Page 50 of 63 FirstFirst ...
40
48
49
50
51
52
60
... LastLast
  1. #981
    Deleted
    Hi, long time reader, first time poster here..

    I would like some advice on my trinkets. Currently I have 522 Bad Juju and 535 Renataki. This week I got 528 Talisman and am wondering if I should replace any of those with this one. I'm running SV crit>haste, have 4set bonus and metagem.

  2. #982
    I've currently got the 522 Shado-Pan trink, LFR JuJu, and LFR Talisman of Bloodlust.

    Am I correct in assuming the first two are what I should be using and that Bloodlust is junk?

  3. #983
    Quote Originally Posted by Hybro View Post
    Hi, long time reader, first time poster here..

    I would like some advice on my trinkets. Currently I have 522 Bad Juju and 535 Renataki. This week I got 528 Talisman and am wondering if I should replace any of those with this one. I'm running SV crit>haste, have 4set bonus and metagem.
    I'd go bad juju and renataki's

    Quote Originally Posted by mistahwilshire View Post
    I've currently got the 522 Shado-Pan trink, LFR JuJu, and LFR Talisman of Bloodlust.

    Am I correct in assuming the first two are what I should be using and that Bloodlust is junk?
    VP trinket + LFR bad juju

    Signature by Geekissexy Check out her Deviantart

  4. #984
    So, I ran with the Rune of Re-origination throughout our normal-mode cleanup raid tonight. I will first detail the summary of the results before diving into the details.

    Summary
    The Rune of Re-origination is not very good for Survival Hunters.

    Configuration
    I ran in a configuration that had 8800 crit, 6800 haste, and 2300 mastery, which simmed well in FD. On proc, this resulted in crit rating of 27000 (over triple my previous crit rating). The test was done in a 25 man raid with full buffs and boss debuffs in a 3.5 hour raid.

    ilvl was 526
    Spec was kept as SV on all fights to have homogenous results in that regard.

    Log: World of Logs

    Uptimes:
    Reorig: (min: 11.2%, max: 31.4%, avg: 18.1%, >30%: 1 bosses, <15%: 2 bosses)
    Juju: (min: 14.2%, max: 35.4%, avg: 23.7%, >30%: 2 bosses, <15%: 1 bosses)

    Reorig specs:
    1467 agi baseline
    (edited for new proc mechanics) Proc: -6800 haste, -2300 mastery, +18200 crit for 10s.

    Bad Juju specs:
    1552 mastery baseline (reforged 620 to crit or haste, which allows for the player to create synergy of 9:6 mastery:crit)
    Proc: 7333 agi proc for 20s, resulting in 1686.59 agi normalized (23.7% uptime)

    Analysis (edited)

    From the parse, the Rune does indeed proc more than it used to, but slightly less than Juju and seems to have issues with uptime due to its short duration. Most fights had under 18% proc uptime for Rune. One of the more peculiar aspects of the parse was the issue of dps trend during the proc, which was always non-increasing (dps either stayed the same or went down). You can verify this trend for yourself by looking at boss fights within the parse, selecting Effinhunter, clicking Buffs Cast, and then clicking the "#" beside Re-origination. While buffs like Bloodlust or Bad Juju's procs are always associated with dps increases, the Re-origination was not strongly tied to dps gains. In fact, gaining such a proc was always associated with stagnant dps trends or dps losses in my particular case.

    My assumption on this is that because the proc was augmenting crit (which is highly RNG), and because the proc time is so low, the chance for bad luck streaks during this short proc of increased crit is high. Additionally, with the loss of haste during the proc, my metagem, 2 piece, and Bad Juju were less likely to proc (the proc rate of Bad Juju during this raid was significantly affected throughout the night, and I seemed to have fewer >30% uptime bosses and more parses within the 18-28% range for Juju).
    Last edited by Effinhunter; 2013-04-26 at 08:56 PM. Reason: Incorporating comments into analysis

  5. #985
    Quote Originally Posted by Effinhunter View Post
    So, I ran with the Rune of Re-origination throughout our normal-mode cleanup raid tonight. I will first detail the summary of the results before diving into the details.

    Summary
    The Rune of Re-origination is terrible for hunters.

    Gear configuration
    I ran in a configuration that had 8800 crit, 6800 haste, and 2300 mastery, which simmed well in FD. On proc, this resulted in crit rating of 17900 (over double my previous crit rating). The test was done in a 25 man raid with full buffs and boss debuffs in a 3.5 hour raid.

    ilvl was 526

    Log: World of Logs

    Uptimes:
    Reorig: (min: 11.2%, max: 31.4%, avg: 18.1%, >30%: 1 bosses, <15%: 2 bosses)
    Juju: (min: 14.2%, max: 35.4%, avg: 23.7%, >30%: 2 bosses, <15%: 1 bosses)

    Reorig specs:
    1467 agi baseline
    Proc: -6800 haste, -2300 mastery, +9100 crit for 10s.

    Bad Juju specs:
    1552 mastery baseline (reforged 620 to crit or haste, which allows for the player to create synergy of 9:6 mastery:crit)
    Proc: 7333 agi proc for 20s, resulting in 1686.59 agi normalized (23.7% uptime)

    Analysis

    Because of synergies between crit and haste and crit and mastery, we have known that stacking pure crit is worse than dividing secondary stats between crit and haste or between crit and mastery (i.e., you get more bang from your ilvl buck by not stacking any one stat--they harmonize together in pairs). From that statement alone, you can probably gather what the experiments showed on the World of Logs parses. What shows up in the log is that every single time the Rune proc'd on a fight like Iron Qon Normal (where there is only one target being attacked), the dps line either stayed flat or went down. The proc from the Rune was a dps loss for me on almost every single cast. The exceptions to this loss were when I was under the effects of some other beneficial cooldown.

    The only beneficial part of the Rune is the passive agility. The proc just shifts secondary stats. On normal, this means the pro of the Rune is 1467 agi.

    In contrast, every other trinket dropped from ToT that we would want has either the same passive agility (e.g. Talisman) or more normalized agility (e.g., Bad Juju and especially Renataki's). Despite what FD and Simcraft seem to be telling me about the ranks of these trinkets (right now, Rune > Talisman in both simulators), I cannot think of a good mathematical argument for why the Rune would ever beat an equivalent ilvl Talisman. The Talisman adds stats with its proc (stacking haste buff, similar to Renataki's agi one). The Rune simply moves stat weights around and does so in what we've modeled as the worst possible way to allocate secondary stats (loaded onto a single secondary stat). Even if Blizzard does up the proc rate again (to let's say double the current proc rate to have almost double the average uptime of Juju), because the proc rate is at best an equivalence (but more likely a loss considering it stacks a single stat), the trinket will gain no additional utility for hunters. Meaning, unless Blizzard changes the proc rate to something useful for hunters (e.g., by adding the mastery and haste totals onto crit while keeping the current mastery and haste), the trinket is only useful for its passive agility.

    Or as Tehstool has so eloquently stated before, "The Rune is terrible."
    You tried hard but I started laughing at "Because of synergies between crit and haste and crit and mastery, we have known that stacking pure crit is worse than dividing secondary stats between crit and haste or between crit and mastery (i.e., you get more bang from your ilvl buck by not stacking any one stat--they harmonize together in pairs)". If you're going to state something like that then you are going to need some sort of proof that isn't based on how it "feels" to play

    Shifting all your stats into Crit is defeinitely a DPS increase. Need proof? Go into FD and sim yourself normally (without Rune of Reorigination) then go into custom stats and adjust your level of Crit, Haste and Mastery to so you have 0 Haste 0 Mastery (before buff) and whatever your crit would become with the Rune procced.

    Then run a sim as usual but with those custom stats and you clearly see that having 0 Haste, 0 Mastery and 16000 (or whatever) crit on gear is a DPS increase.

    This can't give you an exact value of the trinket proc but it clearly shows that the proc is a reasonable DPS increase whereas some people here seem to think it is actually a DPS loss on proc.
    Last edited by Glurp; 2013-04-26 at 11:16 AM.

  6. #986
    Quote Originally Posted by Effinhunter View Post
    - snip -
    You also seem to have forgotten that Rune gives 2x the stats it takes away.

    We're going to make some changes to this trinket based on testing and feedback. Our goal was to make it decent if you didn't game it at all but offer some opportunities for enhanced benefits if you did. We're going to try to meet that goal better with these changes:

    - 10 sec duration (down from 20 sec) but with double the proc rate.
    - Increase your highest secondary stat by 200% of the sum of your two lowest secondary stats. For example: you have 3500 mastery, 5000 crit, 7000 haste. Rune procs, and you get [-3500 mastery, -5000 crit, +17000 haste]. Same logic as before, just double the size of the buff to your highest stat.

  7. #987
    Quote Originally Posted by Effinhunter View Post
    derp
    Lotta stuff you said here is wrong. As the other posters said, the rune doubles your stats, so your analysis on how much crit it gave you was wrong. Also, getting an item with crit+haste is better than an equal ilvl item with only crit solely because of the way ilvl is calculated. The item with crit+haste might give 150 total stats (75 crit + 75 haste) while the item with only crit might give 100 total stats (100 crit). If crit is valued higher than haste, then 150 crit > 75 crit + 75 haste every time.

  8. #988
    Quote Originally Posted by Spiralphoenix View Post
    You also seem to have forgotten that Rune gives 2x the stats it takes away.
    Thank you very much for pointing to the Blue Post about the Rune's changes. This is one that I missed, obviously, and I apologize to the community. I thought they had only doubled the proc rate and not the actual proc effect, which is significantly more now. 2x the secondary stats will be better, regardless of whether or not it's stacked. This oversight obviously affected my analysis of the procs, but does not affect what I saw in the actual dps performance of the proc. I'll go over that a bit later.

    The analysis was done in good faith and was certainly not meant to be misleading. I was simply basing some of my adhoc analysis on why the logs were showing ZERO DPS GAIN during the procs of Rune in every single cast where I was not under the effects of some other beneficial effect or cooldown. I assumed some of you would understand what I was saying here, so I will detail how you can view these losses yourself. Every other type of cooldown/proc, you can see tangible dps differences during a proc. Bad Juju, for instance, every cast is coupled with dps increases (and this makes sense because agility increased the ranged attack power that every ability and shot uses).

    To see what the Rune did on each proc (in tangible dps changes), go to one of the fights (e.g., Iron Qon), and click on "Buffs Cast". Next to the buff "Re-origination", click the "#". Now, this will produce a graph that shows my dps line over that fight with the buff cast being highlighted. The thing to remember about crit is that it is an increased chance to crit for a higher damage amount. It is not guaranteed. Because they reduced the proc duration from 20s to 10s, you have a smaller window to see this increased chance come into effect. Normalized over an entire raid, you may see the increases in crit you were expecting, but unlike other procs where you can say "oh, Juju procc'd, I should use my more damaging abilities to get enhanced effect", this effect of the Rune is just a chance to possibly increase crit chance and may in fact have zero more crits than would have happened without the proc over such a short duration (bad luck is bad luck and it happens frequently).

    As for Glurp, I have modeled extensively in FD. That's the only reason I gave the Rune a try for an entire raid night. If you search for profiles over 200k in the public settings, you will see my profiles. Because I have extensively modeled in FD, I have been able to observe a few things. First, FD is wildly erratic with haste. This appears to come from rounding with haste that in some cases causes small increments or decrements in haste to result in wildly differing results. For instance, if you take the current BiS profile that I created for BM without Ra-Den loot and change the chest enchant from being to haste to being to crit instead, you'll see a dps loss in FD of over 1k dps. If you look at the stat weights, the difference between haste and crit is not a factor (i.e., haste is not 3x better than crit or something). However, FD shows a major loss from a small change. Meanwhile, if you tweak the profile slightly to add more haste instead of crit, you will find a similar dps loss. The reason for this is because of rounding in FD regarding when an extra shot or ability will proc and not ACTUAL dps loss or gain, as seen in the game. Due to the nature of how FD works, rounding is natural and fine. It doesn't mean it absolutely reflects in game play--which is why you should test it in the real world and not just shout about the validity of FD. Because of this rounding of haste, it is often problematic to compare haste changes in FD (which Zeherah has noted). FD is a good approximator of performance and gear changes. You should not be basing an entire argument on plugging numbers into it, however.

    As for the claims made here about a single stat stacking being a bigger dps gain than a synergistic pairing of the exact same ilvl of secondary stats, that goes against theorycrafting that others have done, and I was simply tipping my hat to them. For instance, before RPPM, Nooska showed that the dps delta from increasing pairs in T14 did more dps than increasing any single stat. This behavior was validated by the Simcraft community (Arisen and Lockrick). Now, RPPM may have changed everything to favor haste more than crit for BM and MM to the point that haste is way overvalued now, and this synergy may be less apparent or even rendered obsolete. I'm not sure.

    As for the comment about 150 total stats versus 100 stats in one column--you lost me. If your point was that the Rune is now providing 2x the stats, then I admit I was analyzing under the old proc's effects (which are not true anymore) and that 2x the secondary stats should beat out a synergistic pairing of any type. My apologies.

    Also, the test was done in SV, but I think a better test would be in BM. As someone on EJ recently pointed out, BM has focus frenzy and the dps gain from using it or not using it is neglible (i.e., keeping your pet at 5 stacks of frenzy is about equivalent). One possibility with Rune would be to only use focus frenzy when the Rune procs, so you can get the haste gain from FF, with the full crit stack of the Rune. This should be a dps gain.
    Last edited by Effinhunter; 2013-04-26 at 05:42 PM.

  9. #989
    ^ You said

    Because of synergies between crit and haste and crit and mastery, we have known that stacking pure crit is worse than dividing secondary stats between crit and haste or between crit and mastery (i.e., you get more bang from your ilvl buck by not stacking any one stat--they harmonize together in pairs)
    That's only true because of the way ilvl budget works. If you have 2 items of the same ilvl, one with crit and one with crit+haste, the one with just crit will have less overall secondary stats than the one with crit+haste. That's why an item with crit+haste is superior to items with just crit. But when you're comparing crit vs haste on a point for point basis, then 150 crit will be better than 75 crit + 75 haste, assuming crit is actually better than haste.

    Anyway I'm not doubting that the Rune is shit. This is the first time Blizz has ever implemented something like that and it was either going to be wildly overpowered or more useless than a bucket full of elephant shit. But theoretically, that reason above is why it could be a gain.

  10. #990
    I mean, in theory, I feel like you could play with Rune at a high focus level (around 75-80 focus in SV) and have it work out, so you can minimize your CoS, and by thus "gaming" Rune you might make it passable.

    But that's a risky game. A bad LnL and you'll cap, an AMoC followed immediately by a proc and you'll be starved, ToTH will be right out as a talent. Seems like a lot of effort for not much benefit.

    I am curious though at the possibilities of the Frenzyx5 stacking for Rune procs. The difference between popping FocusFire after BW and not is so negligible that I feel like that may warrant an experiment.

  11. #991
    Quote Originally Posted by Effinhunter View Post
    Also, the test was done in SV, but I think a better test would be in BM. As someone on EJ recently pointed out, BM has focus frenzy and the dps gain from using it or not using it is neglible (i.e., keeping your pet at 5 stacks of frenzy is about equivalent). One possibility with Rune would be to only use focus frenzy when the Rune procs, so you can get the haste gain from FF, with the full crit stack of the Rune. This should be a dps gain.
    Just a note- the value of focus fire over leaving frenzy on your pet should be more significant as you get more RPPM items. Since the haste from focus fire factors into your RPPM rates, you see more value from using it with more significant RPPM gear. In addition, blink strike in 5.3 means more value from focus fire than frenzy, since focus fire translates directly into increased pet regen and thus increased wild hunt uptime or basic attack uptime on your pet.

  12. #992
    Quote Originally Posted by Zeherah View Post
    Just a note- the value of focus fire over leaving frenzy on your pet should be more significant as you get more RPPM items. Since the haste from focus fire factors into your RPPM rates, you see more value from using it with more significant RPPM gear. In addition, blink strike in 5.3 means more value from focus fire than frenzy, since focus fire translates directly into increased pet regen and thus increased wild hunt uptime or basic attack uptime on your pet.
    I completely agree that this should be a dps loss without the Rune in play. The only reason I have some hope of Rune being useful in a BM configuration is because FF seems to proc for me at about the same rate per minute as the Rune does (roughly once per minute). If this is the case, we'll potentially be losing 1-2 focus fires a fight, but we'll be gaining a large haste boost during a very high crit boost.

    I think it's at least worth an experiment. It won't be until next week until I can do a raid's night of testing on it again (I'm definitely not going to try it during progression content for us).

    Also, I'm going to edit my old post so that it does not cause anymore confusion. Apologies again for discussing the old proc of Rune instead of the latest one.

  13. #993
    FF gives a Haste %, not a number so it doesn't do anything with Rune of Reorigination.

  14. #994
    Quote Originally Posted by Glurp View Post
    FF gives a Haste %, not a number so it doesn't do anything with Rune of Reorigination.
    That's the point. They're gonna pop it with re-origination so that they still have some haste during the proc, which should keep the value of crit relatively high.

  15. #995
    Quote Originally Posted by Neazy View Post
    ^ You said



    That's only true because of the way ilvl budget works. If you have 2 items of the same ilvl, one with crit and one with crit+haste, the one with just crit will have less overall secondary stats than the one with crit+haste. That's why an item with crit+haste is superior to items with just crit. But when you're comparing crit vs haste on a point for point basis, then 150 crit will be better than 75 crit + 75 haste, assuming crit is actually better than haste.

    Anyway I'm not doubting that the Rune is shit. This is the first time Blizz has ever implemented something like that and it was either going to be wildly overpowered or more useless than a bucket full of elephant shit. But theoretically, that reason above is why it could be a gain.
    Nooska's post at EJ and the follow up by Arisen and Lokrick was based on taking a static allocation of 1000 secondary stats and putting it into various configurations of crit, mastery, crit/haste, crit/mastery, or crit/haste/mastery. It is not specific to your item discussion. It should be applicable here.

    Also, are you absolutely sure that gear is penalized for a single stat? Beady-eye Bracers do not appear to have that penalty. They seem to have slightly less secondary stats because of the socket and the bonus. For instance, in my armory, I have the Thunderforged Beady-eye Bracers for expertise. They have an allocation of 907 expertise. The Thunderforged Mutagenic bracers have a split allocation of 1182 between haste and mastery. This may appear to be a "penalty" on the bracers of 3:2, but the socket and the bonus are accounted for in the ilvl stat allocation too.

  16. #996
    ^ You even said it yourself: 907 expertise for your bracers vs. 1182 combined haste/mastery for the other bracers. The socket and bonus usually (always?) comes out of the agility allocation, not the secondary stats.

  17. #997
    Quote Originally Posted by Neazy View Post
    That's only true because of the way ilvl budget works. If you have 2 items of the same ilvl, one with crit and one with crit+haste, the one with just crit will have less overall secondary stats than the one with crit+haste. That's why an item with crit+haste is superior to items with just crit. But when you're comparing crit vs haste on a point for point basis, then 150 crit will be better than 75 crit + 75 haste, assuming crit is actually better than haste.

    Anyway I'm not doubting that the Rune is shit. This is the first time Blizz has ever implemented something like that and it was either going to be wildly overpowered or more useless than a bucket full of elephant shit. But theoretically, that reason above is why it could be a gain.
    Effinhunter wasn't talking about the ilvl budget. The experiment by Nooska he linked examined how adding 1k secondary stat would affect dps. This was his result (copy & pasted & refomat from the link earlier):

    Set--------------------------DPS--------Delta / %
    Base-------------------------88513.20---∆ 0
    1k Haste---------------------89442.31---∆ 929.11 / 1.05%
    1k mastery-------------------89629.79---∆ 1116.59 / 1.26%
    1k crit-----------------------89790.17---∆ 1276.97 / 1.44%
    500crit/500mast--------------89996.48---∆ 1483.28 / 1.68%
    500haste/500mast------------89616,76---∆ 1103.56 / 1.25%
    500crit/500haste-------------89590.91---∆ 1077.71 / 1.22 %
    334crit/333haste/333mastery--89771.90--- ∆ 1258.7 / 1.42%

    As you can see, splitting crit/mastery was the best dps gain (even better than pure crit). It is a rather interesting post and keep in mind it was done with pre-5.2 data. Changes to abilities and ilvls may make this no longer valid

  18. #998
    Quote Originally Posted by Tolaez View Post
    Effinhunter wasn't talking about the ilvl budget. The experiment by Nooska he linked examined how adding 1k secondary stat would affect dps. This was his result (copy & pasted & refomat from the link earlier):

    Set--------------------------DPS--------Delta / %
    Base-------------------------88513.20---∆ 0
    1k Haste---------------------89442.31---∆ 929.11 / 1.05%
    1k mastery-------------------89629.79---∆ 1116.59 / 1.26%
    1k crit-----------------------89790.17---∆ 1276.97 / 1.44%
    500crit/500mast--------------89996.48---∆ 1483.28 / 1.68%
    500haste/500mast------------89616,76---∆ 1103.56 / 1.25%
    500crit/500haste-------------89590.91---∆ 1077.71 / 1.22 %
    334crit/333haste/333mastery--89771.90--- ∆ 1258.7 / 1.42%

    As you can see, splitting crit/mastery was the best dps gain (even better than pure crit). It is a rather interesting post and keep in mind it was done with pre-5.2 data. Changes to abilities and ilvls may make this no longer valid
    The quote I was responding to specifically mentioned ilvl, but yeah that is worth knowing.

  19. #999
    Quote Originally Posted by Neazy View Post
    The quote I was responding to specifically mentioned ilvl, but yeah that is worth knowing.
    The quote was not focusing on ilvl. I was discussing stat allocations in a vacuum as discussed by theorycrafters. You had mentioned ilvl in a previous post, and so I think you viewed my subsequent post with that lens.

    Also, a socket is taken from the stat allocation budget, which is dictated by ilvl of the item, which is where agility or any other stat also come from. Blizzard can move the stat budget around as it pleases when designing an item and has even done some creative stat allocation budget accounting in regards to certain items like the Iron Qon ranged weapon, which is overbudget and may one day be nerfed (though, obviously, I hope they don't ;D). However, the socket and its bonus tend to revolve around the stat allocation budget associated with gems available at the start of an expansion. This has allowed us to change items into ones that defy the default stat allocation budget for ilvl by changing 80 agi and ~300 secondary stats, for instance, into more agility (which has a much better return than the secondary stats) through clever usage of both gems (which often have epic and legendary options within an expansion) and the bonus that is available (especially when it awards agility for matching the color). From what I can tell, the socket in the Beady-eye takes away a budget that is reflected in the loss of 80 agi and ~300 secondary stats, with a potential gain of 80 agi / 160 crit or haste and +60 agi modifier--a very favorable trade of secondary stats to more agi and not a penalty.
    Last edited by Effinhunter; 2013-04-26 at 10:39 PM.

  20. #1000
    This is the quote I was responding to (I even quoted you in my first post):

    Because of synergies between crit and haste and crit and mastery, we have known that stacking pure crit is worse than dividing secondary stats between crit and haste or between crit and mastery (i.e., you get more bang from your ilvl buck by not stacking any one stat--they harmonize together in pairs)
    But this is getting ridiculous. We solved the issue.

Posting Permissions

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