Post
    I just tried it out, and as it seems it actually calculates the difference between Mastery and Haste very good atleast. Have to try out more then 10k iterations to see if it actually Simulates it properly.
    http://altered-time.com/forum/viewtopic.php?f=4&t=198

    These are some of the early results we've found. TL;DR In BiS best haste/mastery is somewhere around 9.6k-12k haste and 10k-12.5k mastery.

  16. #16
    Deleted
    Quote Originally Posted by Kver View Post
    Well we have been running extensive simulations of Haste vs Mastery for Fire, and we are seeing a trend very much like we see people reporting on live. I think they linked the specific threads in the OP, but we are trying to explore the secondary stats now.
    Lovely! Ahyea, I tried out a few different settings, and the only actual point where Haste proved better then Mastery was at a Garrosh type of fight. As in more then 9 minutes. Other then that Mastery was always just a little step better. ( Simulations was done around 9.5k haste for the record)

    - - - Updated - - -

    Quote Originally Posted by Dutchmagoz View Post
    http://altered-time.com/forum/viewtopic.php?f=4&t=198

    These are some of the early results we've found. TL;DR In BiS best haste/mastery is somewhere around 9.6k-12k haste and 10k-12.5k mastery.
    Thanks for linking it. Grateful that you guys put in this kind of work! Fun to see that live testings actually worked out well in comparison to theory. I've seen and heard a few people which tend to go even lower then 9.6k haste. Wonder how that will turn out.
    Edit: Just read everything through now. Makes so much sense!
    Last edited by mmoc71ea6f2691; 2014-06-18 at 03:34 PM.

  17. #17
    The Lightbringer Artorius's Avatar
    10+ Year Old Account
    Join Date
    Dec 2012
    Location
    Natal, Brazil
    Posts
    3,781
    Awesome work guys!!!

  18. #18
    Deleted
    we don't need simcraft to calculate what is better for fire we are playing the spec for so long.... and we know that for shorter fight crit>mastery>haste+luck on PBoI and longer (+-4m +) fight crit>haste>mastery - THIS we learnd from PLAYING fire and it is much more important than just some simulation ---> simcraft is goot if we are creating theorycraft not when we are playng the spec for more than a year

  19. #19
    The Lightbringer LocNess's Avatar
    10+ Year Old Account
    Join Date
    Oct 2011
    Location
    Indiana, United States
    Posts
    3,623
    Quote Originally Posted by Quetesh View Post
    we don't need simcraft to calculate what is better for fire we are playing the spec for so long.... and we know that for shorter fight crit>mastery>haste+luck on PBoI and longer (+-4m +) fight crit>haste>mastery - THIS we learnd from PLAYING fire and it is much more important than just some simulation ---> simcraft is goot if we are creating theorycraft not when we are playng the spec for more than a year
    Still always nice to have simulated math back you up. And either way 99% of players don't know how to accurately napkin math and might not want to believe some person on a forums saying "hey it is better this way" (and honestly you should always test things yourself first).

  20. #20
    Deleted
    Quote Originally Posted by Quetesh View Post
    we don't need simcraft to calculate what is better for fire we are playing the spec for so long.... and we know that for shorter fight crit>mastery>haste+luck on PBoI and longer (+-4m +) fight crit>haste>mastery - THIS we learnd from PLAYING fire and it is much more important than just some simulation ---> simcraft is goot if we are creating theorycraft not when we are playng the spec for more than a year
    If that is how you feel, then fine. But you are looking at this in such a narrow way. There are multiple scenarios where Simulations are going to let you see exactly what is better and what is not. The time it takes to actually check that out yourself ingame is so much longer then what it takes to run a simulate. Just for an example. Sha Of Pride, how are you supposed to know what build is best there? Well, either you could go and simulate a fight that is somewhat alike Sha Of Pride, or you could get a group, clear up to Sha Of Pride, then wipe there a whole day just because you want to try out what build is best. What sounds better to you?

    You have to think of this as a long term goal too. Finally we have a group of people that actually puts in dedicated time into doing this, for players like YOU and ME, players that aren't capable of coding, getting SimC right or just don't have the time to do it. They are doing all this just for us. So why would you want to take that effort away from them? By stating, "we already know this". Well, you surely will not know it the day that WoD hits, and you surely wouldn't have known it if not the players stated above had done extensive in game testing already. I really doubt that you've put in more then a few hours to check what build is correct.

    So please, spare us the complains about all this and go on with your ingame testing.

    Really good job with this all of you! Been running multiple Sims now again with the current gear since it's all upgraded now, and the output are truly helping.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •