Summary : "MVP" = Minimum Viable Product. Our development cycle centers around keeping a tight feedback loop with our players, mostly via our Discord server, to hear what they want most in game, and iterating on that feedback every week.
[Writing about DreamWorld's Weekly Release #2, on 08/30/2020]
Dreamers! Cheers and welcome back 😄
Last post we touched on a design philosophy we affectionately and humorously call "Minimum Viable Product (MVPs) all the way down". The phrase MVP is very common in software development, and it's one that's helped us bring DreamWorld's denizens features faster.
Let's illustrate this concept by looking at some practical implementations. In our second weekly release, we had changed the entire map, added a day/night cycle, added a menu with some basic functionality, tweaked atmospherics, and were planning community get-togethers to play and chat.
Why had we gone after this handful of features for that week's release, and why did we push them in the form they were in?
[Also, unrelated, but for the best reading experience, go put on Black Clover's "Black Rover" by Vickeblanka while you read. I listened to it on repeat while writing this, it just totally caught me by the soul.]
- Changing the entire map - This seems like it would be an intense, at least multi-week project, right? So how and why did we change the entire map in under one week? We realized two things; first - our players understood that DreamWorld was not a finished product and would not be for a long time. This was going to be a project they got to have say in every week of its development for years. Second - this change was very quick to make as the map was so small and bereft of things like foliage or almost any other meshes. The map being so small also made it so that the adventurers had already covered every inch of its first form in several hours of play. Changing this would hopefully keep them engaged. Combining this with moving the EEs and several other landmark assets in fun ways kept players looking for cool things for several days in this second week, hoping to be the first to report finding them in the Discord server.
Now here's an important note - we already had the vision for a fully destructible, sprawling map with mobs, tons of materials to build with, and a central city. Why didn't we wait to release new versions of the map until we'd had all that in there, say, a year later? We wanted our players to go on this journey with us, not experience its development asynchronously. We wanted to know what DreamWorld's players wanted along the way so we could use their input to steer towards the most important elements more quickly. What was a fad? What was evergreen? What was most unique? What was totally worthless? We hated the idea of not knowing any of these things until it was too late to not have spent significant time and effort on them. Players have constantly voiced LOVING looking for bugs and ways to get out of the map, even to this day, and it's been extremely helpful. Oh and the map change, with its bald hills, bad textures, and Dr. Seuss - looking inclines and ravines? No one said anything negative about it. In fact, we've had people recently tell us how nostalgic they are for the OG maps 😂
- Adding a day/night cycle, atmospherics : Here was our first stab at trying something unconnected to development pillars (a future blog post for sure) that we thought was going to be intriguing to our players... and we were wrong! Our posit was that a day/night cycle would add visuals that would excite players to stay in DreamWorld longer. Two things happened - 1) we hadn't tested a packaged build before we sent it out to players to try, and there was one massive difference between the editor and production versions. The day/night cycle was about 10 minutes long in editor, and 15 seconds long in the packaged build. Hilarity ensued as people logged into DreamWorld to a veritable visual nightmare. It felt like you were on an asteroid spinning like a top, and looking away from the sky helped only minimally, as searing brightness turned to shadows that shot like horrific outstretched fingers at you before they plunged you into a dark so inky you couldn't tell which way was up. I believe a hotfix went up within a day. 2) When we'd fixed it, not only did almost no one care about the day/night cycle, but the only people who did care were people who said it annoyed them because they liked their builds better in the sunlight.
Let's apply the 'Tight Feedback Loop + ASAP Iteration = Magic' brush. Rather than worry about fixing the day/night cycle any more than we had, we scrapped it all together. Feedback was negative, we knew it wasn't an integral feature to our pillars, so it didn't merit maintenance. In light of many other features we needed to go after, the iteration needed was clear. If we had spent some more time on the day/night cycle and made it more beautiful, reduced nighttime for those who were vocal about liking their builds in the sunlight, might we have gotten an awesomely engaging feature in? There's still room to work there, but we have a pretty beautiful day/night cycle now, a year later (below). Not one person has cited it as a super important feature to them, and we still have people tell us the night should be shorter.
- Planning community Discord meetups : This one was the most obvious of them all. At the time, most of DreamWorld's players most enjoyed the community being built. The community you cultivate surrounding anything can be as or more important than the thing you create. I would argue that without the relationship-building and community that formed around World of Warcraft for example, it would have far fewer players than it still does today.
Family that we're born into and also that we choose may be the most important thing we have in our lives. We will always be aiming to empower DreamWorld's denizens to find new family, forge lasting relationships in the fire of awesome shared experiences, and write their legacies in the memories they create with the people they care about.
- Adding a menu with basic functionality : Feedback - tons of players in the prototype release and first week told us that the most annoying thing was not being able to exit the game via a menu button. They also didn't know how to play. Iteration - add a HUD widget blueprint menu on "P", a writeup there of controls, and an exit button.
We of course get that this was an obvious and easy add, but if nobody had cared that there wasn't a menu with an exit button and liked figuring out the controls themselves, we would have spent time on something else and it definitely would not have been put in the second weekly release. We're actually very excited to hopefully find truths about gaming with this methodology that may have been missed - imagine if we found out that players didn't care about weapons, and really wanted enemies to just die in fun ways when they got near them, as an example. You could structure the game's tactics around being skilled enough to get near foes for that guaranteed kill (wait! This is getting a rainbow mushroom in Mariokart!! 😂).
What you spend your time on is everything, especially in startupland.
MVPs all the way down.
- In the next post, we'll talk about player movement, and production platforms! -