Page 16 of 18 FirstFirst ...
6
14
15
16
17
18
LastLast
  1. #301
    Moderator Yvaelle's Avatar
    Join Date
    Jan 2009
    Location
    London, England
    Posts
    4,380
    Quote Originally Posted by Ferin View Post
    Well raidbots top parses is the median of the top 100 which isn't bad statistically - you're implicitly stripping off the people gaming the mechanics and the shitters by using the median rather than the mean.
    The median of the top 100 (50th) out of like 30 million parses, is still generally an anomaly.
    Youtube ~ Yvaelle ~ Twitter

  2. #302
    Quote Originally Posted by Yvaelle View Post
    The median of the top 100 (50th) out of like 30 million parses, is still generally an anomaly.
    Pretty sure there aren't 30 million parses for a given fight with a given spec (which is what the top 100 is selected from).

    In fact you can check how many samples top 100 is vs all samples on raid bots since it has a samples graph. In the worst case (affl lock) you're selecting the top 1600 of 29151 samples, so that's the top 5.5% of samples. For most specs it's closer to the median of the top 10%. Taking the median of that isn't particularly anomalous IMO.

  3. #303
    I'm not sure on that high skill cap. I know I did get ranks even while not playing flawless.

    My point wasn't that the top ranks are gaming mechanics; they are not necessarily. However I remember in DS one of the factors being just that (Madness in particular for DoT classes), and for example for healers to get ranked using 1 healer less say with aid from off healing classes (omg,found a use for shadow ) you'd do less overhealing and you'll get ranked. So I do know it happens, but there's advanced game mechanics study required per boss to determine this.

    Quote Originally Posted by Ferin View Post
    Well raidbots top parses is the median of the top 100 which isn't bad statistically - you're implicitly stripping off the people gaming the mechanics and the shitters by using the median rather than the mean.
    We should exclude statistical flukes in either direction. If we were to include all the shitters (no offense to anyone ofc) we'd also include those who are not able to play their class well. We really shouldn't include those. We also should exclude those who were lucky on crits/procs. Why? Because they're exceptions to the rule. When we exclude those who were lucky with crits/procs we do not include how often those happen in our statistic; this is a problem in accuracy! Especially for crit-based classes, such as fire mage. This was excluded in the compare done (the one I linked to in previous post) and I am now not so sure if we were right to exclude them.

    I'll give you another example unrelated to crits and procs. I remember fire mage being OP for less than a week (don't remember exactly the details how it worked, doesn't matter). Unfortanately for our mage it was nerfed before the next raid (and even then, he didn't have good gear, being the most unlucky fucker on charms and LFR and drops in N/H I have ever seen). The result was that, next raids, it was much higher for him to get ranked again. Should we include those pre-nerf parses or not? I say not. They were a one-time (less than a week) fluke.

    Your argument however isn't that we should include those statistical flukes on the high end (the top parses); your argument is we should only use those. With the above in mind, given the people who play 75-100% are playing their class reasonably well, we shouldn't discount those in the statistic; on the contrary! Which is why the median of raidbots top 100 parses is a completely irrelevant statistic, and I'm sure I'm not the first one to mention this.
    "When i am done with you, you won't trust your own mind."

  4. #304
    Quote Originally Posted by lolalola View Post
    Your argument however isn't that we should include those statistical flukes on the high end (the top parses); your argument is we should only use those. With the above in mind, given the people who play 75-100% are playing their class reasonably well, we shouldn't discount those in the statistic; on the contrary! Which is why the median of raidbots top 100 parses is a completely irrelevant statistic, and I'm sure I'm not the first one to mention this.
    My argument isn't that at all. Taking the median of the top 100 isn't only taking the statistical flukes; unless you consider the majority of the top 100 a statistical fluke. Given we're talking about the top 10% or so of logs, that's a pretty bad data set if they're all fluke.

    Taking the median is a good idea because it eliminates bias from very high or very low outlying logs. These are pretty common on WoL - if you look at the top 2 or 3 logs for any fight you'll often see a huge gap from those underneath. These are either the luckiest people alive, or a deliberately broken log for the purpose of getting the top spot.

    You can argue that the percentile should be larger than the top 100 yields, or that we should use a fixed percentile, but that doesn't make the top 100 completely irrelevant, just not quite the optimal stat.

  5. #305
    Quote Originally Posted by Ferin View Post
    [...] unless you consider the majority of the top 100 a statistical fluke. [...]
    That was the entire point of my post indeed...

    Given we're talking about the top 10% or so of logs, that's a pretty bad data set if they're all fluke.
    They're flukes compared to those logs where people played well but weren't lucky on crits/procs/mechanic_abuse etc. Those who are not top geared but are still playing their class well do count IMO. They're a much larger group than those who got ranked this week.

    You can argue that the percentile should be larger than the top 100 yields, or that we should use a fixed percentile, but that doesn't make the top 100 completely irrelevant, just not quite the optimal stat.
    I think they're relevant, but not on their own by excluding say the 75-100% because of the reason I mentioned above.
    "When i am done with you, you won't trust your own mind."

  6. #306
    At this rate all of the current threads will be arguments about the statistical validity of picking parses from raidbots showing how behind shadow dps is or isn't. I swear that's all I seem to be reading lately...

    We're behind, that's all that really matters at this point, whether you take the top x% or not. Those of us raiding know we're behind by a significant margin on single target. The exact percent behind, compared across the specs, is a constantly changing value. Blizzard has repeatedly dismissed references to class rankings from log parses when people were asking for balance changes. Discussing the reasons behind the disparities seems to be a more promising way to go, although some good posts have been dismissed on that front also (i.e. stat scaling).

  7. #307
    Devouring Plague initial damage has no base damage anymore and just scales based on spellpower, rather than a base plus a lower amount of spellpower. The DoT component damage has been reduced by 40%.
    They intend to nerf shadow in pvp while buffing it in pve and the solution is to increase DP's upfront burst and reduce the dot? Is this a joke?

  8. #308
    Quote Originally Posted by Cookie View Post
    They intend to nerf shadow in pvp while buffing it in pve and the solution is to increase DP's upfront burst and reduce the dot? Is this a joke?
    Couldn't it mean a buff for PvE and nerf for PvP seeing as PvP gear has less spellpower and it fixes our scaling issues in PvE a bit? I dunno, not much of a theorycrafter.

  9. #309
    Yes, I'm all for increasing the scaling of our spells and that part of the change is great, but also putting the lionpart of the damage at the instant just gives them less room to buff the rest of our damage (without making us too strong in pvp).

  10. #310
    Those of us raiding know we're behind by a significant margin on single target. The exact percent behind, compared across the specs, is a constantly changing value.
    This part I think is not in dispute. I think a lot of the methods to show that priests are behind rely on methods that are iffy, and it's good to attack either ones that are outright silly (for instance, the top 100 is not a good source) or that won't convince the devs (comparing to affliction and arcane, for instance, after GC has said he considers those too high- the devs likely ignore ANY comparison to a spec that they consider an outlier). But even once you throw those out, it's clear that priests are not hauling the numbers for their gear.

    I went into this tier claiming that spriests were going to be a lot lower than in history. I did this when I saw the talents- the priest utility is quite strong, and priests can gain rewards for off healing no matter their spec (for instance). I believe that Blizzard has generally controlled dps factoring in powerful utility, but is simply unwilling to say it (GC eventually had to stop mentioning the hybrid tax, for instance- even after he stickied it, it was still like 15% of forum topics. Now it's in the design, but he can't talk about it.). Their game design is simply back to being more opaque because the players couldn't handle the truth- and also probably because the devs are too scared to have the players prove when they are wrong.

    So for this reason, I've been recommending priests take angles that WILL work even assuming some "hidden variables". Shadow IS too low- probably even by Blizzard's hidden metrics- but we need to demonstrate that to them both on live and on the PTR if they are going to address it. You might point out that shadow, for instance, has a rather poorly defined dps niche on live- some of the hyperbolic posts express the idea of "gosh, I should just be a lock". Pure versus hybrid aside, this is a very good point, and you could probably use it to angle for an edge that a lock doesn't have (given that the lock has several you don't).

    I also don't really see the hardon they have for keeping shadow's single target generally low. I don't get having shadow orbs only do one damned thing really. I mean, as a rogue, you'd feel stupid if your only finishers were like kidney shot and eviscerate. An example additional would be, say, a Mental Destabilization magic-type debuff. This could last 30 seconds at one orb, 70 at two, and 120 seconds at three, and apply both the "weakened blows" debuff and ALSO increase damage taken from the shadow priest by X%- where X is some smallish number, such as 5-10% or whatever it would take. Since this COSTS damage to put up, it wouldn't be a gain in a very short fight (and as such, the buff should be kept low). Since it both costs damage and is dispellable and is of low magnitude, it wouldn't show up much in pvp- maybe under some niche situations. Since it gives you more payoff per orb it would be best to use at three early and then tweak later so you don't throw a 120 second one onto a boss that will last only 35 seconds. It would also give you another use for the orbs, which currently only have one pve use (and honestly, normally only one pvp use too).

    Now, that's not the only fix. There's plenty, and I'm sure some folks hate buffs like that. I think those things are cool, but I'm a rogue. There's a lot of damned fixes that don't up pvp damage, the big question is, can you get Blizzard on board with needing them in the first place? And I think you can.

  11. #311
    I just quit the game due to the state of priests in general, really not enjoying disc, holy I never liked and shadow, well its just bad. I have to work harder on my priest to pull the same numbers on my ele shammy who is lesser geared.

  12. #312
    Guys, I recognize your [our] concerns about Shadow, and I'm trying really hard with GC on the PTR class balance thread (and a other hundreds of Spriests), but seriously: Shadow is not "that" bad. We have some problems with ST and movement and we miss some burst. A buff to PI and making it stack with BL/hero could resolve this (I know it's only a band-aid). And while our damage on ST (especially when there's movement) is really bad, we're not that far. MT we are kinda good, and fights where we can use ToF we excel.
    We need a buff, of course, but saying that we are bad to the point of stop playing... Well, I just disagree.
    Last edited by andre1204; 2013-02-09 at 04:27 AM.

  13. #313
    The f#ck is wrong with them? Yeah lets nerf DP's dot damage... Srsly where is the dot class anymore...? If they do something like this, can they have the freaking decency to change this half assed mastery....

    Quote Originally Posted by andre1204 View Post
    Guys, I recognize your [our] concerns about Shadow, and I'm trying really hard with GC on the PTR class balance thread (and a other hundreds of Spriests), but seriously: Shadow is not "that" bad. We have some problems with ST and movement and we miss some burst. A buff to PI and making it stack with BL/hero could resolve this (I know it's only a band-aid). And while our damage on ST (especially when there's movement) is really bad, we're not that far. MT we are kinda good, and fights where we can use ToF we excel.
    We need a buff, of course, but saying that we are bad to the point of stop playing... Well, I just disagree.
    We never excel, there is no fight where we excel. Locks/mages are always in front of us and always with a noticeable margin. ToF makes us good in certain situation, but not top tiers.
    Last edited by Falsified; 2013-02-09 at 05:07 AM.

  14. #314
    Quote Originally Posted by Falsified View Post
    We never excel, there is no fight where we excel. Locks/mages are always in front of us and always with a noticeable margin. ToF makes us good in certain situation, but not top tiers.
    Well, to be fair, lock/mages are ahead of everyone else by a good margin hehe. GC does make it sound like they are going to bring them back down though with his recent tweets.

  15. #315
    you really should listen to yourself... bitchin about shadow... since when do u play? cata? well if thats the case, ofc.: ur disapointed... but to remember, for a hybrid class, shadow was way too op in cata... we´re actually not comparing well in a single target dps fight, i´ll give u that, but always notice, which other class besides us can give the raid group 40 k hps and 150 k dps in a boss fight? (talking ´bout garalon as an example ) ? before ur bitchin me, i´m playin shadow since end of vanilla, i´m kinda used to be the ,"in your eyes dissapointing specc getting overlooked cause we all wanna be op as in cata",god damnit i played shadow in BC for crying out loud but to summon up, even though we do have problems in single target fights, we offer so much more to a raid then a plain dps class...our offheal is worth so much more then the 5 % more dmg, i know on sims and in the worst case 20%, thats like 0,0001% of the top world warlocks and mages, stop whine... GC is actual aware of the lack referring too single target dps, as u can see in the latest blue posts...how too fix our dps? we got good insights here.. all insights will fix our dmg... but you ever thought about our raid abilities? what would become of a shadow if u would actually take over the dps fixes of this thread? think about it will scale with vampiric embrace and so on... yeah, you already got it... we will be the most op class again... that wouldnt play into blizz cards, actually wanting every class comparing on the same level..., all i am asking for here is see the class you are playing in perspective... we are actually descent on multitarget, fewest bosse are single target where i admit that we are little bit too far behind, otherwise we are a good dps class, pushed and loved for our offheal cooldowns ... thank you for listening... p.s.: i´m not that good in english, sry for any gramma fails ,<3 u all... sincerely flens shadow since vanilla and for the rest of his wow lifetime
    Last edited by flens; 2013-02-09 at 06:17 AM.

  16. #316
    Since when are shadow priests doing 40k HPS?

    Quote Originally Posted by steveyboy View Post
    I just quit the game due to the state of priests in general, really not enjoying disc, holy I never liked and shadow, well its just bad. I have to work harder on my priest to pull the same numbers on my ele shammy who is lesser geared.
    I'm sorry but ele is a lot worse off than shadow is right now. That's probably an l2p issue.
    Quote Originally Posted by Karragon View Post
    I'd like WoW to be a single player game

  17. #317
    Quote Originally Posted by blargh312 View Post
    Since when are shadow priests doing 40k HPS?

    .

    I'm sorry but ele is a lot worse off than shadow is right now. That's probably an l2p issue.
    1.: as stated garalon is an example....

    2.:your right, doesnt concern the priest forum though

  18. #318
    Quote Originally Posted by flens View Post
    1.: as stated garalon is an example....
    Please link
    Quote Originally Posted by Karragon View Post
    I'd like WoW to be a single player game

  19. #319
    For a hybrid class? How about boomkin totally owning in fireland..?
    Oh and can you offer a raid portal....? Do you offer so much self heal that healers just go "meh its a lock, no need to heal"...(jk, but you see my point)
    And utility is by no mean a reason to be rock bottom on single target fights... FYI hybrid tax is supposed to be gone since tlk, only one expansion had it(bc) and it was well understood by every one why(mana battery ftw). If it actually was 5% damage I wouldn't give a crap but its not a minority its the overwhelming majority. As you said the only thing we bring in a raid is HPS utility, which to be fair is the worst utility in game, healing cool-downs or damage reduction cool-downs are much better(VE is utter shit horse in 25m) or raid movements bonuses(locks portal, druid's shout). Also, if what you said was true about our healing utility, then most guilds would bring 2 SP instead of a healer, yet I have serious doubt that anyone did that. You are better off taking a healer and a real damage dealer. You're bitching about what we were, I don't give a crap what happened 5 years ago, no class is like what they were back then.
    Wow is homogenized, It means hybrid tax is gone. The only utility we have is MD, and dispersion for soaking mechanics. Our healing utility is only good with Divine star, and that represents the talent with which we do the least dps, but we have to take it some times(when we can't halo).
    And finally the biggest utility any dps can bring is dps. You bring a tank for tanking, a healer for healing and a dps for dps. And I am only bitching about our single target damage, which MB and DP are near 40% of it. And seeing they nerfed the dot part of DP, yeah i can bitch about it, it was totally uncalled for. I honestly don't see the point of it. And most people here agreed that SP design is flawed, it was said that SP would suck in t14, it was true in single target, but ToF managed to save us on several encounters. Mind sear is good too. But the fact is, we currently got one of the worst single target dps in game, and we can't really switch specs.

    And lastly, GC might be aware, but by no means has he said once that our damage was too low. He's been dodging it for a while now. So yes, we might assume that they know, but they have yet to share anything with us, the one who pays.

  20. #320
    Quote Originally Posted by blargh312 View Post
    Please link
    i´d like to post, not able to though, talking bout the first weeks of content release... links all expired on WoL... looking at the logs now, ofc healers are outgeared u will not find a log confirming my personal experience... anyhow, WoL rank 1 still manages to pull 20k hps even though he does not keep pom on cd and doesnt use symbiosis tranq... nothing more to add...

Posting Permissions

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