Detailed Statistical Analysis of Warzones - Classes are balanced!

[starwaysdude]

Lumping advanced classes like Trooper Commando and Trooper Vanguard together makes your data useless.

 

I don't think so, I have valued damage, protection and healing equally.

[MorbidAri]

Yes, of course... 200k + healing in a WZ probably made no difference at all...

 

He is talking about bubbles. The bubble might not even take a point of damage but still score full amount to their total healing. So yes your data will be skewed because of this fact.

[starwaysdude]

Feel free to test, but I am fairly sure that it only posts healing numbers if it... heals....

[MorbidAri]

Feel free to test, but I am fairly sure that it only posts healing numbers if it... heals....

 

I can confirm having played a Sage leveling that bubbling myself and doing almost nothing else was racking up the healing. And no I wasn't healing myself as i was getting jumped. 3 Imps vs a 16 doesn't give an awful lot of time to heal.

 

Still getting healing medals even when I was barely ever in the battle.

[BobaFurz]

Charts dont mean much...you need single target infos and not from battleground charts.

 

Some skills are AOE, some people spread their dmg, other hit only single targets....

[Cinawyn]

OK, so with all the "OP/nerf" threads here, I decided to shine the cold light of data on this.

 

I ran a few warzones and looked at the results. I did the following

 

I took the highest 5 people based on damage. The reason for this was that I am trying to find the lvl 50 people and this set is the most likely distinguishing criteria.

 

I added the damage+healing+protection

 

Some BG's you end up doing more damage than others. So, I normalized each BG to the one that had done most damage. The highest D+H+P amount was ~377k, so I used that to normalize EACH warzone.

 

For anyone that had zero deaths, I gave them a death of 1 so the kill/death ration would not be infinite.

 

The results are (class/total effect/ KD ratio)

Bounty/Trooper 254,371 10

Imp Agent/Smuggler 291,260 26

Inquisitor/Counselor 270,709 30

Warrior/Knight 253,569 11

 

 

I am actually quite amazed how balanced that is between classes......

 

 

Warzone results

http://www.pdfhost.net/index.php?Action=Download&File=6a43983a702225ce3a52ba3c0863e1bb

 

Initial Warzone normalization

http://www.pdfhost.net/index.php?Action=Download&File=f13e1fa5d6aeb3581a4230fa564b159f

 

Final Class grouped data

http://www.pdfhost.net/index.php?Action=Download&File=ecfb4dffa9dce7f403082a10adb72f70

 

 

A better analysis would have been to have analysed all the classes in the warzone, but I didn't have the stamina for that.

 

All that working out and its all for nothing. I challange you to go 1 v 1 using a seer with any other class with the same level and gear equivelent. You cant do enough damage and you cant outheal the dameage. The seer has to be static to do any moderate damage or moderate healing. dots that do damage do barely any damage and heals that are dots do barley any healing.

Edited January 15, 2012 by Cinawyn

[Jabbb]

This is idiotic. Yeah if I wanna spam my long range dot on 10 people I can do 400k damage, will I have accomplished anything? Not really. Total damage is meaningless in terms of balance. The fact of the matter is, 2 Operatives will kill 2 of another class 100% of the time with ease and slip back into the shadows. Operatives need a nerf. Edited January 15, 2012 by Jabbb

[Hosette]

Yes, of course... 200k + healing in a WZ probably made no difference at all...

 

Pro Tip

 

Lifetap and heal yourself for medals as sorc/sage

[Subcool]

Operatives must feel a lot like WoW's rogues starting off. I remember the days of 1 shotting kids in my same bracket with ease. No potion or gear stacking. Back then everyone cried nerf but it never came because at 60 rogues weren't 1 shotting people or dominating every class.

 

Could a rogue take any class with all cooldowns ready? Yes, a good rogue should be able to. This game isn't about 1v1s. Neither was wow.

 

All these examples of getting killed by operatives 1v1 are simply ridiculous. They hold as much worth as simply saying "I fought an operative and beat them so obviously they are fine." And here is someone attempting to collect data for the good of the community yet people are trolling or calling it useless.

 

I'd take SOME data over NONE any day. Bioware isn't balancing the game around this guys data, so relax on the preciseness. All he did was quickly sample some warzones to find any sore thumbs.

 

Results? None really to be found. No one class archetype is dominating the playing field.

 

Operatives are meant to kill you well, get over it.

[starwaysdude]

Operatives must feel a lot like WoW's rogues starting off. I remember the days of 1 shotting kids in my same bracket with ease. No potion or gear stacking. Back then everyone cried nerf but it never came because at 60 rogues weren't 1 shotting people or dominating every class.

 

Could a rogue take any class with all cooldowns ready? Yes, a good rogue should be able to. This game isn't about 1v1s. Neither was wow.

 

All these examples of getting killed by operatives 1v1 are simply ridiculous. They hold as much worth as simply saying "I fought an operative and beat them so obviously they are fine." And here is someone attempting to collect data for the good of the community yet people are trolling or calling it useless.

 

I'd take SOME data over NONE any day. Bioware isn't balancing the game around this guys data, so relax on the preciseness. All he did was quickly sample some warzones to find any sore thumbs.

 

Results? None really to be found. No one class archetype is dominating the playing field.

 

Operatives are meant to kill you well, get over it.

 

Best post in this thread, thank you for it....

[sildoor]

sorry .. nice try but you can't say that classes balance when you dont take alook at the advanced classes ..

 

so sage+shadow are balanced with

sniper + op .. well maybe op's are 20% better then snipers, u can't see that in your samples..

 

or that a dps sage does more damage then dps shadows thx to aoe ..

 

etc etc etc

 

if there was a way to see the advanced classes you would perhaps see higher differences in performance,

[ltprudii]

sorry .. nice try but you can't say that classes balance when you dont take alook at the advanced classes ..

 

so sage+shadow are balanced with

sniper + op .. well maybe op's are 20% better then snipers, u can't see that in your samples..

 

or that a dps sage does more damage then dps shadows thx to aoe ..

 

etc etc etc

 

if there was a way to see the advanced classes you would perhaps see higher differences in performance,

 

Not looknig at spec or gear. Its hardly highly detailed. Or even has any analysis. Sure you can post numbers. BUT WHAT DOES IT MEAN!? lol

[Subcool]

Not looknig at spec or gear. Its hardly highly detailed. Or even has any analysis. Sure you can post numbers. BUT WHAT DOES IT MEAN!? lol

 

It means that out of the 4 class archetypes, one is not significantly ahead of another. If one archetype was doing far more damage (on average of spec etc...) it would be represented here.

 

And to all these people saying "well this is useless data cause you didn't account for this spec playing like this" etc...

 

Do you guys have any idea what averages mean and how they are useful in understanding data?

 

 

Some people probably are confused by this, but that's not the OP's fault if you fail how to see this is even a tiny bit helpful in relating to a game that, so far, none of us have "run the numbers on".

[EternalFinality]

Operatives must feel a lot like WoW's rogues starting off. I remember the days of 1 shotting kids in my same bracket with ease. No potion or gear stacking. Back then everyone cried nerf but it never came because at 60 rogues weren't 1 shotting people or dominating every class.

 

Could a rogue take any class with all cooldowns ready? Yes, a good rogue should be able to. This game isn't about 1v1s. Neither was wow.

 

All these examples of getting killed by operatives 1v1 are simply ridiculous. They hold as much worth as simply saying "I fought an operative and beat them so obviously they are fine." And here is someone attempting to collect data for the good of the community yet people are trolling or calling it useless.

 

I'd take SOME data over NONE any day. Bioware isn't balancing the game around this guys data, so relax on the preciseness. All he did was quickly sample some warzones to find any sore thumbs.

 

Results? None really to be found. No one class archetype is dominating the playing field.

 

Operatives are meant to kill you well, get over it.

 

That's not true. World of Roguecraft pretty much single-handedly got Rogues nerfed. Any class that ALWAYS wins 1v1 is not balanced by definition.

[Zarthorn]

Go read all the screens that Operatives post in the forums defending that they are OP. Notice an average total of 600-800k (damage and healing together).

 

Funny that's almost double your numbers. They post multiple screens. This it isn't a fluke game. It's over the series of matches.

 

So not only is your test too small, but you weren't seeing any of the serious problem classes/glitches/exploits. So that means you using the small test as "fact" is as anecdotal as everyone else's anecdotes.

 

Now remove about 200k or so probably more if they die a few times and you have your none stacking actual figure.

 

400k for a dps class is about right at endgame with full gear and not abusing the broken stacking issue and that ould likely be for a game where they played extremely well.

[voidsurvivor]

Cybertech grenades, especially when stacked with additional power-ups and other temporary buffs, were creating combat damage burst potential higher than we were comfortable with and caused player versus player engagements in the game to resolve faster than intended. In fact we have additional changes to classes and consumables currently in internal testing to deal with faster than anticipated kill times in player versus player combat in a future patch.

 

Since no one seems to read dev posts...

[fungihoujo]

sample size is too small -check

combining 2 of each of the advanced classes into the same category -check.

 

How many of those IA were snipers? What was the sniper average compared to operative? These were all collected by the same person. Meaning he/she had impact on the variables?

 

As opposed to the normal way of doing things- someone posts anecdotal 'evidence', some incorrect 'facts' 'proving' one side is better than the other, and then calls for a nerf based on it and everyone jumps on the bandwagon with their own 'proof'.

[Lazyllama]

Im certainly not going to flame you for posting what you believe to be a relevant study, but you seem to be ignoring the context of what posters are saying because of the way they are saying it. It is ironic really because by posting the numerical data you believe it is critical unbiased data, but when you are critiqued on it you become defensive and biased toward its relevance.

 

Look man it is nice that you are trying to put some numbers and data behind the study you are trying to achieve here, but straight up it is completely insignificant and irrelevant the way you ran the experiment. Your running chiropractic level science here where you are just grabbing numbers and what you believe to be relevant values and comparing them. Most of the posters in here are absolutely right, showing total numbers from a tiny sample or unspecified classes achieves nothing. And no I am not going to run my own numbers, but that still does not change the fact that the data and analysis that you did means nothing. Just because you wrote numbers down doesn't mean they are accurate or mean anything. You have to identify the relevance and create an experiment that impartially tries to identify the complaints among the community.

 

Im sorry man but the data you NEED to gather, you wont be able to because of the nature of this game, its lack of logging the data and barring any third party software. Your only left with irrelevant data that honestly means nothing outside of stroking the egos of players within the warzone.

[starwaysdude]

It means that out of the 4 class archetypes, one is not significantly ahead of another. If one archetype was doing far more damage (on average of spec etc...) it would be represented here.

 

And to all these people saying "well this is useless data cause you didn't account for this spec playing like this" etc...

 

Do you guys have any idea what averages mean and how they are useful in understanding data?

 

 

Some people probably are confused by this, but that's not the OP's fault if you fail how to see this is even a tiny bit helpful in relating to a game that, so far, none of us have "run the numbers on".

 

The most useful post in this thread

[Quesa]

The glaring problem I see with this is that the classes work in many different ways.

 

For example.

An Operative is an opportunistic killer which spends most of it's time in stealth waiting for the right opportunity. So an Op will unstealth, kill and restealth thus laying out all his damage out in a short period of time on a single target.

 

Then you have classes like a Sorcerer who could essentially tab-dot for low but consistent damage with periods of higher overall DPS through active casting. They also have a powerful AoE that can hit multiple targets.

 

The main issue isn't the overall damage, DPS, survivability or escape mechanics a class has, it's the upfront burst potential of the Operative/Scoundrel which is linked to a hard CC, all coming from a nearly undetectable stealth (which can be buffed +15 with short CD) setting up a situation where a target can be killed in under 8-10 seconds while that target is hard CC'd for nearly half of that duration.

 

So while the data may be correct, the differences in class mechanics and how that classes damage is delivered and HOW MANY TARGETS it's delivered to is something not represented in your data makes it misleading, at best.

 

Killing blows means nothing as I've used corrosive dart on a target, ran away and one of the ticks killed it. Kills doesn't mean much as the mere fact of participating in the death of that person could be anything from CC'ing that target at one point or another to healing someone who killed or participated in the death of that target.

Edited January 16, 2012 by Quesa

[starwaysdude]

Im certainly not going to flame you for posting what you believe to be a relevant study, but you seem to be ignoring the context of what posters are saying because of the way they are saying it.

 

Well.....

 

As opposed to the normal way of doing things- someone posts anecdotal 'evidence', some incorrect 'facts' 'proving' one side is better than the other, and then calls for a nerf based on it and everyone jumps on the bandwagon with their own 'proof'.

 

^ this

[starwaysdude]

The main issue isn't the overall damage, DPS, survivability or escape mechanics a class has, it's the upfront burst potential of the Operative/Scoundrel which is linked to a hard CC, all coming from a nearly undetectable stealth (which can be buffed +15 with short CD) setting up a situation where a target can be killed in under 8-10 seconds while that target is hard CC'd for nearly half of that duration.

 

Actually, its not at all.

 

Burst 1v1 is very frustrating for the person that gets bursted upon.

 

For winning pvp warzones, being able to post 400k + combined damage and healing is what will win in a team situation.....

[Quesa]

Actually, its not at all.

 

Burst 1v1 is very frustrating for the person that gets bursted upon.

 

For winning pvp warzones, being able to post 400k + combined damage and healing is what will win in a team situation.....

 

No, numbers in general mean nothing because there is no context. Just because you put up 400k damage AND healing doesn't mean much in an object-oriented warzone. It's disingenuous to suggest that the numbers portrayed at the end of a match means alot as I've personally seen very low numbers from everyone in a 6:0 game, as well as other MMO wargames.

 

Furthermore, someone like a Merc or Sorc can sit up on a perch and spam AoE damage spells with no quantifiable end effect. That person may put up some impressive numbers but did he do anything to stop the ball carrier? ... did he score a TD? ... did he save a ball-carrier from death thus assisting in a TD? We don't know, all we have is numbers and those numbers have no context.

 

The original post is taking those numbers, averaging them out and using them as evidence of class power/balance. It doesn't mean anything because of how the classes differ in their damage delivery, survivability and player play-style.

 

I can put this another way.

A Powertech can have twice the damage, same kills as an Operative simply because his damage delivery is alot slower thus allowing the target to heal or better utilize skills that slowly regenerate health and prolong the fight.

 

An operative can have low overall damage but same kills because he kills his target in under 10 seconds not allowing much healing or slow regeneration abilities to offset the damage taken.

Edited January 16, 2012 by Quesa

[Dosvidaniya]

Actually, its not at all.

 

Burst 1v1 is very frustrating for the person that gets bursted upon.

 

For winning pvp warzones, being able to post 400k + combined damage and healing is what will win in a team situation.....

 

1. The OP looked at damage, not healing.

2. That depends entirely on where the damage and healing is going. I had a hutball game earlier (we won) where I was stuck with a number of level 20s (and one other 50). There were about 6 level 50s on the other team. At least one was a full healing spec. My little team struggled so very hard. All they really did was stop as many goals as they could. I managed to get two goals. The other team only managed one. We were no where near the same damage and healing. But, my team actually focused their numbers where it counted.

[Aega]

I'll bite. IF these show a true, though very broad, average then it confirms my suspicions. Tanking/high defense classes are doing far too well. What those numbers are not showing is how much of that damage each class can soak up.

 

As a merc, I can unload all the supposed OP dmg people apparently think we have, but get 1 melee on me and its stops very quickly. My dmg gets soaked up and I get controlled and mitigated.

 

I don't pretend to know how to address that. Obviously just calling for nerfs only catches flak. My main option is probably simply to reroll the class I think is far too potent and see for myself. Maybe if this becomse a trend, that will speak for itself.