Page 1 of 2
1
2
LastLast
  1. #1

    [3v3] PvP Representation (EU)

    [EU] 3v3 Representation

    Introduction
    Hey guys, just like to start this off by saying this isn't a 'flame x class' thread. I'm just throwing some figures and graphs out there for you guys to see. I'll maybe do some analysis on a bit of it too. I'm all up for discussing spec balance, ways to fix outliers in representation and other topics in this thread, but I don't want to have to hand out infractions on a thread that I've made.

    I'm making this thread as I was looking at representation myself and thought it would be cool to share with you all, and to see how it compares to in a few weeks when arena sites get updated and we get a new set of numbers.



    Contents
    • Class Representation
    • Class Representation (Graph)
    • Healer Representation (Graph)
    • Damage Dealer Representation (Graph)
    • Analysis




    Overall Class Representation
    This is all classes with all specs. There is 1002 total characters (I presume they were aiming for 1000) on the EU ladder, and simply by filtering each spec you get a hit with the number of characters listed under that spec. The numbers are as followed:

     Death Knight - 88
     Druid - 56
     Hunter - 89
     Mage - 84
     Monk - 13
     Paladin - 67
     Priest - 202
     Rogue - 24
     Shaman - 149
     Warlock - 66
     Warrior -164

    Found here: http://eu.battle.net/wow/en/pvp/leaderboards/3v3





    Class Representation (Graph)



    As you can see, Priests are trailing far infront of everyone else. This is mainly due to all 3 of their specs being viable at the moment (48 Disc's, 89 Holy's, 65 Shadow's). Next up is Warriors; which seem to be the most dominant non-hybrid class which are then followed by Shamans which, like Priests, have 3 viable specs. After that, it's a pretty large jump from 14.9% to 8.9% where you get Hunter, DK, Mage, Paladin, Warlock and Druid all fairly close to one another. Monks and Rogues both seem heavily under represented - which seems to be a common theme for them throughout MoP.





    Healer Representation (Graph)



    This is a graph purely comparing the healing specs of the 5 classes. Arguably the most shocking bar on this chart is the Mistweaver Monk column, whose class is repping a whole 0 players in the top 1002. The second most shocking is Priest, with a whopping 42.8% of all healers in the top 1000 being them. The split is mostly in Holy's favour, although Disc is a large proportion of it. Shamans are very high too, considering the nerfs they have received and though Druids were arguably the strongest healer for a long time on PTR, they seem to be very underpar. Remember, if balance was perfect, healers would all be around 15-17% representation each.





    Damage Dealer Representation (Graph)



    If you've played much 3v3 so far this season there is nothing surprising in this graph. Warrior representation has skyrocketed since last season, while Hunter, Feral and Ele have all dropped significantly. Death Knight, Warlock, Shadow Priest and Mage representations have all risen slightly - although this could be put down to their synergy with Warriors. A comparison graph with data from the last week before the Tyrannical season ended can be found here.





    Analysis
    It's obvious which classes are performing better than others. The most surprising figure for me is the Mistweaver percentage, with not a single mistweaver in the top 1002 players in EU (which is only 2k rating). Warrior representation is very high, this is probably down to not only the strengths that the class has but also their synergy with nearly every spec in the game. The handful of specs Warriors doesn't work well with (Rogues, Rets, WW Monks) have all seen drops in representation. Shadow Priest representation is still high, which could be the reason for Resto Shamans holding their own on ladder - as you don't want to double stack Priests in 3s.

    Hunter representation is interesting because nearly every Hunter on ladder is now survival (65 of 89), with only 17 playing beast mastery and yet the class is still performing above average.

    DPS classes with percentage decreases in representation are (biggest to smallest): Feral Druid (-71%), Balance Druid (-69%), Ret Paladin (-69%), Elemental Shaman (-48%), Hunter (-37%), Rogue (-36%), Enhancement Shaman (-24%) and then Windwalker Monk (-9%).

    DPS classes with percentage increases in representation are (biggest to smallest): Warrior (+324%), Death Knight (+73%), Mage (+31%), Warlock (+16%) and Shadow Priest (+5%).


    Trivia
    • There is currently 3 more Blood Death Knights than Combat Rogues, Mistweaver Monks, Frost Death Knights and Fury Warriors added together.
    • There is only a 57.2% chance that the enemy healer in 3s isn't a Priest.
    • Mathematically, it is almost guaranteed that all of the games you play will also have a Warrior in them (24% * 4 = 96%). Statistically speaking, if you don't have a Warrior in your team... then the enemy will in theirs.
    • If you take all of the Arcane Mages, Fire Mages, Demonology Warlocks, Assassination Rogues, Balance Druids and Marksmanship Hunters on ladder, then you still can't kill a Warrior (this one's a joke).
    Last edited by Snuggli; 2013-09-23 at 09:12 PM.

  2. #2
    Nice numbers-work there !
    Not talking about balancing issues, but representation should give a general overview of what will end up high on the ladders I guess ?

  3. #3
    Quote Originally Posted by Aevyn View Post
    Nice numbers-work there !
    Not talking about balancing issues, but representation should give a general overview of what will end up high on the ladders I guess ?
    Generally speaking, yeah. If something is over represented it normally means it is overpowered (or had a lot of hype), by looking purely at the most played DPS class, then picking one of the other most played DPS classes and then putting it with the best healer you get comps back like: Warrior/Hunter/Priest, Warrior/Mage/Priest, Warrior/Lock/Priest, Warrior/DK/Priest and Warrior/Spriest/Shaman - which are probably the comps that they are playing to reach the top 1000.

    There is sometimes seasons when comps are strong but underplayed so you should always take representation with a pinch of salt, but it's definitely a good indicator of what is performing well.

  4. #4
    Can you promise to do another one of these to compare the first week stats of 496 geared players to the full griev geared players later on? I am thinking some classes are a little under performing right now because of the damage suppression of higher damage reduction but not much damage increase from gear just yet. Hell it would be even interesting to see the end of next week when all those melee classes pick up their weapons.

    I can speak as a DK that my survivability is much stronger because of the slower damage output, but my pressure has remained somewhat constant because of the healing nerf as well. I am scared that once the gear ups the damage a bit that I will fall back out of favor because it will be easier to burst me down during a long cross CC chain.

  5. #5
    Quote Originally Posted by Tyrean View Post
    Can you promise to do another one of these to compare the first week stats of 496 geared players to the full griev geared players later on? I am thinking some classes are a little under performing right now because of the damage suppression of higher damage reduction but not much damage increase from gear just yet. Hell it would be even interesting to see the end of next week when all those melee classes pick up their weapons.

    I can speak as a DK that my survivability is much stronger because of the slower damage output, but my pressure has remained somewhat constant because of the healing nerf as well. I am scared that once the gear ups the damage a bit that I will fall back out of favor because it will be easier to burst me down during a long cross CC chain.
    Yeah definitely. Longer games obviously favour some classes over others - same with some classes scaling better than others with higher ilevels. I'll probably do one in a few weeks when people have weapons and teams sorted out - and once the initial "omg reroll" hype is over. This is only if websites like worldofwargraphs haven't updated themselves to handle the new ladders though.

  6. #6
    Field Marshal XMD's Avatar
    Join Date
    Apr 2013
    Location
    Bavaria
    Posts
    75
    Just wait for it, hunters never played survival spec in arenas. They are currently "learning their new class", in order to come out on top in a few weeks (after warriors receive their little corrective).

    Btw. nice work, I like statistics, even if they cannot reveal the whole truth due to how complex and dynamic the whole system is.
    US statistics would also be interesting, for comparison.
    Last edited by XMD; 2013-09-23 at 09:20 PM.

  7. #7
    The Patient
    Join Date
    Sep 2011
    Location
    New Hampshire
    Posts
    229
    some of the thing that I found strange about the graph. you split some of the dps and stack the other. ie enc and ele while dk is grouped up.

    otherwise its a good read.

  8. #8
    Quote Originally Posted by MuskiER View Post
    some of the thing that I found strange about the graph. you split some of the dps and stack the other. ie enc and ele while dk is grouped up.

    otherwise its a good read.
    It's mainly for pure DPS classes where there is either not that much difference between the specs, or where one of the specs is played wayyyy more than the other.

    Here are how many players there are under each spec:

     Death Knight - 88: 85 unholy, 3 blood, 0 frost
     Hunter - 89: 65 survival, 17 beast mastery, 7 marksmanship
     Mage - 84: 78 frost, 4 fire, 2 arcane
     Rogue - 24: 17 subtlety, 7 assassination, 0 combat
     Warlock - 67: 60 affliction, 5 destruction, 2 demonology

  9. #9
    Is the Season 13 gladiators by class data out yet guys?

  10. #10
    Stood in the Fire Kals's Avatar
    Join Date
    Jul 2011
    Location
    Vermont
    Posts
    481
    Nice graphs, I love the class decreases and increases percentages; shows that a LOT of people sacrifice playing a class they've played for a long time in order to play something OP/better. Blizzard balance for you though I guess. That jump in warriors is insane (albeit expected) tho.

  11. #11
    Quote Originally Posted by Kals View Post
    Nice graphs, I love the class decreases and increases percentages; shows that a LOT of people sacrifice playing a class they've played for a long time in order to play something OP/better. Blizzard balance for you though I guess. That jump in warriors is insane (albeit expected) tho.
    When I got the Warrior % increase number I double and triple checked it because it was so much bigger than the rest that I thought I had done my maths wrong. It's not like they were that poorly represented last season either, there were still more Warriors than Ferals above 2.2k last season.. and to have such a huge % increase in 1 week, it's crazy.

  12. #12
    The Patient
    Join Date
    Sep 2011
    Location
    New Hampshire
    Posts
    229
    Quote Originally Posted by Snuggli View Post
    It's mainly for pure DPS classes where there is either not that much difference between the specs, or where one of the specs is played wayyyy more than the other.

    Here are how many players there are under each spec:

     Death Knight - 88: 85 unholy, 3 blood, 0 frost
     Hunter - 89: 65 survival, 17 beast mastery, 7 marksmanship
     Mage - 84: 78 frost, 4 fire, 2 arcane
     Rogue - 24: 17 subtlety, 7 assassination, 0 combat
     Warlock - 67: 60 affliction, 5 destruction, 2 demonology
    thanks ! was kind of curious to see the fotm spec.

    Also just text my bud to get ready to be one of the first mw above 2k. on us anyway.

    just too busy last week or so to make any real push

  13. #13
    Moderator Yvaelle's Avatar
    Join Date
    Jan 2009
    Location
    Vancouver, Canada
    Posts
    5,466
    Awesome post Snuggli

    I'd be very interested to see a follow-up when full grievous is common at high rating toward the end of the season.

    If I may speculate - higher item level but comparable resilience will lead to higher burst and a shift in the balance of certain survivability cooldowns which are very effective right now. Some classses have sufficient survivability to survive today's burst, but won't gain real survivability out of the item level increase. For example, Shadow's survivability is essentially fixed - resilience won't significantly rise, dispersion is a near immunity, and the small armor on items will result in a very tiny (almost trivial) survivability increase. By contrast, Sacrificial Pact scales off 100% of health, so gear will increase stamina - and thereby increase max health, and sacrificial pact size.

    So I would expect Shadow to become squishier as the season goes on and item level rises, but Warlock survivability should remain relatively comparable (I'm oversimplifying, but Dark Regen, Drain Life, Corruption, etc all scale off max hp - which should rise significantly). Based on that I'm expecting my Warlock to do better than my Spriest, and on a large scale - I think we will see Warlock representation grow as the season goes on (and possibly Spriest rep fall).

    Similarly, warriors are doing absurdly well again - but that is a function of Second Wind + Defensive Stance + 75% Resilience. I'd speculate that as burst rises, the effectiveness with which people can burst through Warrior Phase 2 (Second Wind phase) will become more common - and warrior effective survivability will decline as the season continues.

    All speculation, of course
    Youtube ~ Yvaelle ~ Twitter

  14. #14
    By the looks of it I guess this is going to be another awful season then.

  15. #15
    Titan apepi's Avatar
    Join Date
    Dec 2008
    Location
    Mostly harmless
    Posts
    12,651
    Nice math there, Have you checked at the US one to see if we are much different from eu? Also you said if something is over represented it means that they are overpowered, so if someone is underrepresented then it is underpowered? Or does that not have a correlation?
    Time...line? Time isn't made out of lines. It is made out of circles. That is why clocks are round. ~ Caboose

  16. #16
    Pandaren Monk docterfreeze's Avatar
    Join Date
    Dec 2012
    Location
    Finding a stranger in the alps.
    Posts
    1,989
    priest #1 healer. Color me surprised.

    Words to live by.

  17. #17
    Stood in the Fire Kals's Avatar
    Join Date
    Jul 2011
    Location
    Vermont
    Posts
    481
    Quote Originally Posted by apepi View Post
    Nice math there, Have you checked at the US one to see if we are much different from eu? Also you said if something is over represented it means that they are overpowered, so if someone is underrepresented then it is underpowered? Or does that not have a correlation?
    I think underrepresented in most cases either means not very viable in PvP or hard to play (MW, Combat Rogues, Balance Druids) or had mechanic changes which made them less enticing to those who played them last season (Feral clone change, mage icicles, ele RNG "fix", ect).

    Also the factor that some of those classes/specs as previously stated not only don't go well with warriors but might be completely countered by warrior teams (hard casting classes such as Ele and Doomkin for example).

    Furthermore some classes scale way better with gear, and especially weapons (rogues and ferals are big) which means the resil change sort of screwed them over until they get more grievous, or at least weapons.

    Just my 2 cents on that at least.

  18. #18
    One more warrior season inc ! gotta love it

  19. #19
    Moderator Yvaelle's Avatar
    Join Date
    Jan 2009
    Location
    Vancouver, Canada
    Posts
    5,466
    Quote Originally Posted by apepi View Post
    Nice math there, Have you checked at the US one to see if we are much different from eu? Also you said if something is over represented it means that they are overpowered, so if someone is underrepresented then it is underpowered? Or does that not have a correlation?
    Representation and power are correlated, but the correlation is indirect and inconclusive.

    If Priests have triple the active pvp community of say, Rogues - if we assume that gladiator-level players are equally common amongst both populations - then it should be fair to assume that priests would have triple the representation of rogues at high ratings, not because they are overpowered - but simply because they are a more popular/fun class/playstyle. Alternately, just because rogues have consistently low representation at high ratings, doesn't necessarily mean that rogues are consistently underpowered - they could be way stronger than priests and still not hope to achieve the tripled representation that priests would have over them merely because of the larger population size: it works both ways.

    Then you have the distribution of knowledge - for example - some specs and classes have prolific communities which catalogue and share every tip and trick and example of how to succeed with that class: other classes do not. To use priests again, we have a very active pvp community - tons of priest-specific pvp resources - and frequent interactions with many of the top priests in the world: either through Twitch, or Youtube, or Forums such as these. By contrast, hunters often suffer from a lack of information sharing between their best and their beginners - it can be very hard to learn how to play a hunter better in pvp, simply because the knowledge isn't well documented and the experts are few, or keep to themselves. We can watch Talbadar or Theed or Hydra play Priest and see how they play it, what tricks or positioning or builds they choose against various comps - someone performs verbose math to ascertain the best gearing path, or best talents for various situations. By contrast, hunters (thinking specifically of Cata more than MoP) have had many in low brackets - and then a handful of rank 1's: with little or no middle ground. Gaps like that are the result of either knowledge gaps, or skill gaps.

    Skill gaps occur when a class has a very high skill ceiling, and/or a very sharp learning curve. If the class is useless until you can play it well, and then it's good - that's a learning curve - alternately if anyone can pick up a class and do well with it - that's a relatively flat learning curve. If anyone can pick up a class and do well with it, but even the masters struggle to gain greater performance out of it - that's a low skill ceiling: there is only so much the class is capable of doing. In contrast, some classes are extremely difficult to master - you can always learn something new you could be doing better: that's a high skill ceiling.

    I should be saying spec here, not class, sorry. So, a spec can be 'easy to learn, hard to master', or 'hard to learn, hard to master', or 'easy to learn, easy to master', or 'hard to learn, easy to master'. I won't try to label my opinion of specs into those categories because it's subjective - people will find different classes more suitable to them and that changes the quadrant in which they belong.

    Lastly (the last I can think of at the moment anyways), we have situational context - every pvp encounter occurs within a context, and it's possible to master one aspect of combat without necessarily excelling in another. Similarly - but worth mentioning - this also means that a spec can be overpowered in one context and not in a different context. Contextual keys to consider are your own spec, your comp, your enemies comp, and the bracket (ie. 3v3) - adjusting any of these can adjust balance of power of the spec itself: for which we often use the terms synergy, and counter-comp.

    In a 1v1 fight between a Subtlety Rogue and a Blood DK, it's probably safe to say a Blood DK is overpowered - but a Blood DK isn't necessarily overpowered in a DPS role in 3v3, or when fighting a Frost Mage in 1v1. Overpowered (and by extension, underpowered) can only exist within a very specific context; Blood DK's aren't necessarily overpowered in 1v1, but are overpowered when fighting melee specs or pets specs who are likely to feed them Death Strikes (or ranged who can't kite). Mistweaver Monks are generally pretty strong in 1v1, 2v2, and 10v10 - but are terrible in 3v3 and to a lesser extent 5v5. Resto Druids in 5.2 had massive synergy with both Frost Mages and Shadowpriests - so they weren't "overpowered" in 3's, but were overpowered in Shatreeplay in 3's.

    We could assign weights for how much each of these factors matters and then create an algorithm for representation - but the significance of the weights themselves seems to shift - so the formula would probably have to be constantly amended. Though... now I'm interested to try to work backwards from the current stats to see what are the current values of those weights...
    Last edited by Yvaelle; 2013-09-24 at 03:06 AM.
    Youtube ~ Yvaelle ~ Twitter

  20. #20
    Off topic but I'd happily see the overall representation of warriors to half of what it is if it can finally have a second 2.2k viable spec like literally every single other class in the game most of which actually have multiple rank 1 viable specs.
    Quote Originally Posted by High Overlord Saurfang
    "I am he who watches they. I am the fist of retribution. That which does quell the recalcitrant. Dare you defy the Warchief? Dare you face my merciless judgement?"
    i7-860 @2.8GHz | Radeon HD 7770 | 8GB DDR3-1333MHz | Corsair CX 430W |

Posting Permissions

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