Buffalo Joe: Post-Mortem

Posted by (twitter: @oreganik)
August 28th, 2012 4:43 pm

The Buffalo Joe LD 24 Compo Post-Mortem

A common question I get is, “What does Buffalo Joe have to do with Evolution?” That’s a good question! Because, I mean, all concepts evolve, right? And, well, the evolutionary concept at the heart of the game sort of caved in to outside pressure and, you know, didn’t survive? It’s actually sort of meta! Right?  =D

The Numbers

Ludum Dare Competition Count: Second

Tools: Unity (C#), Blender, Photoshop, Audacity, iPhone Voice Memo

Hours spent on:

  • Sleeping: 9
  • Meals with wife: 3
  • Exercising: 1
  • Solitary Eating Rituals: 1
  • Wondering about it all, man, like: 0.9
  • Fruitlessly Flogging Flocking Fixes: 2
  • Audio: 1.1
  • Texture/Model Creation: 3
  • ??? (missed time due to likely abduction by forces unknown): TBD
  • Fire (Discarded): 1.5
  • Level creation/design: 1.5
  • Remainder: 23 hours

Choosing the Concept

My wife and I went on a dinner date right after the theme was announced, and because she’s a bubbly font of creativity, we soon had a long list of good ideas. Ultimately, I had to decide between:

  • Polytoad: Evolutionary Frogger! Your body adapts to match the cars which crush you. Mack truck? Throbbing metal hindlegs. Speedbike? Tapered waist. Honda Accord? More beige!
  • Guy On A Buffalo: “Evolutionize” your flock in unspecified ways to meet unspecified dangers!

Really, in my heart of hearts, I knew Polytoad was the game to choose. It was Doable. But this damn song kept playing in my head:  Guy On A Buffalo! Right. Who could resist the Call of the Buffalo?

 F*cking Flocking

Right away, I made an LD n00b mistake by wasting time on tech I’d never used before. Flocking is a pretty basic computer science thing, but I was an English major. Oops. That time could have been spent on getting evolution in the game. Speaking of…

The Original Game Design

  1. A fire is raging across the plains. You must escape across the river to safety.
  2. There are packs of wolves / Indian hunting parties / Pioneer wagons barring your way. You can run them down if your herd is strong enough.
  3. Eat grass to get baby buffalo and grow them.
  4. Find special plants — Fireweed, Cloudpuffs, and Rustleaf — to start adding special DNA to your new baby buffs.
  5. These plants will evolve your herd into stronger, faster, more powerful animals, which you’ll need for later levels.
  6. You are personally a man on a buffalo who can shoot bad guys (slowly) and can’t get killed.

This had everything I needed:

  • A clear goal (get across the river)
  • A clear obstacle (enemies blocking the river will try to kill you)
  • A clear way to overcome the obstacle (grow the herd so you can overrun them)
  • Time pressure (fire, which was effectively a timer)
  • Dynamic pressure (wolves would spawn at an increasing rate)

The one problem I recognized — but hoped would solve itself — was that evolution was not necessary for this game design. It would make your buffalo stronger, helping you achieve certain goals, but it wasn’t necessary. But instead of solving it then and there, I stuck with production, hoping I’d figure it out.

Issues Encountered

  • I made sloppy states for my leader, buffalo, and wolves. This made finding bugs difficult. I ended up rewriting everything as components, making it more like a state machine. LESSON: Start with state machines.
  • I got fire working, functionally (burned plants, animals, moved across the plains), but there was no way to make it look good in 3D in the time I had. That was about 60 to 90 minutes wasted right there. LESSON: Make sure game objects can be visualized in your game for a low cost. Cut them if they can’t.
  • I spent a lot of time getting wolf and buffalo interactions to feel more life-like. In other words, buffalo would kill wolves if hit head-on, otherwise wolves would hurt buffalo. And wolves would eat dead buffalo, so you could sacrifice a buffalo to get away. Oh, and baby buffs would be weaker, right? And maybe couldn’t kill wolves… TLDR: Too complicated. Did not contribute to high concept, nor was it required. LESSON: Core gameplay must be simple systems. Expand later.

Final Game Design

  1. Get your herd of buffalo over the deep canyon to escape the marauding wolves
  2. Do this by eating enough Cloudpuffs to give your buffalo wings, so they can fly over.
  3. Baby buffalo eat grass to make more buffalo.
  4. Wolves kill baby buffs. You kill wolves.
  5. Wolves will spawn at a continually faster rate, creating time pressure.
  6. Your base score is number of buffalo divided by time.

This was a great design for a Ludum Dare. Too bad I figured it out so late!

  • Goal is still clear and physical (get over there).
  • How to achieve goal is clear (fly).
  • Wolf / Buffalo / Leader interaction is simple and clear.
  • Resources make sense (grass = health, blue grass = power up)

Everything got in… except the cloudpuffs and flying! So then the super final game objective became:

  • Protect your herd while it eats grass to create more buffalo. Win when you get ten!

Ah well. Had fun, made some stuff, let’s do it again!

Ted Brown / oreganik.com


Leave a Reply

You must be logged in to post a comment.

[cache: storing page]