Gamedev Diary: Testing All The Way
Going back through the earliest conversations about the development of JOKKA! Made me realize that… they were conversations. I didn’t know how strange this was until I started reading and researching gamedev, because inevitably there’s a chapter about alpha testing. It always starts with something like “Here’s how you find alpha testers!” and “here’s how you collect data!” and “here’s how you reconcile user feedback with the game you’ve spent months developing in your fortress of solitude!”
What was I doing from the very beginning? “Look, here’s my toy! It’s online! Come poke it!”
Obviously, this is an artifact of having chosen to develop a web-first game; it’s a lot harder to have this kind of interaction if your game is desktop or console-based. But I did what I did, and because of that, from the very beginning, I was working with alpha tester feedback. There are conversations with 10+ people asking me questions, reporting something weird, telling me what would be fun and what would be awful, reacting to my plans. This is, I’ve come to understand, not the usual way you develop a game.
It was a lot of fun, though.
The biggest and most obvious pitfall of doing this is dilution of your vision, which is just a fancy way of saying ‘you let other people pull you in too many directions, most of which have nothing to do with what you wanted in the first place.’ But I am stubborn, and I never felt obliged to implement every suggestion I heard. I picked and chose the ones I liked, discarded the ones I didn’t, and listened to the problems and concerns with particular interest.
But what a gold mine user feedback is!
What would the game have been like had I not been operating with continuous user feedback? I have no idea. But I probably would have created something cool, sprung it on a bunch of people, and then gotten swamped with everything from minor issues (“I can’t make a name with an apostrophe in it!”) to game-breaking (“this is boring”).
To wind things back in time, the early toy involved breeding characters and seeing how the offspring turned out. This was the point where I confronted the question: “So what?” Many pet-collection/breeding games have more mechanics than ‘pair off two characters, get a baby’, so what did I want to do afterwards? I had multiple conversations with my players about this; I said in one conversation: “The recursive loop was always “make more puppets and dress them” but I begin to worry about the art workload.”
Which led to “…it would have to be fun in some other way besides breeding at that point.”
This, I think, was the moment when the game design began to pivot. Looking back at it now, I thought I was building Flight Rising, Wolvden, Lioden, and would continue to think so for a long time.
What I’d actually be building, though, was Dwarf Fortress.
But for now, it was still a toy.
“I have a clan page that loads everyone, lets you rename, release (delete), and click-to-breed. Clicking on the character takes you to the character screen. And then there's the cavern screen where you can do the actual breeding and getting the baby. And all of it works! I guess even if it never gets much more complicated, it might still be fun. Like... maybe I can think 'it's okay to settle for wordle, you don't have to build multiplayer MMO scrabble.'”
Yeah, yeah. Keep telling yourself that, Jaguar…!