Here’s a post-mortem for the Minotaur entry, a platformer we made as a team of 3 based in… France? UK? Australia? Ehem let’s say “Earth”.
This entry is the 3rd one I made with fellow programmer Manu, and 5th overall, so one good thing we can say is we’re quite getting a grasp on how to manage our time/energy and the scope of the game. We already knew the engine we were about to use (Phaser), and most of the tooling around it (TileEd, etc.), plus we’re starting to have a sense of what can be done in 72 hours, to avoid scope issues.
The week-end was far from being uneventful though, so here’s a few thoughts about specific topics.
Physics-based platformer? You’re gonna have a bad time.
If there’s one thing we didn’t realize when choosing the idea, it’s that we needed to actually learn to work with physics libraries.
Up to now, we always used Phaser’s “Arcade” physics engine, which is pretty basic and works well for simple things. But we quickly realized that it was nowhere near the requirements for a game where every element of the level can actually move. More precisely, the Arcade engine doesn’t handle well situations where more than 2 bodies are colliding together (ex. : cube stack).
Yay…
Conclusion: we reluctantly switched to the more powerful P2 library, that’s also packed with Phaser. That’s when we left the comfort zone…
BETTER THAN GOAT SIMULATOR! PARTY HARD! GET HYPE!
Fortunately, getting the basics to work took less time than expected. However, when looking back at it, we overall spent hours and hours of tweaking all through the week-end to make the gameplay actually stable and fun. The hard part, with physics libaries, is to actually understand which lever solves which problem:
- Having troubles making walls slide past each other? Lower the friction.
- Don’t want the player to be able to push stuff just by walking against it? Lower its mass.
- Want to force the player to follow a moving wall? Create a constraint.
- Want to make sure walls don’t move an inch until you grab them? Make them static.
Not mentioning the various bugs that took hours to solve, be it about world bounds, collision groups, detecting collision directions…
But in the end we actually got it to work, and that was incredibly rewarding. There’s still a few glitches but the game is mostly playable, so we’re now proud to say: YES, we know how to make 2D physics games!
Graphics vs. Level design
One of the main events for this entry is we teamed up with Jack, a talented graphic designer we met the day before on /r/ludumdare. It was pretty improvised, but still a great opportunity, since “Graphics” is the one category where we’re always struggling the most. As a pair of programmers it can be hard to make good-looking art… Plus given how long art can take to make, it’s often been hard to find the good compromise between programming more features/contents and making better art trying to make better art.
Working with a teammember dedicated to graphics was mostly new to us, and we couldn’t help notice how it changes the dynamics of game dev a bit: suddenly you have to discuss upfront how the game will look, what are the capabilities you’re looking for, which limitations you can accept, the dimensions of the tiles… And then stick to that. Plus, even if you try to thoroughly discuss the theme/style you’re looking for, you’re suddenly blind to a big aspect of the game, being how it will actually look.
Well, until you actually get some art:
The first artwork Jack shared with us. Exciting!
That first artwork was really exciting, and helped us adjust the story, figure out the shaders we’d need, and in general plan things out a bit more precisely. And of course we were motivated by knowing how good the game could look in the end!
That’s about when we made a little organization mistake: what would have been best at this point was to get a tileset done so we could start designing levels. The focus was instead put on the character animation, which took quite some time. It was totally worth in the end (try the game to see it, importing a gif wouldn’t do it justice :P), but the side effect is we started really late the work on the final levels.
On the good side, the tileset was pretty nice in the end, and a great moment occurs when you finally integrate the sprites into the game. Suddenly everything looks 100 times better! Yay, I’m making an actual game! After that, we eventually managed to produce enough content to make the game enjoyable, but there’s still some improvements that could be done about the levels. Mainly:
- I do believe some aspects of the base mechanic are underexploited ;
- We didn’t fully take advantage of the art provided by Jack, as most backgrounds for instance were patched together in a hurry.
The final level 2 background, hacked in less than a minute, half an hour before the deadline. Sorry Jack.
In the end I think the game still looks pretty good, and we definitely couldn’t have reached a such level of polish as a team of two. When comparing the game with our previous entries, the difference in art quality is pretty obvious I think.
About the game idea…
I’ll end the post with a little confidence about how we found our idea: the week before LD, we actually brainstormed ideas for most of the themes of the final round (here’s our file, beware it’s in French). I’m not sure how fair it is but there was that post featuring a ranking of the themes after the last round, and because of that we were almost sure that the theme would be either “Companion” or “An Unconventional Weapon”. I’ve never seen a such post before but it was definitely a bit of a spoiler I think! So we did have some ideas in advance.
On the other side, the telekinesis mechanic was an idea for the “You are the Power Source” theme, not the final one. Since I really wanted to try that platformer game, we eventually chose it over the other ones, and plugged the theme on the story more than the mechanic. But hey telekinesis is actually used to destroy robots in the game, so technically it is an unconventional weapon!
Thanks for reading this, if you haven’t already we’d be glad to get your feedback on Minotaur. And congrats to all contestants, we’ve played some really awesome games already. Cheers!
