Does ArenaNet actually "test" anything before they release a skill "balance"?

Here's a list of bugged skills that were within in the last two major skill balances: December 2008, and March 2009.

So, let's see how many skills were actually changed in the last two skill balances:
40 (December 2008) + 19 (March 2009) = 59

Let's see how many skills had functionality changes:
37 (December 2008) + 4 (March 2009) = 41

Let's see how many bugs were the result of functionality changes? Oh wait, all of them.

8 skills.

8/41 is around 20% (19.51...%). Around 20% of their functionality skill changes are bugged.

So, I have two questions to ask:

1. Does ArenaNet actually test anything when they release a skill balance? Or, do they just release something to us and have us test it and find the bugs for them?
2. When they say "we don't have enough time to test", when all they need to do is change numbers around (which has a bug rate of, for the past two months, 0%, AND they are literally spoon-fed the exact updates they need to do), would you believe them?

20% bugged would be an amazing record with literally zero testing.

I actually don't, if every Skill is tested, but "big updates" like the VoD-Update, Zaishen-Update, the implantation of the AT's had been tested on a Test-Server with Players from the Community.
Of course it is impossible, to test everything, especially now, while nearly everyone works on GW2.

From 8 years working in games dev I can tell you that even with large teams of dedicated QA working round the clock its almost impossible to eliminate all bugs from a large complex game system. You concentrate on the crashes and vulnerabilities, then the game blockers, then, way further down the list, comes the notion of 'balance'.

Not saying thats right or wrong, but its the way it is when you are under time and money pressure from the bean counters at the top who don't give a flying **** about a few guys QQing on a forum or whether you can farm 2 or 20 ectos per hour with a skill set.

They also shadow update fairly often. "It's Just A Flesh Wound!" used to give you energy for the target and the user, even after the functionality update. They fixed it recently but no update notes for that.

Many of the skills you listed had fairly minor issues. For example, most of the hexes didn't cause degen until the update, and that simply got overlooked. The game has handled rounding in the same fashion for years. Mark of Insecurity making short-duration stances and Patient Spirit simply not work was a problem with the skill's new mechanics being retarded, not a bug itself.

In fact, the only bugs on that entire list that I think were actually problematic were Recurring Insecurity and Cultist's Fervor; those were pretty major oversights by QA. The rest of your complaints are nitpicky; relevant, sure, but not quite as major as you seem to be making them out to be.

Meh, you have to consider they cant afford to work much on the game anymore, its going to be quite a bit of lying to keep the community at bay - thats about it id say

Honestly though, how can you not notice a degen hex doesnt make the bar pink if you are testing skills..

Mark of Insecurity was incredibly bugged. It made enchantments/stances expire 90% faster, which made things expire much faster than Air of Disenchantment which made enchantments expire 300%? faster. So, there was a clear misunderstanding and a failure in testing. That skill should have probably been the most tested skill. I would even say that the skill needed to be tested with every enchantment and stance in the game before being released (which takes like, what, 1 hour?) to make sure it interacts correctly. Clearly, this didn't happen (lol Patient Spirit).

Testing is not hard. You read what the skill says, and you see if that's what the skill actually does. Then, you look at the side cases and try to exploit it. If ArenaNet was a smart company, they would design a special map just for skill testing. Or, they would let AIs play with the skill and see what they do with it. AIs look through all potential targets of a skill, and if someone sees an AI using Signet of Ghostly Might on another ally (as when it was incredibly bugged), it would probably raise a "holy shit, we need to fix this."

Not turning bars pink is a clear indication that they never really used the skill on anyone, or absolute retards are "testing" the skills (either is quite plausible). The pink bar is pretty damn obvious. That kind of oversight just makes me wonder if they even put it on their bar and click it. Ever. I wouldn't mind if they forgot things like "Lingering Curse doesn't work correctly with Aura of Restoration" - that kind of thing, even though it should never happen in a real game, takes more than just 1 hour of testing. All these bugs are the FIRST thing I would notice about a skill. "It's just a flesh wound"'s applying speed boost even without conditions is probably the first thing I would have tested. No pink bar on a degen hex? That's as clear as glass. Come on, this shit is obvious.

If you really believe they make changes without testing then there is no point being here and little point playing the game.

I reckon the situation goes something like this.

The number of skills in the game is huge a couple of people working on making changes to those skills can only check so far and can only put so much time into testing.
Mistakes will be made and assumptions will be wrong.

To put it another way there are maybe 4 people working to fix problems and many thousands working to create them.
I assume Anet are well aware of this problem and will make GW 2 more robust in this area.

You know very little about "testing". They have hundreds of possible combinations, it's not as simple as reading a piece of text and testing whether the skills actually does that, if it affects enchantments you have to try it on dozens of enchantments with many different professions and considering all other enchantment-affecting situations, you have to combine the skill with other skills in various order (test that it does what it's supposed to do and doesn't do what it's not supposed to do), repeat previous tests done for previous updates to see whether the new update hasn't broken something done before, you have to plan for the unexpected in terms of game mechanics by considering the skill changes from a purely software point of view (I guess devs suggest what to test and it can be VERY different from reading the skill text). It does NOT take the "1 hour" you quote, tester is a full-time job, 8h a day (more when you're closer to release), 5 days a week. As Regina said recently (about Anet hiring testers for GW2), it's an extremely repetitive task that requires meticulous and precise execution, reports must be written where you check hundreds of boxes and you're expected to improve the testing techniques constantly.

I don't think you read my post. I don't care about those small interactions that you might come across very few times. Even though I expect most good games to fix those, I don't expect ArenaNet to.
I expect them to fix the clear, obvious bugs. Which, they don't.
And, I do know QA pretty well. I did QA for a summer in high school for Blizzard. See, we spent hours on one skill, going over obvious stuff first. ArenaNet doesn't even do the obvious stuff.



Here, ANet, I'll spoon feed you exactly what you need to do for the update (only number changes, so it's easy):

Aegis: 30s -> 45s recharge
Warrior's Endurance: 14 second duration
Strength of Honor: 25s recharge, reduce +damage by 4
Lingering Curse: 10s duration, 15s recharge
Peace and Harmony: 0...3 hexes/conditions removed (3 at 12 divine), remove 90% duration reduction, 3/4 cast, 5s recharge
Mind Blast: 5s recharge, -1 energy on the energy gain
Aura of Restoration: Require 10 Energy Storage for 1 energy gain (Mind Blast eles lose 4s Distortion)
Vampiric Spirit, Vampiric Gaze, Angorodon's Gaze, Unholy Feast -> reduce life stealing by 15-20 for each, increase recharge by 4s.
Mirror of Ice: Smiter's Boon this. or at least: 20s duration, 40s recharge, reduce damage by 5.
Glyph of Renewal: 15 or 20s recharge
Power Attack: reduce damage by 3, recharge to 4s

I did. But you didn't really understood what I was saying:

the QA guys have thousands of things to test, going through a very lengthy and painful process; they may have not noticed some bugs due to the time taken to check all their stuff; they may also know the bugs you mention but the devs may not be able to fix them with a quick fix (more likely).

Testing is a cruel task, you can't get directly rewarded, since doing your job nicely means nothing happens ingame. No one come here to talk about the several hundreds of skills with each various aspects (time, energy, features) working perfectly well.

You must be a superdev! It's well-know that the skill numbers are in a big file, so it's just a 30mins change! Well done.

Sarcasm? If so, assuming changing these values (provided the guy gets them spoon-fed like this) will take longer than 30 mins is really ridiculous

sure they test it. how extensively? lol, who knows. prolly 2 guys working on it by themselves.

No, the ridiculous thing is that you think they would just take your word, change the numbers on all those skills, and not do any testing. I see why we need a storage update; some of the players around here have such enormous egos, they just can't fit anywhere right now.

Ask yourself this: when you were doing your internship at Blizzard, did you have more than 2 people working on QA? I'm going to assume more than one full-time employee was helping you. If so...what the hell do you know about ANet's process? I'm sorry, but if you have a QA team of 2 people, you can't reliably spend hours on one skill, as you say.

bro - first of all, blizzard probably had 100 other interns just like you who were all working around the clock on skill updates.

second, stop bitching about skill bugs - the vast majority of the ones you pointed out are trivial, at best.

send in a report, then step away from the computer if it bothers you that much.

Alot of that is nitpicking skill descriptions, which are hard to fix because they need to be localized to all languages.

E.G. MoI- the mechanic reduces enchant/stance duration, not increasing expiration. It reduces by 90%, leaving 10% of the original duration.... or in other works, 1000% faster expiration. It was always intended to work this way, that's how it functioned when it said 50% faster expiration, and the current version still does: the description just isn't terribly clear. Peace and Harmony is exactly the same.

This is no different from Frenzy saying "you attack 33% faster", when it's actually a 33% frame reduction, which really translates to 50% faster. This is just one of many points on which the skill descriptions are inconsistent - that doesn't mean the skill mechanics aren't working as intended.

Of course, very nice skill balance, stomp every second skill to the ground. Now read again, skill BALANCE is the name for a reason.
Half of these skills deserve a nerf, but not like that.

Actually it sound more like you don't know anything about testing. Each of those combinations and variations is just a test case, sure, they have a lot of them, but they are just items in a test plan, most of which should be fully automated or even ran as unit tests.

What the hell has the size of a file got to do with it? Most modern programs are huge, but the code, data and resources are organised so that they are easy to find and change, otherwise nobody would be able to change anything!

Of course, if the code is badly organised, and there are no formal unit tests or automation, then yes, very hard to make changes and test thoroughly. So the high fault rate would indicate either poor/lack of software testing, or poor dev practices. It could just be down to lack of resources. Either way 20% fault rate on trivial changes is quite startling.

it's a common pattern with anet in general: excellent ideas, seriously lacking execution. the entirety of GW is testament to that.