do they test this game?

[chrion]

I mean from all perspectives, before season 6 started did they try every single combination of class's vs all other class's like they obviously should be? And if they dont WHY THE HELL NOT! If they for example tested arcane mage/rogue vs DK/tree druid, they would be like wait wait this is wrong, not balanced at all, and then they could make simple adjustments. If its this simple (and it is) a child could figure out that testing before release is probably a smart thing to do. Im sure you all could name plenty of other combinations of x/x vs y/y that just couldnt have possibly been tested before they let that garbage get into our hands. This is really the core problem, If a shaman/ret faces a shaman/hunter and the ret team can win the match with ease in comparison to the RNG/skill/luck needed by the hunter team, then make slight adjustments until it feels right, then if that same shaman/ret team has an easy time against a rogue/priest, where the shaman/hunter team has a hard time, then make slight changes to CD's and dmg healing #'s that bring them in line with eachother, then go back and test the ret vs the hunter team again, then move onto a diff comp and blah blah blah, you get the point. If they wanted to spend the time to balance the game they could, they certainly spend the time to make damn sure starcraft is as perfectly balanced as possible, and they certainly can do it with WOW if they made it there goal.

[Narit]

you should make your own game and do that then!

[chrion]

nice generic canned responce...when anywhere did I say I could do it the actual testing better? Im starting a convo about blizzards testing process, and thats the only thing you could respond with? Im sure you could be more interesting than that....atleast I would hope...

[Degrador]

I shouldn't even bother, but for some reason I am...

If they for example tested arcane mage/rogue vs DK/tree druid, they would be like wait wait this is wrong, not balanced at all

Blizzard has stated before it's too difficult and they're not going to try to balance for 2v2. So in all likelihood, given that position, they're not going to test 2v2 combinations.

and then they could make simple adjustments

Oh really? 'Simple adjustments' huh? Riiight... It doesn't take a genious to realise that even the smallest of changes in this game can have wide implications, especially given the largely different aspects of the game that such changes generally apply to (eg, soloing, 5 mans, raiding, 3v3, 5v5, BGs).

If its this simple (and it is)

No it's not.

a child could figure out that testing before release is probably a smart thing to do.

Do you honestly think Blizzard could get into the position they have without the slightest clue about testing? Pretty stupid thing to be stating really...

If a shaman/ret faces a shaman/hunter and the ret team can win the match with ease in comparison to the RNG/skill/luck needed by the hunter team, then make slight adjustments until it feels right, then if that same shaman/ret team has an easy time against a rogue/priest, where the shaman/hunter team has a hard time, then make slight changes to CD's and dmg healing #'s that bring them in line with eachother, then go back and test the ret vs the hunter team again, then move onto a diff comp and blah blah blah, you get the point.

Umm, do you realise the number of combinations possible for 2v2? Assuming 4 distinct classes, there are over 500 class combinations to test. Now consider that there are 3 different talent trees for each class, often with 5+ actual viable PvP specs per class, you're now looking at over 500,000 combinations. From there, if you wanted to do a thorough test, you'd have to have players who are experienced at the classes they're using, and capable of understanding the nuances of the various specs they're using. Assuming you've got those, you'd want to have at least 20, more likely 100+ games of each 2v2 combination to get accurate results (even then, 100 games is still a relatively small data set to base a conclusion on). Hence we are now looking at millions of 2v2 games needing to be played. From there, you would then need to make these so called 'slight changes' and retest everything again.

Now, as I said before, Blizzard has stated that they're not going to balance around 2v2 as it's just never going to work. If you want to go through the math on doing thorough testing for 3v3 as well as 5v5, you'd find the amount of tests required is just an impossible amount.

If they wanted to spend the time to balance the game they could, they certainly spend the time to make damn sure starcraft is as perfectly balanced as possible, and they certainly can do it with WOW if they made it there goal.

3 races, no talent points, only 1v1 fights. This means there's only 3 possible combinations to deal with. Doing 20-100 tests of each of these combinations per build is actually possible, doing it for WoW is not.

[Sartecho]

You'd think they could've at least figured out that resto druids would be absurdly broken.

[Tremelizzer]

If they didn't test, it would be MUCH MUCH worse than now (and currently it isn't bad even). So, continue crying.

[Mysmere]

you dont have to balance 2v2 if you balance 1v1.

[Sigmundus]

I remember seeing a Blue post during tbc that said Blizz balance arena around 5v5, not 2v2 or 3v3. So complaining about 2v2 and 1v1 isn't going to get you anywhere, just like it didn't before.

[Ravasha]

Public Test Realm maybe?

[gehennaspawn]

I'll consider reading your post next time, if you use paragraphs.

[Kramar]

TL;DR: no.

Wall of Text:

so lets see, 10 different classes, lets say you would like to have each tree a viable PvP spec, that's 30 specs, now we have to take 2 of those specs, and it can be the same spec twice which gives us:

30!/(2(30-2)!)= 435 comps for 2v2.

then how many line-ups can we get? seeing as it's 2 teams and yet again we can have the same comp twice, there are:

435!/(2(435-2)!) = 94,395 line-ups.

you'd want to test these line-ups for about 100 times each to get a good dataset, and if you want an even better dataset, you'd have multiple skilled people play a certain line-up so it's not just one team's skill that affects the outcome.

94,395 line-ups * 100 tests * 4 teams to test each line-up = 37,758,000 2v2 matches to get a good dataset.

each 'minor' adjustment would have to be tested like this again, you can see where this is going I hope.

so, with the same numbers for 3v3 and 5v5 lets see:

3v3: (30!/(3!(30-3)!) = 4,060 4,060!/(2(4,060-2)!) = 8,239,770 8,239,770 * 100 * 4 = 3,295,908,000 tests each change

5v5: (30!/(5!(30-5)!) = 142,506 142,506!/(2(142,506-2)!) = 1,015390877*10^10 1,015390877*10^10 * 100 * 4 = 4,061563506*10^12 tests each change


4,061563506*10^12 = 4,061,563,506,000 which is quite a large number, so no, blizzard will not test each comp and each line-up till the death to get an idea of how balanced things are, doing so would require more time than wow is old.

Yes, they test, but not by just pairing up all different comps and seeing what happens, they just get the played games data and draw conclusions from that, and any changes they make will have a ??? effect on the game since there might allways be someone out there that finds another way to use/counter a change.

[Deleted]

In the ideal world they would manage to balance all setps against every other setup. However in practice its virtually impossible. Its why blizzard has a "close enough" attitude. Its the only realistic way of obtaining anything halfway to balanced. I test games for a living. Its my job. Yes I'm lucky but until you actually do it yourself you have no concept of the amount of work involved with balancing a game like WoW.

PVE balance is the easiest thing to obtain. There are far fewer factors and variables. That being said its not easy to achieve. While there are fewer variables to control they are still difficult to control. A great example is a feral druid. A well geared, well played feral druid can top the DPS meters. Hoiwever feral druids have probably the toughest prioity list to control and manage of any class going. As a result there is considerable variation between the performance of 1 feral druid and another.

Now lets consider PVP. Think about it. You have 10 classes. Now lets assume in the ideal world each class has 1 PVP spec. I'm well aware most classes have 2 or maybe even 3 possible specs they could use but for simplicity lets keep it at 1. A bit of maths tells us its a possible 100 different build combinations they would need to balance and thats just in 2v2. It just goes into the stratosphere when lookingat 3v3 and 5v5. That is a simply HUGE amount of work. In fact I would say its practically impossible.

Lets say we look at your first example of Arcane Mage + Rogue (Team X) against a DK + Resto druid (Team Y). You could test that combination until it is perfectly balanced. Now lets introduce another team setup say Arms Warrior / Holy Pally (Team Z). When Team X faces Team Z you see that team X is completely dominating (just an example I have no real idea if they would or not). So you nerf Team X and buff Team Z a bit. Now you notice on further testing Team X is now being owned by Team Y due to the nerfs. You have just broken the perfect balance you worked so hard for in the first place between Team X and Team Y.

In short WoW is far too complicated for perfect balance. Its just not realistic. Blizzards approach of "close enough" is the only realistic method of balance.

EDIT: I am crap at maths so my numbers are probably off but it gets the point across.

[Mandible]

Public Test Realm maybe?

People don´t "test" things on the ptr - they go there to see the stuff before it goes live. I highly doubt most people who does arena has an extra character lying around or even consider using the possibly premade characters available to try out odd combinations for arena teams.

On a sidenote - World of Warcraft wasn´t made to handle pvp, it was added in the middle.

[Hrethel]

I do love seeing crazy complaints blown clean out of the water by the application of a little bit of solid maths.

/applaud

Honestly, Kramar's maths should be stickied so that any time people bitch about Arena comps being untested can be pointed at it and asked how they would set about testing those 94,395 line-ups.

[Arganaut]

WoW's success has been built upon an incredible PVE environment. PVP was there and something to do but it was Arena that changed everything.

When you consider the vast numbers involved and possibility of spec's/combo's I personally think Blizzard are doing a pretty damm good job at balancing PVP.

It is bloody difficult to reverse engineer something on this scale.

[Deleted]

Another complaint-thread made by Chrion.
Are you trying to make complaining your trademark, buddy?

You're lucky I'm not a moderator here, as I would kindly ask you to take your continuous complaining somewhere else.

Here are some examples of your whining. Please note that these are only the topics you made specificly to complain, cry, bitch and whine. It's not even including all the posts you made on existing topics where you blame everyone and everything for what is wrong with the world.

[Nixia]

you dont have to balance 2v2 if you balance 1v1.

Ofc they have. One class could be crap in 1vs1 but very very strong in 2vs2.

Here is one example:
Sl/sl warlock in tbc, they could beat any class in 1vs1 but where still balanced in 2vs2.

[CSJenova]

Here are some examples of your whining.

Oh wow...my personal favorite was the comparison of Starcraft balance to 2v2 balance...but that is quite a line-up right there.

[Mysmere]

close enough as the blizzard term isn't close enough.
there are massive margins everywhere.

i believe the best way to balance a class would be to do it through the calculation of scenarios. aka "what does this class need to counter someone casting" or "what does this class need to counter evasion"

theoretical dmg output should remain the same for any dps spec. the difference between melee and ranged would simply be a lowering of dmg output by the ranged when the melee is on them.
you start with the duration of the fight you want and lets say on average you want a ranged dps to be at range for 60% of the fight.
you simply construct scenarios until the ranged has the exact correct combination of abilities to do this against the melee's counter abilities.

after you have achieved this, you give them abilities to enable the correct dmg ratio for ranged and melee, and health ratio factoring in heals, auto heals, and mana (you'd want them to be x% mana at the end, assuming they use the correct abilities, and oom if they didnt)

giving any class cc that breaks on the first hit is a good way to make teamplay completely irrelevant. cc until win. a decent counter to this would be to simply make it break when the cc'er damages anyone at all. goodbye cc until win.
giving any class an ms effect should only affect dmg that person does (goodbye sit on the healer until win) and would be part of their utility aka no interrupt
burst is just a fallacy created by over scaling of dmg to hp and would be completely removed if pve were balanced around pvp.

you are correct in saying pve balance is the easiest to obtain. you can simply base pve around this balanced pvp.
yes, itd be hard, but last i checked programming/development is a job, and jobs should be challenging.

[khlg]

why do people cry so much,people who complain so much
are ruining the game for both pvp and pve people by crying
when they lose to a diffrent class then start posting threads on
websites like this crying for nerfs to the class they lost to.
If they try nerf some class for pvp that will affect pve people
even if it is a small nerf for the class in pvp could be a big
nerf to pve people.

so stop crying about it its a game you dont have to play it
if you dont like the way blizz do it play a diffrent game.........