Agile Game Development?

[saricden]

Hi everyone,

I'm starting to work on a new game that will be open-source with a blog component sort of coupled together with the game, and I was wondering if people take similar approaches to developing games as they do more general software.

With games (in this case an adventure-RPG platformer), there's a ton of stuff that needs to be done, such as:

• Plot development
• Character profiles and world design
• Making actual assets such as sprites, tilesheets, and (although I think this can be added last) music & sfx
• Designing special gameplay mechanics (which I think can be inspired by aspects of the plot, ie: cool items that give you powers that make align with the context of the story)
• Implementing basic gameplay stuff, ie: maps, preloading, physics, menus, inventory system, tilesheets, world travelling, etc. (MADE WAY EASIER by awesome frameworks such as Phaser )
• Implementing the special gameplay mechanics

There's certainly more than what I listed above, but the point is there's a lot to do. While a waterfall development approach would work, wherein you do everything step by step ordering stuff by what depends on what, I've found in past experiences that it's pretty easy to get bogged down in this approach. There's a good chance that this is because my favourite (and I'm guessing a lot of your favourites as well) part of the whole process is the development.

That's why I've been thinking that a more agile workflow might be a good way to keep yourself engaged with your own project (instead of jumping around from new project to new project, without really getting much done). An example of this might look like:

1. Not 100% sure what the entire plot is yet, but I know what the main character should look like
2. The MC is going to need to walk around in a 2D sidescroller environment, let's code that w/ some dummy sprite
3. Alright! Working well with a dummy sprite, time to animate a decent sprite of our character
4. (3 weeks later) Phew, those 20 frames were tough, but we've got it! Look how awesome that MC looks!!

And you continue to sort of compartmentalize different parts to keep up your velocity.

As I'm writing this I'm starting to realize it might be kind of nuts to go it alone. Maybe I should rename this thread to "anyone wanna collab..?"

TL/DR: What design-development approaches have you taken when creating a game? How did they work out? How have you kept yourself organized & engaged while creating a solo project with many components?

[b10b]

Hey fellow Islander!

By way of getting the ball rolling ... the first question I ask is "Do I want to finish this game?".  If not, Agile is the way to go

Whereas if completion is more important then I will need to impose many restrictions to my resources - specifically time, costs, features.  All three must be defined before the game can be declared as "finished".  Working backwards is quite easy ... with a deadline, finite resource, minimum viable feature set I can quickly work out the milestones, the epics and stories needed to get her done.  I can create a GDD and scope that forms a reliable reference, and using an established framework and toolchain there are few technical surprises.  From there the finer details (implementation & creativity) can follow an Agile approach as best suits the team.

 

[saricden]

Makes sense! 

The way I'm thinking of trying out would in theory be without deadlines and would be quite open to feedback from beta testers (totally open betas to anyone who wants to play).

Idk if this would still be considered Agile or if it would maybe be some weird middle ground between the two, but I'm thinking specifying some decent goals / milestones in some GDD-esque initial document (or probably in this case, a blog post) would be a good way to work towards a complete game. Doing something like that would also probably help identify any requests / feedback that's wildly unrelated, or is pulling the game in a drastically different direction.

I just posted this topic here that better describes this imaginary model I'm on about, hopefully that'll help give a better context lol 

[Jordgubben]

Quote

agile

(verb) To work on something when you don't understand the problem space, the market, the actual goal of the product or really any engineering at all.

I think the term "agile" is  becoming a buzzword. As such it is gradually loosing all value as a usable concept.

But there are at least three underlying principles that are worth salvaging. They are all valuable in isolation, but more importantly they tend to complement and strengthen each other:

Eliminate uncertainties. This one is critical if you are making a new game. You do not know how your game works or how your players will react to it.

Deliver value early. Getting feedback (and money) earlier is always better then later. We are gradually seeing this even in the AAA industry.

Self organization. Good news. If your a solo project then you get half of this one for free