More testing needs to be done on new abilities before they make it to live

The Ecdysiast·Oct 12, 2017·1 votes·205 views

That includes reworks, updates, and new champions. Mainly, there needs to be more testing for bugs. I'm sure they've got a team available that plays the champion, but it's not too much to ask that they test out their interaction with all abilities that may have an odd bug.
Then there's things like the visuals compared to the hitbox. Some abilities even have a strict outline but don't fall inside it, like Galio's R or Illaoi's tentacles. And I know this is more of an issue of abilities that require their edge hit the enemy's edge rather than their center, but there's no pattern at all for which abilities do that.

5 Comments

The Last BalladOct 12, 20172 votes

That includes reworks, updates, and new champions. Mainly, there needs to be more testing for bugs. I'm sure they've got a team available that plays the champion, but it's not too much to ask that they test out their interaction with all abilities that may have an odd bug. Then there's things like the visuals compared to the hitbox. Some abilities even have a strict outline but don't fall inside it, like Galio's R or Illaoi's tentacles. And I know this is more of an issue of abilities that require their edge hit the enemy's edge rather than their center, but there's no pattern at all for which abilities do that.

Unfortunately, there is no way to do this. Remember on Kayn's release, where if Leona died while being pulled by her e she would be able to walk through all walls without restriction?

How on earth were they supposed to anticipate that.

It just isn't feasible to test a skill out against every other skill in 138 champions, for every situation. So they don't, and they instead test out what occurs to them/most obvious interactions and fix bugs on things that pop up.

The EcdysiastOct 13, 20171 votes

It's definitely possible to test, ESPECIALLY something like that.

My post suggested interactions with other abilities where a bug is likely to occur. Your example is actually an obvious one. Abilities that follow other champions through all movement. Warwick's Q and Leona's E are the only examples, and testing them dying isn't hard either.

I'm not asking that they test every possible situation, but obvious ones like that are left to go to live when they should've been caught. You do mention that they test obvious ones, but apparently they don't.

The Last BalladOct 15, 20171 votes

It wasn't just them dying, it was them dying while in transit of their dashes(which is rather rare). Not to mention, that was brought to be because of how terrain pass though was coded to allow Kayn's e to exist. So it wasn't even something that directly involved Kayn being in that specific game, just him existing in general.

And I will repeat,

They cannot test interactions that they don't think of. A Leona being able to ignore walls for an entire game because she died during her e isn't exactly something you can anticipate.

Look at the Mord bug that has been going around! How on earth can you even conceive that a champ will replace another champ's ult because of a revive passive? And it was only discovered recently on live, despite having the capacity to have happened for a while now.

The EcdysiastOct 16, 20171 votes

Henceforth, they have no excuse not to think of an interaction like this.

The Last BalladOct 17, 20171 votes

I don't think you quite understand how difficult it is to predict the bountiful yeilds of spegetti code.

The fact is that Mord's ult is replacing other champs ults makes no sense.

Just because you write something doesn't mean you will understand nor anticipate how every bit of code will interact with other portions, and the play testers? They don't even write the code. They are hired based on their ability to play the game, not on their ability to write and understand code.

Just because they do this for their jobs does not mean they have a special ability to anticipate what things will cause problems, and probe for them(because guess what? they don't necessarily know coding!). Just to document all of Morde's bugs, the guy that has been doing that had to play hundreds, if not thousands, of hours of Morde, and he is a dedicated fan. There are 336 hours in a pbe cycle, 80 of which are business hours, when the play testers would be doing their jobs(assuming they do 40 hours a week). And 138 champs that can be potentially be affected by the changes on the pbe. If they were to be testing a champ's interactions with other champs, including things like "what happens if x champ kills y champ with z skill, while champs h,k, and t are in the game?" there just wouln't be enough time to check all the interactions! For the next champ, we are talking about 13,906,484,270,000,000,000 possible team combinations! How the hell are the playtesters supposed to test that many potential interactions? HOW?

It is far more efficient to test the obvious things that would occur to people, and let the millions of players find the rare bugs that have effectively 1/1,000,000 chance of occurring due to how specific their requirements are for them to spawn. Because spending less than a millionth of a second testing each interaction sounds like it would accomplish a whole lot of nothing.