As a team, we're driven by inspiration. We feel inspired and excited to make a change, and so we work in short spurts to implement what we're excited about. We work on a feature, make sure it works, and then it's done and really we want to get it out as quickly as we can so people can start playing with it, because that's the really rewarding part of it all - there's no better feeling than to see players interacting with your creation and enjoying it.
We experimented in the past with having a 'feature design subforum'. The idea being that any changes which had a mechanical impact should be posted there first for review before they went live. Developers weren't keen to adopt this idea because it added a significant hurdle to their work, and often it would devolve into bikeshedding about inane and irrelevant facets of the change which completely killed inspiration.
Then we experimented with a 'balance group', where a number of players were invited into a private forum to offer feedback on changes before they went live. We encountered two problems with this. One, developers were unhappy with players holding authority over what changes went live, and two, the requirement to post in that forum was optional, as was the requirement to follow that feedback, which ultimately rendered the group (mostly) useless.
We've moved away from the idea of these two forums, though the feature design forum does still exist and we encourage its use.
Now we have a vague concept of a 'mechanical lead' - e.g. me. The idea is that any important mechanical changes should be run by me first. I don't hold the same interest in mechanics as Peppermint does, nor am I as experienced in that area, but I can generally spot bad things before they happen and I can filter them out.
In the case of 1) Feylock changes (before I fixed them), 2) PDK balance, and 3) mage BAB changes, ultimately it was my responsibility and thus my failure that they went live in their original state.
Change 1) went live because I rejoined the team on the same evening, and the change had already received approval from the other three admins, so I offered some minor feedback (which was actioned) and assumed the rest of the change was okay. I should have spent more time inspecting the change before I approved it. This is why I felt responsibility to fix the class shortly after the original developer left the team, and why the minor quibbles in the feylock feedback thread made me so stressed (this isn't MY mess yet I have to clean it up!) and eventually why the thread ended up locked after I unfairly let my frustrations out.
Change 2) and change 3) went through because I was very busy at the time and hadn't found the time to sit down and scrutinise about the changes, so I gave Mithreas the go-ahead.
Change 2) also received significant feedback from the balance group (when it still existed) - feedback which wasn't actioned. I don't blame Mithreas for this because sometimes it's very difficult to tell when somebody is debating about minor trivial opinions versus balance shattering truths, and I doubt he wanted to spend his limited time debating the finer points of game balance anyway. It was my responsibility to read over the changes in detail and say 'no' and then propose fixes. On that note, Peppermint and I are working hard to balance the path, and we will be making some changes in the near future.
Change 3) shouldn't have gone through. Irongron and I were ambivalent, and Dunshine wasn't a fan of the change. At the time Dunshine was busy and there was a miscommunication which led me to believe that the other two admins were fully aware of the proposed changes and were okay with them.
Seven Sons of Sin wrote:Again, I'll point to Overwatch. You know Jeff Kaplan gives the uodates. you know he speaks for his team, but addresses the arguments within the team. But you know when things are coming down the pipe because of their Public Test server etc.
Jeff Kaplan also has access to a team of hundreds (thousands?) of QA employees who are paid to test his game, and the resources to create and maintain a Public Test Server through which all changes flow for weeks or sometimes even months before they hit the live servers. Even in these massive, well-funded, AAA projects, broken or silly changes regularly hit the PTR and are later changed or reverted.
I think it would be wonderful if we had a PTR and a fantastic transparent change process and all of great stuff. Thing is - we aren't a massive AAA project. None of us are paid for our time. I don't think anyone has any motivation to set up and maintain a PTR, and that point presumes we have the time in the fisrt place to manage such a thing properly, even if we did have the motivation.
It might be better for your sanity if you consider Arelith to be -the- PTR. Sometimes we'll make mistakes and then we'll correct those mistakes when we have to. It sucks when players get caught in the crossfire, and whenever possible we're trying to mitigate that, now more than ever - see the relevel system we implemented and how they were distributed to feylocks as proof of that.
So lots of words to say this: we're doing the best we can with what we have.