63 replies to this topic

[Synkz]

[quote name='brosearch' timestamp='1361906567' post='3853815']
what would these results show if 5v5 titles were not included?
[/quote]

Fistao said they only show 3v3

[Fistao]

I didn't add v5 titles because, according to Starcookies calculator, there are no Gladiators in most realms.

So I don't have tools to check teams eligibility.

[dionim]

thats what i see when queuing 2400 heh

[jonadefty_3724427]

Quite interesting results although not really surprising with the gross amount of warriors in 3v3. I'm actually quite surprised that spriests didn't dominate the dps results a little bit more.

As for the debate of basing the results on class population, the huge swings in representation don't just take place at the very top of the ladder. I'd definitely say that sub 2.2k players are more influenced by class balance than gladiators/r1 players as it can clearly be seen when classes go through phases of being ridiculous. High profile streamers and tournaments (online and Blizzard run) attract average players to new classes because they want to be able to push more and more rating and "fotm" rerolling seems like an easy solution.

The switch around of glad representation also shows this happening at a higher level. Think about rogue and lock glad representation last season compared to this one. Think about how low the amount of warriors were glad in s11 and how they have taken a comfortable top spot now. There's always going to be dominant classes in this game and players across all rating brackets will reroll them and the representation graphs will shuffle around.

[Saikx]

that warrior %, quoting justin bieber from rage thread:

Snuggli, on 26 February 2013 - 06:31 PM, said:

Made this, started out as KFC but ended with just warrior.

[Braindance]

Somewhat outdated but general guidelines still apply

[Treetrunksxoxo]

That's a good mage guide.

[Radejjj]

I think that these types of statistics are really deceptive. When you see how bad rogue/monk/dk/warlock representation is, you think that means those classes/specs are terrible, but that isn't necessarily true. Warriors/hunters/spriests are severely over represented, if they get nerfed, everyone elses numbers go up, but what always happens, is the op class gets nerfed, and the underepresented class gets a buff which just completely swaps everything. 5.2 is going to be rogue/dk/monk/disc priest.

As good as hunters are they also only have one viable comp, KFC. Ferals are tied with spriests as being the stupidest spec right now and there are really not even that many of them, for whatever reason. Not to mention classes/specs with shit representation can be overpowered as hell. Fire mages are still stupider than fuck all.

And hybrid classes tend to look weak in things like this. Boomkin and enhancement shaman are both really strong, but resto shaman and feral druids are better, so people who can play those classes at a really high level gravitate towards the stronger spec. The same goes for disc priests, and to a lesser extent ret paladins, though I think ret needs more help than the former class/specs. Monks also really are not that bad but because of the stupid gearing system it became impossible for anyone to level/gear one late in the season.

tldr: numbers can be deceiving, math is racist.

Edited by Radejjj, 26 February 2013 - 10:20 PM.

[Hackattack3]

Braindance, on 26 February 2013 - 05:08 PM, said:

Because it doesn't take the total class population into account. Only by normalizing over the total class population can you get some hints on what is "stronger" or not.

You have for example 39 hunters getting glad in EU. If 40 hunters play the game you can say they are overpowered. If 100000 hunters play the game, even if the percentage you report is 13.26%, the percentage of hunters getting glad is 3.9% which would mean hunters are actually weak.

tl;dr: no normalization=no relative class strength conclusions

It's actually very easy to pull up the general class population statistics to gut check these results:

source:  http://www.worldofwa...tition-0-0.html

Last Database Update : 25 Feb 2013
- Level 90 players - 522,868 guilds and 32,609,663 players in the database
Class % (Global) Paladin 11.7% Hunter 11.3% Druid 10.6% Death Knight 10.4% Warrior 9.8% Priest 9.7% Mage 9.2% Shaman 8.2% Warlock 7.4% Rogue 6% Monk 5.6%

It's clear to see that there is less than a 3% global representation difference between pallys, hunters, druids, dks, warriors, and priests.  The two classes that are the most under-played are rogues and monks and GC has addressed this discrepancy in the last few weeks.  He said people play rogues less b/c there is less pvp and "stealth ganking" and the monk buffs are to get people to actually PLAY them.

There is also a HUGE difference in leveling characters.  Hunter has historically been the easiest class to level and the most casual friendly.  Try to get a friend/co-worker/girlfriend into wow and they tend to be drawn to easy to level classes.  Many casuals, from my experience, favor hunters due to smooth leveling experience.  This is another factor which affects general population.


TL;DR - The discrepancy is MARGINAL.  If you want to draw any conclusions you could 2xmultiplier for rogue glads b/c low population class and normalizing shammy to pally population would show shammys are still the best healers.

[Hackattack3]

Radejjj, on 26 February 2013 - 10:11 PM, said:

tldr: numbers can be deceiving, math is racist.

But it is also important not to throw out the baby with the bathwater.  Don't take the data as the end all be all but understand that there is some variability that has not been accounted for and look for quantities of scale.

All the marginal things people brought up (i.e. general population) cannot hide the fact that there will be 3 times as many warrior gladiators as any other dps class (excluding spriests and hunters).

Edited by Hackattack3, 26 February 2013 - 10:27 PM.

[Radejjj]

Hackattack3, on 26 February 2013 - 10:25 PM, said:

But it is also important not to throw out the baby with the bathwater.  Don't take the data as the end all be all but understand that there is some variability that has not been accounted for and look for quantities of scale.

All the marginal things people brought up (i.e. general population) cannot hide the fact that there will be 3 times as many warrior gladiators as any other dps class (excluding spriests and hunters).

Absolutely, I'm not really denying warriors/hunters/spriests being too strong as much as the other classes being too weak. CLEARLY those specs/classes are over represented. Because of how absurdly overrepresented those specs are, everyone else looks weak. If the op shit is nerfed, and everyone else left the same, the numbers would look a lot better without having to buff rogues/monks/dks/priests

Edited by Radejjj, 26 February 2013 - 10:31 PM.

[Djandawg]

Radejjj, on 26 February 2013 - 10:11 PM, said:

I think that these types of statistics are really deceptive. When you see how bad rogue/monk/dk/warlock representation is, you think that means those classes/specs are terrible, but that isn't necessarily true. Warriors/hunters/spriests are severely over represented, if they get nerfed, everyone elses numbers go up, but what always happens, is the op class gets nerfed, and the underepresented class gets a buff which just completely swaps everything. 5.2 is going to be rogue/dk/monk/disc priest.

As good as hunters are they also only have one viable comp, KFC. Ferals are tied with spriests as being the stupidest spec right now and there are really not even that many of them, for whatever reason. Not to mention classes/specs with shit representation can be overpowered as hell. Fire mages are still stupider than fuck all.

And hybrid classes tend to look weak in things like this. Boomkin and enhancement shaman are both really strong, but resto shaman and feral druids are better, so people who can play those classes at a really high level gravitate towards the stronger spec. The same goes for disc priests, and to a lesser extent ret paladins, though I think ret needs more help than the former class/specs. Monks also really are not that bad but because of the stupid gearing system it became impossible for anyone to level/gear one late in the season.

tldr: numbers can be deceiving, math is racist.

You are one kind of unique spastic aren't you?. You posted this shit in lock forums, priest forums, now here.
-Numbers are not deceptive because they represent this season. No one gives a shit about nerfing classes x,y,z in 5.2. It is about S12 and it is almost finished.
-A resto shaman/holy paladin representation are each 6 times higher than disc+monk representation.
-Warrior representation is more than every single melee combined.
-How do you think people feel when they had to skip an entire arena season because their class can't do arena? Are they like, oh well as Radejj pointed out in 3 different sections of AJ irrelevantly , we couldn't do  arena this season but at least we have the huge relief of knowing the fact (that again was spotted by Radejj) that our class is actually not bad, others are too strong.

It doesn't matter how things go after nerfs/buffs or if  a class with no representation is not bad in essence even though it gets raped with its not so bad tools in arena by amazing classes and this is what those S12 charts are about, the overall result.

-Also you coincidentally forgot to mention holy paladins, probably cause you are scared witless about shitty classes getting buffed. Tldr of all your posts is: Nerf overpowered, don't buff underpowered but don't touch my class?

Edited by Djandawg, 27 February 2013 - 12:16 AM.

[Hackattack3]

Radejjj, on 26 February 2013 - 10:30 PM, said:

Absolutely, I'm not really denying warriors/hunters/spriests being too strong as much as the other classes being too weak. CLEARLY those specs/classes are over represented. Because of how absurdly overrepresented those specs are, everyone else looks weak. If the op shit is nerfed, and everyone else left the same, the numbers would look a lot better without having to buff rogues/monks/dks/priests

I would agree with you if this was Cata, that was pretty close to balance.  If they got pve gear, triple dps, laddder/mmr bugs, and RLS under control that would have been the best pvp expansion.  The problem now is the game is soo fucked up and certain classes depend on broken mechanics that it is really difficult to balance without rebuilding the expansion.  Example:  Where do you even begin to re-design warlocks?  They revolve around a broken mechanic (blood fear) but they still don't have great representation.

[Kushi]

GO GO RETS! Oh wait...

[Radejjj]

Hackattack3, on 26 February 2013 - 11:33 PM, said:

I would agree with you if this was Cata, that was pretty close to balance.  If they got pve gear, triple dps, laddder/mmr bugs, and RLS under control that would have been the best pvp expansion.  The problem now is the game is soo fucked up and certain classes depend on broken mechanics that it is really difficult to balance without rebuilding the expansion.  Example:  Where do you even begin to re-design warlocks?  They revolve around a broken mechanic (blood fear) but they still don't have great representation.

Warlocks are not that bad, warriors/bm hunters/tremor totem just shit all over them. If you fix the op shit, warlocks/rogues/monks/discs/dks would all naturally go up in representation when warriors/hunters/spriests dont make up 60% by themselves.

[Nightmonkey]

I think everyone missed something here.  Take another look at the graphs.  The first thing that jumped out to me was just how low melee representation is outside of Warriors.  Of course Warrior representation is obnoxiously high too.

My theory is that Warriors have only been able to dominate because of their exceptional defensive abilities.  Other melee classes have higher utility or damage, but nothing else can really match Warrior defensive abilities and mobility.

Seeing as how Warrior defenses are being nerfed and other melee classes are receiving buffs, do you guys think we will see fewer Warriors and more of the other melee classes next season?  Or will we just see no melee and only ranged classes?

[Dakkrothy]

Representation can mean a lot but it can also mean jack shit


Lets put it this way.

Priests in season 5 could almost only play RMP to be competitive, and when I say competitive I mean it could beat anything and get rank 1 anywhere. Why was that?

1. Did priests get absolutely carried by rogue & mage?
2. If well played, was synergy enough to make the comp viable?

It was obviously 2, if it wasnt, you'd take any of the other healers that were stronger in general.


Personally I think it's more important that you focus on a classes strength in comps which they're usually viable in rather than the class itself; WE DON'T WANT DUELS TO BE BALANCED. If you look at the wrong things you could very easily end up overbuffing them or overnerfing them (depending on how powerful/weak they are).



Another thing to look at would probably be what x class is struggling against if its a weak class. Lets take an example from the current season.

Rogues, very weak because of warrior strengths. Does it mean you have to nerf warriors, buff rogues or both? It depends how big the leap is. Once you've nerfed warriors in this case you have to think what rogues would be like WITH the warriors nerfs before you buff rogues. If they remain weak you should start buffing them slightly til you find a decent balance point.


It's extremely hard balancing wow with the amount of abilities, classes & constant changes there is in the game. It's impossible to say the least. I just wish blizzard actually hired really high rated, unbiased, PvPers. Mainly because it HAS to be balanced around the very top ratings if you want a game as balanced as it could be. This would probably end up pissing a lot of casual players, but then again it's just a l2p issue in this case, which would be way more acceptable in the general scene of WoW I think.

I recommend you to read the last article posted in http://www.veevsvaul...ll-cap-dilemma/ to know what I am talking about. but if you CBA I'll just do a TL;DR - A class extreamly strong around 1900 ratings doesnt mean it's nearly as strong at 2.5k ratings



peace!

Edited by Dakkrothy, 27 February 2013 - 02:51 AM.

[fant0m8]

A better way of "normalizing" the data would be to find out the percentage of classes on PVP teams, perhaps PVP teams over 1500. And then compare that to glad representation.

Just looking at these percentages is a bit deceiving because the total number of people in the population is ridiculously small. 400 per region? Wow. There used to be more gladiators than that in one battlegroup in one bracket.

This isn't much better than looking at the rank 1 population of 5v5 teams and drawing conclusions about class balance.


(I'm not saying that balance is good, just that these stats aren't good evidence one way or the other.)

[Nightmonkey]

This data tells a pretty compelling story.  It's not the whole story of arena this season, but it is a nice little snapshot of the top 3v3 teams right now.  

Do you guys feel like something isn't accurate here?  KFC dominating, Warriors everywhere, Paladin and Shaman most common healers, Shadow Priests also super strong with a decent showing from Mages.  Everything else seems kind of low representation, especially melee other than Warriors.

Seems like an accurate description of S12 to me.

[Vanguards]

I thought someone said Rets were doing better than Rogues this season in terms of representation. Maybe it wasn't meant at Gladiator level?